Effects of Fungicide Chemistry and Application Timing on Fusarium Head Blight and Deoxynivalenol in Soft Red Winter Wheat

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Daisy L. D’Angelo, B.A.

Graduate Program in Plant Pathology

The Ohio State University

2013

Master's Examination:

Dr. Pierce A. Paul, Advisor

Dr. Larry V. Madden

Dr. Mike A. Ellis

Daisy L. D’Angelo

2013

Abstract

Demethylation Inhibitors (DMI) and Quinone Outside Inhibitors fungicides are important components of wheat disease management programs. However, only members of former group are usually recommended for the control of Fusarium head blight (FHB) and deoxynivalenol (DON). Since wet, humid conditions during anthesis and early grain fill are highly conducive to FHB development, DMI fungicides are commonly recommended for FHB management when anthesis coincides with wet weather. However, there are several practical limitations to applying a fungicide at anthesis. Research is needed to determine whether applications made after anthesis will provide adequate control of this disease and its associated toxins in soft red winter wheat under field conditions in the

U.S. Midwest. In the case of the QoI, these are not usually recommended to FHB and

DON management because some members of this group have been reported to increase instead of decrease DON in harvested grain. However, it is unclear whether all QoIs affect DON in the same manner and whether DON response to QoIs vary with application timing. The objectives of this study were to: 1) determine the effect of post- anthesis applications of 19% tebuconazole + 19% prothioconazole and 8.6% metconazole on FHB and DON in soft red winter wheat (SRWW) under different naturally infected and artificially inoculated field conditions, 2) determine whether the magnitude of FHB and DON responses to post-anthesis fungicide applications varied with active ingredient,

cultivar, and baseline disease and toxin levels, and 3) estimate the efficacy (based on mean percent control of FHB index and DON) of post-anthesis treatments relative to untreated and anthesis reference treatments.; 4) determine the effects of two QoI active ingredients (22.9% azoxystrobin and 23.6% pyraclostrobin), relative to untreated checks

and DMI (19% tebuconazole + 19% prothioconazole) reference treatments, on DON

contamination of wheat grain and spikes; 5) determine whether the effect of QoIs on FHB index and DON was influenced by application timing; and 6) determine whether the relationships among index, DON, and Fusarium damaged kernels varied with active ingredient and application timing.

iii

Dedicated to

My Husband, Craig E. D’Angelo

Acknowledgments

I would like to thank my advisor, Dr. Pierce A. Paul, for his unending support and guidance throughout my time at OSU. His encouragement and instruction have added great input to this thesis.

My student advisory committee members, Dr. Laurence V. Madden, for his help with thesis revisions; and Dr. Michael A. Ellis, for his instruction on fungicide chemistry and management, and help with reviewing this thesis.

My lab mates Kelsey Andersen and Jorge David, post docs Jessica Engle and

Katelyn Willyerd, and lab technician Kristin Davies for their help conducting experiments and collecting data.

Bob James, for his help and support maintaining greenhouse experiments, and the farm crews at Snyder and the OARDC research stations for planting and setting up all field experiments.

My Husband and family, without their encouragement and support I would have never decided to return to school and pursue a graduate degree.

Vita

June 2002 ...... Reynoldsburg High School

December 2006 ...... B.A. Biology, Capital University

2011 to present ...... Graduate Research Associate, Department

of Plant Pathology, The Ohio State

University

Fields of Study

Major Field: Plant Pathology

Table of Contents

Abstract ...... ii

Dedication………………………………………………………………………………...iv

Acknowledgments...... v

Vita ...... vii

List of Tables ...... ix

List of Figures ...... xi

Chapter 1: Fusarium Head Blight of Wheat and Deoxynivalenol: A Review ...... 1

Chapter 2: Efficacy of Post-Anthesis Fungicide Application against Fusarium Head

Blight and Deoxynivalenol in Soft Red Winter Wheat ...... 14

2.1. Introduction………………………………………………………………….14

2.2. Methodology………………………………………………………………...19

2.3. Results……………………………………………………………………….25

2.4. Discussion…………………………………………………………………...30

Chapter 3: Influence of Fungicide Chemistry and Timing on Deoxynivalenol

Contamination of Wheat Grain by Fusarium graminearum……………………………..46

3.1. Introduction………………………………………………………………….46

3.2. Methodology………………………………………………………………...51 vii

3.3. Results……………………………………………………………………….57

3.4. Discussion…………………………………………………………………...62

References ...... 78

viii

List of Tables

Table 2.1 Summary of inoculation and treatment application protocols for experiments

conducted to evaluate the effects of post-anthesis fungicide treatments on Fusarium head

blight of wheat ...... 36

Table 2.2 Summary statistics from linear mixed model analyses of data from field

experiments conducted in multiple location-years to evaluate the effects of post-anthesis

fungicide treatments on Fusarium head blight index, incidence, Fusarium damaged

kernels, and deoxynivalenol in soft red winter wheat as influenced by wheat cultivar, fungicide active ingredient, and inoculum density………………………………………37

Table 2.3 Summary statistics from linear mixed model analyses of post-anthesis treatments effects on Fusarium head blight index and deoxynivalenol content of harvested grain………………………………………………………………………….39

Table 2.4 Log of the response ratio, percent control and corresponding statistics for the effect of fungicide treatments on Fusarium head blight index and deoxynivalenol in soft red winter wheat………………………………………………………………………….41

Table 3.1 Pairwise comparisons of untreated check and DMI reference treatments with

QoI treatments from linear mixed model analyses of fungicide effects on Fusarium head blight index, Fusarium damaged kernels, and deoxynivalenol from wheat plots spray-

inoculated at anthesis with a spore suspension of Fusarium graminearum under field

conditions ...... 68

Table 3.2 Pairwise comparisions of untreated check and DMI reference treatments with

QoI treatments from linear mixed model analyses of fungicide effects on Fusarium head

blight index and deoxynivalenol for wheat spikes inoculated at anthesis with a spore suspension of Fusarium graminearum under controlled conditions……………………..70

Table 3.3 Regression coefficients from linear mixed model analyses of the relationship between Fusarium head blight index or Fusarium damaged kernels as continuous covariates and log-transformed deoxynivalenol content of harvested wheat grain as fungicide treatments……………………………………………………………………..72

List of Figures

Figure 2.1Fusaruim head blight index for different treatments applied to soft red winter wheat in field experiments conducted at the Ohio Agricultural Research and

Development Center (OARDC) Snyder Farm near Wooster, Ohio in 2011 (A), 2012 (B), and 2013 (C); at the OARDC Western Agricultural Research Station near South

Charleston, Ohio in 2011 (D), and 2013 (E); and at the Illinois Council for Food and

Agriculture Research facility in 2012 (F) and 2013 (G) ...... 42

Figure 2.2 Deoxynivalenol content of harvested grain for different treatments applied to soft red winter wheat in field experiments conducted at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio in 2011 (A), and

2013 (B), at OARDC Western Agricultural Research Station near South Charleston, Ohio in 2011 (C), and 2013 (D), and at the Illinois Council for Food and Agricultural Research facility in 2013 (E) ...... 44

Figure 3.1Mean Fusarium head blight index (A, C, and E) and deoxynivalenol content of wheat grain (B) and whole spikes (D and F) ...... 74

Figure 3.2 Relationship between Fusaruim damaged kernels and log-transformed deoxynivalenol content of wheat grain (A) and between Fusaruim head blight index and log- transformed DON content of wheat spikes (B and C) ...... 76

Chapter 1

Fusarium Head Blight of Wheat and Deoxynivalenol: A Review

Fusarium head blight (FHB), commonly known as head scab, a disease caused by

the fungus Fusarium graminearum and related species, affects small grain crops such as

wheat, barley, and rye throughout the world. Robert W. Stack presented a succinct

summary of the history of FHB with an emphasis on North America in a chapter of a

book called Fusarium Head Blight of Wheat and Barley (Leonard and Bushnell, 2003).

He divided his summary into four eras. The first began in the 1880s, around the birth of plant pathology, when the disease was first being described and reported. The second era was initiated with a single event that occurred in 1908, when experimental reproduction of the disease by artificial inoculation was accomplished. He went on to describe this period as the “classical era”, during which pathogenicity was described and epidemics were more thoroughly documented. The third era, which began in the early 1950s, was described as the “dark ages” of FHB in North America, and it was thusly named for the lack of research at that time, even though epidemics of FHB were still being reported.

The fourth era began around 1980 and was described as the modern era. Stack believed that the so-called modern era of FHB was spurred by the identification of aflatoxin. This newly discovered mycotoxin was associated with human and animal illness, and naturally led researchers to investigate mycotoxins associated with other fungal species that could be hazardous to human and animal health. Concerns about mycotoxin contamination,

along with the development of new, sensitive laboratory detection and

quantification technologies, brought about a new wave of research, including screening

for resistance to FHB. Stack concluded his historical account by inferring that the early

2000s (the time his chapter was written) marked the transition of FHB from the modern to the molecular era.

FHB causes significant losses in many regions around the world. These losses are interconnected, and are due to factors such as floret sterility, smaller shriveled, light-

weight, and discolored kernels (collectively referred to as Fusarium damaged kernel

[FDK]), and grain contamination mycotoxins (Windels 1999). Floret sterility and reduced

kernel size and weight lead to a reduction in grain yield and quality, which ultimately

leads to grain rejection and price discounts. FDK also affects seed quality, resulting in

reduced germination, seedling blight, and poor stand establishment (Gilbert and Tekauz

2000). Of the mycotoxins associated with FHB, deoxynivalenol (DON), a trichothecene compound, is the most frequently occurring (Tanaka et al 1988). DON and other toxic metabolites produced by Fusarium species may cause alimentary toxic aleukia (Yagen et al. 1976, Rotter et al 1996), foodborne illnesses in humans, and feed refusal and vomiting, especially in non-ruminant animals (Goswami et al. 2004, Diekman et al.

1992). It inhibits protein synthesis when ingested, leading to growth retardation, reproductive disorders, and compromised immune functions (Rocha et al. 2005, Adams et al. 1989). Consequently, in the United States, the maximum allowable limit for DON in harvested grain is 2 ppm, and grain lots with higher levels are either rejected or priced down, leading to economic losses for producers (McMullen et al. 1997).

FHB and DON are also a major problem for millers and bakers, as they reduce the

milling and baking quality of flour produced from F. graminearum-infected and

mycotoxin-contaminated grain (Dexter et al. 1996). DON is a water-soluble and heat-

stable compound with minimal degradation at high temperatures (150oC), consequently,

it often persists throughout the processing and cooking of flour (Rotter et al. 1996,

Bullerman and Bianchini 2007). Millers are concerned about the effects of FHB and

DON on flour specifications, such as ash content, which is a measure of the mineral content and brightness of the flour (Dexter et al. 1996). FDK is often associated with darkened flour with high ash content (Dexter et al. 1996), which reduces its quality and market value. Bakers are concerned about contaminated flour during the baking process, as flour contaminated with DON can reduce gluten protein strength, resulting in over mixing of the dough and poor baking quality (Dexter et al. 1996).

THE CAUSAL AGENT OF FUSARIUM HEAD BLIGHT

Fusarium species cause head blight of wheat, barley, rice, and oats; stalk and ear rot of maize (Leslie et al. 1990), as well as diseases of non-cereal hosts, such as soybean, canola, potato, and sugar beet (Goswami et al. 2004, Ellis et al. 2011). Many species of

Fusarium are known to cause FHB, however, F. graminearum is the primary causal agent in most wheat-growing regions of the world. F. culmorum is the next most common causal agent of FHB, particularly in cooler growing regions such as Europe (Snijders et al. 1991, Parry et al. 1995). Other Fusarium species known to cause FHB include F. poae, F. avenaceum, F. sporotrichiodes, F. oxysporum, F. verticilloides, and F. sambucinum (Magan et al. 2002). These latter species produce toxins that are far more

potent than DON, however, they are less effective as pathogens on wheat, and do not

predominate in the Fusarium species complex (Parry et al. 1995).

Based on their toxin production profile, F. graminearum may be classified into

different groups called chemotypes (Ichinoe et al. 1983). The most commonly produced

trichothecene is deoxynivalenol (DON) and its acetylated derivatives 3-

acetyldeoxynivalenol (3-ADON) and 15-acetyldeoxynivalenol (15-ADON) (Goswami et al. 2004). Nivalenol (NIV) and 4-acetylnivalenol (4-ANIV) are also produced by F. graminearum. These latter trichothecenes have been shown to be more toxic than DON, however, they are encountered less frequently (Desjardins et al. 1993, Rotter et al. 1996).

There are regional differences in the frequency and distribution of F. graminearum chemotypes. The 15-ADON chemotype is more predominant in North America than the

3-ADON chemotype (Gale et al. 2007), however, there is evidence that populations are shifting, and 3-ADON is becoming more prevalent (Ward et al. 2008, Puri et al. 2010). In

Japan and France, both NIV and DON chemotypes occur frequently (Ichinoe et al. 1983).

DISEASE CYCLE AND EPIDEMIOLOGY

F. graminearum overwinters on crop residue, which serves as the main source of inoculum for disease development. The pathogen overwinters either in its sexual state as perithicia or asexually as mycelia or chlamydospores (Sutton 1982). In the spring, inoculum (ascospores, hyphal fragments, or conidia [Parry et al. 1995]) produced in crop residue is either wind (Parry et al. 1995, Gilbert and Fernando 2004) or rain splash (Paul et al. 2004) disseminated to wheat spikes where infection occurs. Inoculum production and dispersal is heavily dependent on temperature and moisture. For instance, perithicia

forcibly discharge ascospores, and wind and rain help to disperse them to the wheat spike

(Trail and Common 2000). Ascospores may travel long distances from surrounding

fields; a dispersal mechanism that is heavily dependent on wind (Parry et al. 1995).

Conidia and mycelial fragments may be disseminated from within the wheat canopy by

rain splash (Paul et al. 2004). Spores landing on the infection court germinate, become

established, and developing on and within spike tissues, both above and below the point

of penetration (Yoshida et al. 2007). As infection progresses, F. graminearum produces

DON ahead of the infection front (Peraldi et al. 2011).

Wheat is most susceptible to infection from the pollination stage (Feekes 10.5.1)

up until the hard dough stage (Feekes 11.2) of development (Sutton 1982, Yoshida et al.

2007). This is because the pathogen is able to exploit and penetrate susceptible flowers,

anthers, embryos, and natural openings such as stomates (Pritsch et al. 2000, Boenisch

and Schäfer 2011). Other spike tissues such as the lemma and palea are less susceptible to

penetration and infection due to their waxy surface, which help to protect the developing

flowers (Walter et al. 2010). Consequently successful penetration and infection relies

heavily on the time of flower opening or anthesis, referred to as Feekes growth stage

10.5.1 (Bai and Shaner 1994). Spikes become less susceptible during the hard dough stage of development.

FHB development is favored by wet, rainy conditions during anthesis and early

grain fill (De Wolf et al. 2003). Optimum conditions for pathogen development include

high relative humidity and temperatures between 25 and 30oF. However infection has

been shown to occur at cooler temperatures when high humidity persists for 3+ days

(McMullen 2008). As the fungus becomes established, infected spikes are unable to

obtain the nutrients required to develop healthy kernels. The result is usually manifested

as bleached, scabby, and light weight spikes (Bai and Shaner 2004, McMullen et al.

1997). Although FHB is generally regarded as a monocyclic disease (Fernando et al.

1997), early infection and uneven flowering may lead to secondary infection and limited

polycyclic spread of the disease. After harvest, perithicia again develop in crop residue,

serving as inoculum for the following year’s disease cycle.

TRICHOTHECENE BIOSYNTHESIS AND ITS ROLE IN FHB DEVELOPMENT

Trichothecenes are a diverse family of sesquiterpenoid mycotoxins produced by

Fusarium and other fungal species (Desjardins et al. 1993). They are broadly classified

into two primary groups (Type A and Type B) based on their chemical structure. Type A

trichothecenes have hydrogen, hydroxyl or esters at C-8, while type B trichothecenes

have a keto group at C-8 (Desjardins et al. 1993). DON and NIV, as well as their

acetylated derivatives, are classified as type B trichothecenes (D’Mello et al. 1998). The

first step in the biosynthetic pathway of all trichothecenes is the conversion of the

molecule farnesyl pyrophosphate into trichodiene via the enzyme sesquiterpene cyclase,

which is mediated by the Tri5 gene (Desjardins et al. 1993, Edwards et al. 2001). This

was the first gene identified to be associated with trichothecene synthesis, and since then,

11 other genes associated with subsequent steps in the biosynthetic pathway have been identified (Desjardins et al. 1993). In F. graminearum, these 12 biosynthetic genes are

found in a core cluster on chromosome 2, with three additional genes (Tri1, Tri 101, and

Tri15) located on chromosomes 1, 4, and 3, respectively (Hallen-Adams et al. 2011).

In FHB development, DON is produced by F. graminearum to aid in the

colonization of wheat tissue. However, DON is not required for infection (Hallen-Adams

et al. 2011). Tri5-deficient F. graminearum mutants (unable to produce toxins) have been

very useful for understanding toxin function and biosynthesis. Non-DON-producing

mutants are still able to cause disease (Desjardins et al. 1993), albeit to a lesser extent

than toxin-producing wild-type strains. This suggests that DON act as a virulence factor

for FHB development. In one study in which green fluorescence protein (GFP) was used

to track a toxin deficient mutant as it colonized the host tissue, a reduced ability on the

part of the fungus to spread from the infection point was evident (Boenisch and Shäfer

2011), further supporting the role of toxins as a virulence factor. In another study

investigating gene expression during wheat kernel colonization, Hallen-Adams (2011) observed that the Tri5 gene was turned on during initial infection. In addition to providing evidence that toxins aid in spread and colonization, results from this study also

showed that Tri5 may be re-activated later in the growing season, contributing to elevated

levels of toxins in harvested grain, even in the absence of visual symptoms. This suggests

that F. graminearum can rapidly resume toxin biosynthesis in dried infected grain under

favorable conditions.

MANAGEMENT OF FUSARIUM HEAD BLIGHT

Cultural control: FHB is best managed by employing good cultural practices

along with planting resistant cultivars and using an effective fungicide. Good cultural

practices include crop rotation with non-host crops, tillage, and weed control. F.

graminearum is a necrotrophic, facultative saprophyte, and as such can survive in crop

residue on or above the soil surface (Fernandez et al. 1993). However, tillage to reduce

inoculum in the field is often at odds with other environmental concerns that tilling

erodes the soil faster, reduces its water holding capacity, and increases CO2 emissions

(Dill-Mackey and Jones 2000). Consequently, farm legislation in the United States was

effected in 1990 which required farmers to adopt conservation tillage plans which

required 30% residue coverage at the time of crop emergence for land classified as

subject to soil erosion (Anonymous 1991). This legislation has likely contributed to the

increase in FHB and DON in the U.S. and has generated new interest in developing

alternative strategies for managing FHB (Dill-Mackey et al. 2000). Similar to what was

observed in the U.S., Koch et al. (2005) noted that the wheat crop in Europe has

increasingly been affected by FHB and DON, and suggested that the adoption of

conservation tillage practices was the primary factor responsible for higher FHB and

DON levels.

No-till or reduced-till farming must therefore rely on other management strategies

to reduce inoculum in and around the field, such as chopping or grinding up residual crop material in order to speedup decomposition (Summerell et al. 1990, Bateman et al. 2007).

Another strategy that can be used to help prevent inoculum buildup is rotation with non- host crops (Dill-Mackey et al. 2000). Staggering planting dates as well as planting cultivars of different maturities could also help to slow down or limit disease development. With these latter two cultural practices, a subset of the plants in the field not at the critical stage for disease development (flowering or anthesis) when environmental conditions become conducive for FHB may simply escape infection (Fry

1982).

FHB Forecasting: Although good cultural practices are recommended to reduce inoculum in the field and reduce FHB severity, Lori et al. (2009) suggested that favorable weather for FHB is likely to be more important than tillage practices, fertilizer treatment, and crop rotation. They suggested that an integrated management approach using as accurate FHB prediction model would be the best approach for managing FHB.

According to De Wolf et al (2003), FHB is well suited for risk assessment modeling because of the narrow infection window, the sporadic nature of the disease, and the fact that epidemics are often compounded by high mycotoxin levels. There is often a significant, positive association between FHB intensity and DON (Paul et al. 2005); therefore, FHB models may be useful tools for wheat growers to decide if a fungicide application is warranted for both FHB and DON control. Predictive models have now become an important component of FHB and DON management programs in the U.S.

Models often consider a range of crop-related variables, including crop growth stage, cultivar resistance, previous crop, as well as environmental variables, such as temperature, rainfall, and moisture (Prandini et al. 2008, Shah et al. 2012). Predictive models do have limitations, however, particularly in determining mycotoxin contamination, since mycotoxins may still be produced post-anthesis and in harvested grain (Pradini et al. 2008).

Cultivar resistance: Resistance to FHB is described as being partial and non- specific, meaning that no cultivar is completely resistant, nor do they show different resistant responses to different F. graminearum populations (Snijders et al. 1992).

Resistance in wheat to FHB was first described in 1963 as being either type I, which is resistance to initial infection, or type II, resistance to spread within the spike. Type II

9 resistance was thought to be more common than type I (Schroeder and Christensen 1963).

Type I resistance is usually evaluated by visually assessing disease following spray inoculation, which closely mimics natural infection. Type II resistance can be evaluated by performing point inoculations, where wheat spikes are directly penetrated and infected with the pathogen and disease spread within the spike can be visually evaluated. Work completed in the 1980s by Miller et al. (1985) led to the description of type III resistance, defined as resistance to DON accumulation. Type III resistance was quantified by measuring ergosterol, a unique component of fungal cell walls, and DON, and comparing the ergosterol:DON ratio among cultivars with known levels of type II resistance.

Susceptible cultivars (based on type II resistance) with high ergosterol:DON ratios were thought to have type III resistance. Miller et al. (1985) hypothesized that type III resistance could be due to some cultivars’ ability to degrade toxins after they are produced in host cells. However, this type of resistance was not thought to be entirely independent of type II resistance. D’Mello et al. (1998) suggested that what is referred to as type III resistance could in fact simply be type II resistance. In the late 1990’s, additional types of resistance were described that were thought to be independent of resistance types I, II, and III (Mesterhazy 1995, Bai and Shaner 2004). Type IV resistance was described as resistance to kernel infection and type V resistance as tolerance. Type

IV resistance can be visually assessed by quantifying % Fusarium diseased kernels

(FDK) or fungal biomass (Rudd et al. 2001, Bai and Shaner 2004).

Fungicides: A well-timed fungicide application is also useful for managing FHB and DON. The recommended application time is at anthesis (Haidukowski et al. 2005,

Beyer et al. 2006) when plants are most susceptible to infection. Usually only a single

fungicide application is justified, as the cost of additional treatments is often not offset by

increased yield (McMullen et al. 1997). Additionally, applying a fungicide in years not conducive to disease development may not be necessary (Weisz et al. 2010). Application rates and concentrations should also be evaluated to ensure adequate coverage of vulnerable wheat spikes, as partial or incomplete coverage may significantly reduce disease control (Hooker et al. 2004). Applying sub-lethal doses of fungicides has been shown to lead to resistance, which may be accompanied by increased mycotoxin levels

(D’Mello et al. 1998). However, fungicides alone are not a reliable FHB management strategy. Efficacy varies considerably from one study to another. While many studies have reported effective FHB control (Paul et al. 2008, Haidukowski et al. 2005, Beyer et al. 2006), other studies have reported the opposite (Milus et al. 1994), or asymptomatic grain with high levels of toxin in plots treated with some fungicides (Hollingsworth et al.

2008). The literature has emphasized an integrated management approach using a fungicide in combination with a resistant cultivar (McMullen et al. 1997, D’Mello et al.

1998, Willyerd et al. 2012, Wegulo et al. 2011).

Several factors need to be taken into consideration when selecting a fungicide for

FHB management. Fungicides should not only be able to reduce FHB development, but they should also have the ability to control DON contamination (Paul et al. 2008). To date, research has shown the most effective fungicides available for controlling FHB and

DON are collectively within the chemical class known as the Demethylation Inhibitors

(DMI’s), or triazoles (Zhang et al. 2009, Paul et al. 2008). Triazoles act to inhibit ergosterol biosynthesis in fungi, specifically cytochrome P450 sterol 14α-demethylase, which is an essential enzyme. Ergosterol is a very important structural component in the

11 cell wall of ascomycete and basidiomycete fungi, and its disturbance compromises fungal membrane integrity (Becher et al. 2009).

Quinine Outside Inhibitors, commonly referred to as strobilurins, comprise another class of chemicals that is of interest for controlling wheat diseases. Strobilurins were discovered to be produced by several of basidiomycete wood rotting fungi.

Synthetic alteration of these naturally occurring chemicals has allowed high selectivity towards target oomycete, ascomycete, and basidiomycete fungi (Ypema and Gold, 1999).

This class of chemicals has the ability to inhibit mitochondrial respiration by binding to cytochrome b on the inner mitochondrial membrane of fungi and other eukaryotes

(Bartlett et al. 2002). By binding to cytochrome b, they disrupt the energy cycle, and subsequent production of ATP. As a consequence, these chemicals tend to be most effective at controlling fungi during high energy states of development such as spore germination and early germ tube growth. This is an interesting contrast to triazole fungicides, which work to inhibit ergosterol biosynthesis. Spore germination and germ tube development uses ergosterol reserves from the spore, so DMIs are generally not effective at controlling diseases during the early stage of development (Bartlett et al.

2002).

Contrary to DMI’s, QoI’s are generally not recommended for FHB control due to reports of increased DON contamination of grain in response to a QoI application

(Magan et al. 2002). Two studies attributed such an increase to the selective elimination of non-toxin producing Michrodochium species, which allows toxin-producing Fusarium species to colonize (Simpson et al. 2001, Zhang et al. 2009). Beyer et al. (2006) suggested that applications made several weeks before anthesis could be more harmful to

12 non-toxigenic microorganisms, which would facilitate in the proliferation of toxigenic

Fusarium species. However, Ruske et al. (2003) and Blandino et al. (2009) refuted the idea of fungicides selectively targeting fungal species as an explanation for higher DON in plots treated with QoIs, as they found no correlation between fungal biomass and DON in contaminated grain, but a disproportionate amount of DON compared to visual disease.

They hypothesized that increased toxin levels could be due to the fact that strobilurins have been shown to delay senescence in plants, and this could give the fungus additional time to colonize plant tissue and produce DON at the infection front.

Fungi tend to develop resistance against fungicides with high specificity much faster, and as a result, they may become less effective faster than fungicide with low specificity. Resistance has been reported in both DMI (Becher et al. 2009) and QoI fungicides (Parnell et al. 2006, Yun-Sik-Kim et al. 2003), although resistance to DMI tends to develop more gradually than resistance to QoI (Yun-Sik-Kim et al. 2003).

Parnell et al. (2006) noted that after the strobilurins were first introduced in 1996 to control a range of cereal pathogens, it took only four years for high frequencies of resistance to be detected over large areas of Germany, France, and the United Kingdom.

It is not only important to understand how fungicides affect disease development, it is equally important to investigate fungicides with different modes of action in order to use then in a responsible chemical management program through tank mixes (Hobbelen et al.

2011) and rotation of modes of action in order to reduce the risk of resistance.

Chapter 2

Efficacy of Post-Anthesis Fungicide Application against Fusarium Head Blight and Deoxynivalenol in Soft Red Winter Wheat

2.1 INTRODUCTION

Fusarium head blight (FHB), commonly known as scab, is a fungal disease caused by Fusarium graminearum and other relates species (Parry 1995) that affects small grain crops such as wheat, barley, and rye in many regions around the world. FHB causes significant losses in grain yield and quality due to floret sterility, production of small, shriveled light-weight kernels, and contamination of grain with mycotoxins, especially deoxynivalenol (DON). Cultivation of FHB-affected seeds may lead to poor stand establishment as a result of reduced seed vigor and germination, and seedling blight

(Gilbert and Tekazu 2000). In addition FHB is a concern for the milling and baking industries as it affects the quality of the flour produced from infected grain (Dexter et al

1996). It may also have negative effects on animal health, especially non-ruminant animals, if mycotoxin-contaminated grain is consumed (Diekman et al 1992).

Efforts to manage FHB and DON are most effective when cultural practices such as crop rotation with non-host crops and tillage are employed along with genetic resistance and a well-timed fungicide application (Willyerd et al 2012, Wegulo et al.

2011). Lori et al. (2009) suggested that although cultural practices alone may be effective at reducing in-field inoculum and FHB severity, under favorable weather conditions, an integrated management approach that includes an accurate prediction model is necessary to effectively manage FHB. The use of such a model is now an integral part of FHB management programs in the U.S., providing estimates of disease risk in 30 states to help guide fungicide application (De Wolf et al 2003, Shah et al 2013). Usually only a single anthesis application is justified, as the cost of additional treatments is often not offset by increased yield in U.S. wheat production systems (McMullen et al. 1997). Depending on the level of disease and susceptibility of the cultivar being treated, a well-timed application of 41% prothioconazole (Proline 480 SC; Bayer CropScience, Research

Triangle Park, NC), 19% tebuconazole + 19% prothioconazole (Prosaro 421 SC; Bayer

CropScience, Research Triangle Park, NC) or 8.6% metconazole (Caramba 90 SL; BASF

Corporation Agricultural Products, Research Triangle Park, NC) may provide between 40 and 70% reduction of FHB and DON (Paul et al 2007, Paul et al 2008, Willyerd et al

2012).

High relative humidity, rainfall, and surface wetness during anthesis and early grain fill, the period of greatest host susceptibility to infection, are environmental risk factors for FHB development and DON contamination. These are the conditions required for spore production (Dufault et al 2006), dissemination (Paul et al 2004, Sutton et al

1982), and infection of the wheat spike (Bai and Shaner 1994). Consequently, fungicide application for FHB management is especially emphasized when anthesis coincides with wet weather (Haidukowski et al. 2005, Beyer et al. 2006). However, under wet field conditions, when fungicides are most warranted, current recommendations may be

difficult to adhere to due to physical limitations of spraying a fungicide in the rain and driving equipment through soggy fields. These limitations have led to questions being asked about the benefit of pre- and post-anthesis fungicide application for FHB and DON management. Edwards and Godley (2010) evaluated the effects of pre-anthesis application of 41% prothioconazole (Proline 480 SC; Bayer CropScience, Research

Triangle Park, NC) on FHB and DON in winter wheat in the UK, and reported that treatments made at Zadoks GS31 (stem elongation), GS39 (full flag leaf emergence), and

GS65 (mid-anthesis) reduced FHB incidence and DON by 50, 58, and 83%, and 27, 49, and 57%, respectively, relative to the untreated check. Applications made at all three growth stages provided 97% control of FHB incidence and 83% control DON. Based on results from a similar study conducted in Canada in which 38.7% tebuconazole (Folicur

3.6F; Bayer CropScience, Research Triangle Park, NC) was applied to hard red spring wheat at GS39 and GS60 (early anthesis), Wiersma and Motteberg (2005) also reported that, across several cultivars, the best timing for FHB control (as well as foliar diseases) was at early anthesis, but concluded that FHB levels were too low to adequately compare treatments. Evaluating the effects of post-anthesis fungicide applications on FHB, DON

and nivalenol (NIV) in artificially inoculated field experiments, Yoshida et al (2012)

observed that an anthesis application of 70% thiophanate-methyl (Topsin M; Nippon

Soda Co. Ltd., Chiyodaku, Tokyo) resulted in significantly lower levels of FHB than the

untreated check and applications made at 10, 20 and 30 days after anthesis (DAA).

However, in one of the two years of the study, they observed that a single application

made at 20 DAA or an application made at anthesis followed by a second at 20 DAA

significantly reduced mycotoxin contamination (DON + NIV) relative to the check and

the anthesis treatment, without reducing FHB severity. This led them to speculate that late-milk application (GS 11.1) was probably the best timing for mycotoxin control in wheat, since toxins may continue to accumulate during grain development.

It is clear from the studies cited above that pre- and post-anthesis treatments may

contribute to FHB and DON reduction, but the anthesis treatments still stood out as being the most effective for FHB in all cases. However, 19% tebuconazole + 19% prothioconazole and 8.6% metconazole, the two most effective fungicides again FHB

(Paul et al 2008), were not evaluated in any of the aforementioned studies. In addition,

none of the studies were conducted in soft red winter wheat production regions of the US.

Given that fungicide effects on FHB and DON have been shown to vary considerable

among active ingredients, locations, and wheat market classes (Paul et al 2008), further

research is needed to determine whether applications made after anthesis will provide

adequate control of this disease and its associated toxins in soft red winter wheat under

field conditions in the US Midwest. If foliar fungicides are still able to adequately control

FHB and DON when applied after anthesis, recommended application times could

potentially be extended, allowing producers the flexibility of applying treatments after

rain events. The objectives of this research were to i) determine the effect of post-anthesis

applications of 19% tebuconazole + 19% prothioconazole and 8.6% metconazole on FHB

and DON in soft red winter wheat (SRWW) under different naturally infected and

artificially inoculated field conditions, ii) determine whether the magnitude of FHB and

DON responses to post-anthesis fungicide applications varied with active ingredient,

cultivar, and baseline disease and toxin levels, and iii) estimate the efficacy (based on

17 mean percent control of FHB index and DON) of post-anthesis treatments relative to untreated and anthesis reference treatments.

2.2 METHODOLOGY

Plot establishment and experimental design Wooster, Ohio. Experiments were

conducted during the 2011 (WO11), 2012 (WO12), and 2013 (WO13) wheat growing

seasons on the Snyder Farm at the Ohio Agricultural Research and Development Center

(OARDC), near Wooster, OH. On 8 October, 2010 and 7 October, 2011, moderately

resistant and susceptible soft red winter wheat (SRWW) cultivars, Malabar and

Hopewell, respectively, were planted in 1.5 x 6-m plots, with 1.5-m border rows of the

cultivar Truman planted between adjacent plots to minimize inter-plot interference. Seeds

were planted using a Kincaid planter at a seeding rate of 4 x 106 seeds/ha.

In 2011, the experimental design was a randomized complete block, with three

replicate blocks and a split-split-plot arrangement of cultivar as whole plot, fungicide application timing as sub-plot, and inoculum density as the sub-sub-plot. In 2012, a similar experimental design and the same three treatment factors were used; however, they were arranged differently. Inoculum density was the whole-plot, fungicide timing the sub-plot, and cultivar the sub-sub-plot.

For the experiment conducted in 2013, plots were established on 25 September,

2012. However, only the susceptible cultivar Hopewell was planted as described above.

The experimental design was also a split-split-plot, with three replicate blocks. Inoculum density was the whole plot, fungicide active ingredient the sub-plot, and application timing the sub-sub-plot.

South Charleston, Ohio. During the 2011 (SC11) and 2013 (SC13) growing

seasons, similar experiments to those conducted in Wooster were conducted at the

OARDC Western Agricultural Research Station, near South Charleston, OH, 220

kilometers southwest of Wooster. On 8 October, 2010 and 13 October, 2012, the susceptible SRWW cultivar Bravo was planted in 3 x 6-m plots at a seeding rate of 3 x

106 seeds/ha, with 3-m border strips between adjacent plots. In 2011, the experimental design was a randomized complete block, with four replicate blocks of seven treatments.

Treatments consisted of the application of two different fungicides, each at three different growth stages (Table 2.1), plus an untreated check. This study was naturally infected. In

2013, the experimental design was a randomized complete block, with a split-plot arrangement of inoculum density as whole-plot in four replicate blocks and nine fungicide treatments as sub-plot. Treatments again consisted of two fungicides applied at four different growth stages, plus an untreated check (Table 2.1).

Urbana, Illinois. Field experiments similar to those described above were conducted at the Illinois Council for Food and Agricultural Research facility, near

Champaign-Urbana, IL, during the 2012 (IL12) and 2013 (IL13) growing seasons. FHB susceptible SRWW cultivar Pioneer 25R47 was planted in 1.5 x 6 m plots on 10 October,

2011 and 11 October, 2012. Plots were planted with a Great Plains no-till drill

(3P606NT) planter at a seeding rate of approximately 3 x 106 seeds/ha. In both years, the

experimental design as a randomized complete block, with 22 treatments randomly

assigned to plots within each of four blocks. Treatments were in a factorial arrangement

of two fungicides, five application timings, and two inoculum densities, plus an untreated

check at each of the two inoculum densities (Table 2.1).

Fungicide treatment application. The Demethylation Inhibitor (DMI) fungicide

19% tebuconazole + 19% prothioconazole (Prosaro 421 SC, abbreviated as

TEBU+PROT) was used at label-recommended rate (475 ml/ha) in all seven experiments, while the DMI 8.6% metconazole (Caramba 90 SL, abbreviated as METC) was used, also at label-recommended rate (1,023 ml/ha), in 5 (WO13, SC11, SC13, IL12 and IL13) of the seven experiments. In all cases, a nonionic surfactant, Induce (Helena Chemical Co.

Collierville, TN), was added to the fungicide mixture at a rate of 0.125% v/v, and applications were made at 50% anthesis in all experiments, plus 3-4 post-anthesis treatments, applied at either 2, 3, 4, 5 or 6 days after anthesis, depending on the experiment. Hereafter, anthesis treatments will be abbreviated as A, and post-anthesis treatments as A+2, A+3, A+4, A+5, and A+6, respectively. Table 1 provides treatment details for each experiment.

For WO11, WO12, and WO13, treatments were applied using either a tractor mounted or backpack (R&D Sprayers, Opelousas LA) sprayer, with booms fitted with

Twinjet XR8002 nozzles or paired XR8001 nozzles, mounted at an angle (45°) forward and backward, and calibrated to apply at a rate of 187L/ha at 207 kPa. Treatments at

SC11 and SC13 were applied using a CO2 backpack sprayer with a 3-m boom, fitted with

6 Teejet drift guard nozzles DG 100-02, calibrated to apply at a rate of 187L/ha at at 207 kPa. For IL12 and IL13, applications were also made using a backpack sprayer and handheld boom, equipped with 3 TwinJet XR8002 nozzles, at a rate of 187 L/ha at a spray pressure of 276 kPa.

Inoculum preparation and Inoculation. For experiments conducted in Ohio, inoculum was prepared using 10 aggressive isolates of Fusarium graminearum collected

from naturally infected wheat fields in Ohio. Isolates were first grown on nutrient rich media (PDA, Komada), and then transferred to either mung bean agar (MBA) for macroconidia production, or carrot agar (CA) for ascospore production. Macroconidia were produced by incubating cultures for 7-10 days at room temperature under ultraviolet lights, with a 12-hour photoperiod. Ascospore production was stimulated by using a sterilized rubber policeman to remove mycelium from CA plates (after about 4 days) in order to induce self-fertilization. Plates were then incubate at 22-25oC with a 10 hour

photoperiod for another 10-14 days until perithicia were visible on the surface of the

media. Sterile deionized water was then added to each MBA and CA plate, and

macroconidia and ascospores were harvested by gently scraping the surface of the media with a rubber policeman. Spore concentrations were determined using a hemacytometer

(Reichert-Bright Line, Hausser Scientific. Horsham, PA 19044) and stock suspensions were stored at -20oC until used for inoculation.

With the exception of SC11, where the field was naturally infected, all

experiments were artificially spray-inoculated at early anthesis with inoculum

concentrations ranging from 2 x 104 to 14 x 104 spores/ml (Table 2.1). For experiments

conducted in Ohio, equal volumes and concentrations of ascospores and macroconidia

were mixed prior to inoculation. Applications were made using a backpack sprayer (R&D

Sprayers, Opelousas, LA) fitted with 3 Teejet Flat Fan nozzles on a 3-m-long boom. By

adjusting the pressure and application speed, the equipment was calibrated to administer

approximately 67 ml of inoculum per m2 of plot. In Illinois (IL12 and IL13), inoculum

preparation and plot inoculations were similar to that described above for trials conducted

in Ohio. However, only macroconidia were used.

Data collection and Analysis. Disease intensity was visually evaluated during

the soft dough stage (GS 11.2) of crop development, approximately 3 weeks after

anthesis. A total of 100-120 spikes were evaluated per plot. For WO11, WO12, WO13,

IL12 and IL13, five clusters of twenty spikes were arbitrarily selected per plot, for a total of 100 spikes, and evaluated for disease symptoms. For SC11 and SC13, six arbitrarily selected clusters of twenty spikes, for a total of 120 spikes, were scored for FHB symptoms, accounting for the larger plot sizes. In all cases, FHB index (plot severity) was visually estimated as the mean percentage of diseased spikelets per spike, and FHB incidence as the percentage of diseased spikes out of all spikes sampled. In Illinois, diseased-head severity and incidence were evaluated on 10 spikes per plot and then used to estimate index as described previously (Paul et al 2005, Stack and McMullen 1998).

Depending on the location and year, plots were harvested between late-June and mid-July, using research plot combine harvesters. Grain samples were collected from each plot and evaluated for Fusarium damaged kernels (FDK) by visually estimating percent diseased kernels (small, shriveled, and discolored). Additional grain samples were ground using a coffee grinder (Braun Aromatic KSM 2 B, Gillette Commercial,

Boston, MA), or a laboratory mill (Laboratory Mill 3033, Perten Instruments,

Springfield, IL) for DON analysis, and sent to the US Wheat and Barley Scab Initiative mycotoxin testing laboratory in the department of plant pathology, University of

Minnesota, St. Paul, MN, for DON quantification by gas chromatography-mass spectrometry (GC-MS).

All data were analyzed using PROC GLMMIX in SAS (SAS, Cary, NC) (Littell et al. 2006) in order to evaluate the effects of fungicide application timing and, where

23 applicable, cultivar, fungicide active ingredient, inoculum density and their interactions on FHB, FDK, and DON. Prior to the analyses, FHB incidence, index and FDK data were arcsine-square-root-transformed and DON was log-transformed to stabilize variances.

Each experiment was analyzed separately, with application timing, cultivar, fungicide and inoculum density treated as fixed effects, and block and all interactions involving block (in the case of split-plot experiments) treated as random effects in the analyses. For all statistically significant effects, contrast and lsmestimate statements in GLIMMIX were used to compare post-anthesis treatments with the untreated check and the anthesis treatment. Percent control of index and DON relative to the untreated check was estimated as described by Paul et al. (2007, 2008) for each treatment. For experiments with resistant and susceptible cultivars (WO2011 and WO2012), the untreated susceptible check was used as the reference treatment for estimating percent control.

2.3 RESULTS

FHB and toxin levels. Mean disease and toxin levels varied among years and

locations. During the Wooster, OH, 2011 (WO11) wheat field season, averaged across

inoculum density, cultivars, treatments and replicates, mean disease index levels (IND, mean proportion of diseased spikelets per spike) ranged from 1.6% to 10.4% (Fig. 2.1A), and mean DON levels from 0.13 to 1.56 ppm (Fig. 2.2A). In 2012 (WO12), an unusually dry season, both IND and DON levels were much lower, with mean IND ranging from

0.4 to 2.4%, and mean DON levels from 0.01 to 0.57 ppm (Fig. 2.1B and Fig. 2.2B) . In

2013 (WO13), the corresponding mean ranges were 0.6 to 4.2% for IND and 1.11 to 2.21 ppm for DON. It is interesting to note that the upper limit for mean DON for the 2013 experiment was in anthesis treated plots, not the untreated check (Fig. 2.1C and Fig.

2.2B).

For the 2011 experiment conducted at South Charleston, Ohio (SC11), a location and year with natural rainfall during anthesis and natural infection, averaged across fungicide active ingredients and inoculum densities, mean IND levels ranged from 2 to

12% (Fig. 2.1D), and mean DON levels from 2.6 to 7.7 ppm (Fig. 2.2C). In 2013, mean

IND levels in an inoculated trial at that same location (SC13) ranged from 2.3 to 9.4%, and mean DON from 0.91 to 2.5 ppm (Fig. 2.1E And 2.2D). The corresponding mean disease and toxin ranges (averaged across inoculum densities and active ingredients) for experiments conducted in Illinois in 2012 (IL12) and 2013 (IL13) were 0.3 to 30.8% and

0.13 to 0.38 ppm for mean IND and DON in 2012, and 17.3 to 37.8% and 5.8 to 11.8 ppm for mean IND and DON in 2013.

In all seven experiments, with the exception of the Wooster 2013, very similar trends were observed, with the upper disease and toxin limits being derived from mean levels in the untreated checks (Fig. 2.1 and 2.2). In general, all plots treated with PROT +

TEBU, both at and after anthesis, had considerably lower mean IND and DON than the untreated check (Fig. 2.1 and 2.2). For experiments with different cultivars (WO11 and

WO12) and different fungicides (WO13, SC11, SC13, IL12 and IL13) (Prosaro and

Caramba; PROT+TEBU and METC, respectively) similar trends were observed, with treated plots having lower mean levels of IND and DON than the check. However, the susceptible cultivar (Fig 2.1A and 2.1B, and 2.2A) and plots treated with PROT+TEBU

(Fig. 2.1C, 2.1D, 2.1E, 2.1F, 2.1G, 2.2B, 2.2C, 2.2D, and 2.2E) tended to have higher mean disease and toxin levels than the moderately resistant cultivar and plots treated with

METC, respectively.

Fungicide, cultivar, and inoculation effects on disease and toxin levels. Results from linear mixed model analyses of the effects of one or more of the following fixed- effect treatment factors or combination of factors - cultivar (CV), fungicide chemistry

(FUN), application timing (TRT) and inoculum density (INOC) - on arcsine-square-root- transformed FHB index (arcIND), incidence (arcINC) and FDK (arcFDK), and log- transformed DON (logDON) are summarized in Table 2.2. For trials with three fixed- effect factors (WO11, WO12, and WO13), the three-way interactions were not statistically significant for any of the measured responses (Table 2.2). Similarly, for trials with at least two of the four tested treatment factors (WO11, WO12, WO13, and SC13),

in general, none of the two-way interactions were statistically significant for any of the measured responses. For SC11 in which treatments consisted of a factorial arrangement

of FUN and TRT, and IL12 and IL13 in which treatments consisted of a factorial

arrangement of FUN, TRT, and INOC, results from linear contrasts of least squares

means showed what the effect of TRT was not influenced by FUN, INOC or FUN*INOC

(data not shown).

For all experiments, pairwise contrasts of LS means for each of the reference

treatments (untreated check and anthesis treatment) with each of the post-anthesis

fungicide treatments were performed based on the results of the linear mixed model

analyses. Contrast results were used to determine fungicide effects on arcIND, arcINC,

arcFDK, and logDON. Since the results for arcINC and arcFDK were very similar to

those observed for arcIND, only the latter response and logDON are shown in Table 2.3

and discussed hereafter. In general, similar trends were observed across all experiments,

for all years and locations. In all cases, mean arcIND was significantly lower in the

anthesis and post-anthesis fungicide treatments compared to the untreated check (Table

2.3, P < 0.01). With two exceptions, there was no significant difference between the

means for the anthesis treatments and means for any of the post anthesis treatments. The

only exceptions were for SC11, the naturally-infected experiment, in which it rained

during anthesis (33 mm on the day the fungicide was applied and 39 mmm on the day

after), and WO13, the study in which it was very cold during and shortly after anthesis

(mean, max, and min temperatures 8.5, 22.8, and -1.6oC, respectively, for 48 hours,

beginning at the time of treatment application). In the former experiment, averaged across

PROT+TEBU and METC, applications made at 2 days after anthesis had significantly

lower mean IND than the anthesis application. A similar, but marginal effect (P < 0.10) was observed for application made at 5 days after anthesis. For WO13, averaged across

PROT+TEBU and METC, all post-anthesis treatments (A+2, A+4, and A+6) had significantly lower mean arcIND that the anthesis treatment.

Very similar trends were observed for DON, with all fungicide treatments generally having significantly lower logDON than the untreated check (P <0.001), and generally with no significant difference between the anthesis and post-anthesis fungicide treatments (for trials with > 1 ppm DON in the untreated check) (Table 3; WO11, IL13, and SC13). SC11 and WO13 were again the exceptions, showing responses very similar to those described above for logIND. However, with toxin, the A+5 treatments for the

SC11 trial had significantly higher logDON than the anthesis treatment (Table 2.3). For

WO13, there were elevated DON levels in the anthesis treatment compared to the untreated check as well as all of the post-anthesis treatments, with the difference being statistically significant for all post-anthesis treatments (P < 0.001; Table 2.3).

Percent Control of FHB Index. Percent control, defined as C = ([X̅ check - X̅ Treated]

/X̅ check)•100, where X̅  represents mean for a treatment or control, is a useful matrix for

growers to determine the efficacy of a particular treatment in terms of disease control.

Data from the seven experiments were therefore used to estimate mean percent control

(C̅ ) of index and DON for all treatments relative to the untreated check and for post- anthesis treatments relative to the anthesis treatment, based on multivariate random-

effects meta-analyses (Paul et al 2008, Willyerd et al 2012). Since two- and three-way

interactions were not statistically significant (Table 2.2), whole-plot factors (where

applicable) were treated as separate experiments, and mean IND and DON, averaged

across inoculum densities (where applicable), as well as their standard error were used for

the meta-analyses. Sampling variances of the means were estimated from linear mixed model analyses (Littell et al. 2006) of the individual trial data and the meta-analytical

models were fitted with log response ratio (L) as the effect size, following the approach

described by Willyerd et al (2012). Mean percent control for IND and DON and their

95% confidence intervals were estimated from mean L (L̅ ) values and their confidence

intervals as C̅ = [1-exp(L̅ )]100.

Results from the meta-analyses are summarized in Table 2.4. Based on the standard normal test (Z statistic in the table), log response ratios between the check (CK) and the fungicide treatments, regardless of timing, were statistically different from zero

for both FHB index (IND) and DON (P < 0.001). The treatment applied two days after anthesis (A+2) had the highest negative L̅ values, -1.18 for IND and -0.77 for DON, and correspondingly, the highest C̅ values relative to the check, 69 and 54%, for IND and

DON, respectively. C̅ for the other treatments were 56 and 50% for the anthesis

treatment, 62 and 52% for the treatment applied four days after anthesis (A+4), and 62

and 48% for the treatment applied six days after anthesis (A+6) (Table 2.4). For IND, L̅

values between A and post-anthesis treatments were only significantly different from

zero for A+2 and A+6 (P < 0.05), and marginally significant for A+6 (P = 0.09). Relative to the anthesis treatment, A+2, A+4 and A+6 further reduced FHB index by 30, 13 and

14%, respectively. Post-anthesis treatments were not significantly more effective than the anthesis treatment against DON (L̅ not significantly different from zero), although the

A+2 and A+4 treatments reduced DON a further 8 and 6%, respectively, relative to A.

2.4 DISCUSSION

Our findings summarize multiple years of field experiments conducted under a range of soft red winter wheat (SRWW) growing condition in Ohio and Illinois to assess the benefits of applying a fungicide up to 6 days post anthesis (dpa) for Fusarium head blight (FHB) and deoxynivalenol (DON) management. This was the first comprehensive

US-based study to formally assess the efficacy of 19% tebuconazole + 19% prothioconazole (TEBU+PROT; Prosaro) and 8.6% metconazole (METC; Caramba), the two most effective fungicides again FHB and DON (Paul et al. 2008), when applied within a narrow window (up to one week) post anthesis. We consistently found significantly lower levels of IND, DON, and FDK in anthesis as well as all post anthesis fungicide treatments compared to the untreated check in all seven field trials. In general, there was no degradation of fungicide efficacy by extending the application up to 6 day after 50% anthesis. In some instances, post-anthesis fungicide treatments provided better control of IND or DON than the treatment made at anthesis. The effects of cultivar

(moderately resistant and susceptible) varied among experiments, but did not affect post- anthesis treatment effects on IND or DON based on the non-significant interactions.

Across years and locations, there was no significant interaction of fungicide chemistry and application timing, or of inoculation densities and application timings, further emphasizing the consistency of the post-anthesis treatment effect. Overall, there was an average of 56 to 69% control of FHB index and a 48 to 54% control of DON when either

19% tebuconazole + 19% prothioconazole or 8.6% metconazole was applied at or up to 6

days after anthesis.

Several factors may have been responsible for the observed success of post-

anthesis fungicide applications on FHB and DON, including successful control of FHB

after infection has occurred, successful control of post-anthesis infections, heterogeneous crop development in wheat plots, and the interaction of these factors with weather conditions. Several studies have demonstrated that “late” infections of wheat spikes do occur (Lacy et al. 1999, Hart et al. 1984, Del Ponte et al. 2007, and Yoshida et al. 2007), which could affect both FHB symptoms and DON production. However, several of these studies also showed that the highest level of FHB was observed when infections occurred at anthesis compared to after anthesis. So, anthesis would seem to be the optimum time for applying a preventative fungicide treatment to control FHB as well as for spray inoculating plots as was done in this investigation to evaluate the efficacy of such a treatment. However, accurately determining when anthesis actually occurs in a population of wheat spikes in the field is a challenge. The 50% anthesis fungicide

recommendation often used for FHB control and for spray inoculation commonly refers to the growth stage of primary tillers. Therefore, anthesis for primary tillers may in fact be pre-anthesis for secondary tillers. Similarly, post-anthesis for primary tillers may actually be anthesis for secondary tillers. Synchrony of crop development and tillering are influenced by several factors, such as cultivar characteristics, weather conditions, and crop husbandry (row spacing, planting density and fertility program), which as a consequence may affect fungicide performance and the success of artificial spray inoculations.

All plots in 6 of the 7 experiments in this investigation were spray inoculated with

fairly high spore densities at mid-anthesis, approximately 12-36 h after the anthesis

treatments were applied, but one to five days before the post-anthesis treatments.

Therefore, it seems reasonable to assume that the resulting disease and toxin levels were

largely due to the artificial inoculations rather than natural inoculum and to infections at

anthesis (at least for primary tillers) rather than late infections. Furthermore, given the

inoculation and treatment protocols used in these studies, one may further assume that

post-anthesis-treated plots, not subjected to the preventative or protective action of the

fungicides, were equally likely to be infected during the critical anthesis growth stage as

the untreated check. The observed efficacy of post-anthesis treatment against index and

DON relative to the untreated check could therefore be attributed, at least in part, to

reduction of fungal spread within the spike and visual symptom development after infection had occurred. The application of a fungicide after infection, but before symptom development, is often referred to a curative treatment effect (Ivic 2010, Mueller and

Bradley 2008), where the fungicide is absorbed and affect mycelial growth within the

plant tissue. Curative and protective activities are the physical modes of action of

Demethylation Inhibitor fungicide such as 19% tebuconazole + 19% prothioconazole

(TEBU+PROT) and 8.6% metconazole (METC). Although there is no documented report

of the curative activity of TEBU+PROT and METC against Fusarium head blight of

wheat, this type of effect has been demonstrated for other pathosystems with sister active

ingredients such tebuconazole, epoxiconazole, difenoconazole, and propiconazole

(Dahmen and Staub, 1992, Godoy and Canteri, 2004).

The fact that post-anthesis treatments were just as effective in the naturally

infected as well as the artificially inoculated experiments suggest that possible curative activity was probably not the only explanation for the observed responses. Coupled with the postulated curative effect on infected spikes, applications made “after” anthesis (for primary tillers) likely provided protection to secondary tillers which may have actually been at anthesis at the time of application, and also protected primary tillers from late infections. The magnitude of the “late” application effect and occasional greater control with post-anthesis treatments (relative to anthesis treatment) in some studies could be due to differences in the relative proportion of primary and secondary tillers in a plot or field.

If, for instance, this balance shifts in favor of secondary tillers (due to the weather-, cultivar-, and cropping factors mentioned above), then “post-anthesis” applications (using the primary tillers as a reference) may perform as well or even better than an anthesis application. If the proportion of secondary tillers is relatively high, then anthesis applications based on primary tillers will actually be treating secondary spikes at pre- anthesis, which some studies have shown to be relatively less effective than anthesis or post-anthesis treatments (Yoshida et al. 2012, Weismera et at. 2005).

Differences in tiller development or the length of the anthesis window as a whole were probably the reasons for relatively superior performance of all of the so-called post- anthesis treatments relative to the anthesis treatments in the WO13 experiment. In that particular experiment, conditions were unseasonably cold during and after anthesis of the primary tillers, which likely slowed down crop development and extended the anthesis window, allowing the post-anthesis treatments to protect more spikes at the critical anthesis growth stage, while anthesis treatments protected spikes at pre-anthesis when

fungicides are usually less effective against FHB. Rainfall at anthesis also could affect

results. For the other experiment in which post-anthesis treatment out-performed anthesis

treatment (SC11), it rained before (53 mm on the two days before), during (33 mm), and

after anthesis (39 mm). Rain at the time of anthesis likely reduced fungicide efficacy,

which then allowed applications made at 2 (0.51 mm of rain) and 5 days (0 mm of rain)

after anthesis to provide better FHB and DON control than the anthesis treatments.

Results from a recent study conducted by Andersen et al. (2013) showed that rain

immediately after TEBU+PROT application may reduce the efficacy of this fungicide

against FHB. Although SC11 was the only location-year that actually had natural rainfall

during anthesis, it still serves to highlight the potential negative effect of rainfall on

fungicide efficacy if an application is made during the rain and the value of post-anthesis

application when rainfall prevents effective fungicide application at anthesis. The

inability to treat fields under wet soggy conditions is not the only justification for

considering a post-anthesis application, reduced fungicide efficacy is another.

Although there have been multiple previous studies evaluating the effects of pre- and post-anthesis fungicide applications on FHB and DON (Edwards and Godley 2010,

Wiersma and Motteberg 2005, Yoshida et al. 2010), none of those studies evaluated the effects of a single fungicide application within the one-week post-anthesis window evaluated in this study. For instance, while Yoshida et al. (2010) determined that a fungicide application at 20 days after antehsis effectively controlled mycotoxin levels in

Japan, this recommendation is not as useful in SRWW production areas in the Midwest

due to pre-harvest interval (PHI) restrictions. The time interval between applying a

fungicide at 20 days after anthesis and harvest (approximately 21 days in SRWW

growing regions) would not meet the legally required PHI for TEBU+PROT or METC

(30 days). In addition, TEBU+PROT or METC were not evaluated in any of the previous

studies, nor were the studies conducted in the U.S. Midwest. Since fungicide effects on

FHB and DON have been shown to be affected by fungicide active ingredients, weather conditions, growing regions, and wheat market classes (Paul et al. 2008), it is important

to take these factors into consideration when evaluating post-anthesis fungicide

applications.

All wheat growing regions and their associated wheat class each have their own

unique set of growing conditions and environmental variables that need to be addressed

when considering management decisions. In the US Midwest, fungicide spray

recommendations are based on applications being made at anthesis; however, ideal

conditions for FHB development are not often congruent with recommended fungicide

applications, due to the physical limitation of spraying after rain events. While it is

important to continue to actively monitor anthesis dates, our study showed that fungicides

are still able to adequately control FHB and DON when applied up to 6 days post-

anthesis. This implies fungicide application times could potentially be expanded, which

would allow producers the flexibility of applying treatments after rain events. While this

study presents consistently lower IND and DON in post-anthesis fungicide applications

relative to untreated controls, further research is still needed to determine if these effects

are consistent across wheat classes and growing regions across the United States.

Table 2.1 Summary of inoculation and treatment application protocols for experiments conducted to evaluate the effects of post-anthesis fungicide treatments on Fusarium head blight of wheat

Inoculationb Fungicide Treatmentc Expa Date Density Fungicide Timing Date WO11 5/30/11 1,2,3,4 x 104 TEBU+PROT A 5/30/11 A+2 6/1/11 A+4 6/3/11 A+6 6/5/11 WO12 5/18/12 2,6,10,14 x 104 TEBU+PROT A 5/18/12 A+2 5/20/12 A+4 5/22/12 A+6 5/24/12 WO13 5/24/13 2,5,8,12 x 104 TEBU+PROT; METC A 5/24/13 A+2 5/26/13 A+4 5/28/13 A+6 5/30/13 SC11 Naturally infected TEBU+PROT; METC A 6/1/11 A+2 6/3/11 A+5 6/6/11 SC13 5.29.13 2,10 x 104 TEBU+PROT; METC A 5/28/13 A+2 5/30/13 A+4 6/1/13 A+6 6/3/13 IL12 5/4/12 2,12 x 104 TEBU+PROT; METC A 5/3/12 A+2 5/5/12 A+3 5/6/12 A+4 5/7/12 A+6 5/9/12 IL13 5/24/13 2,12 x 104 TEBU+PROT; METC A 5/23/13 A+2 5/25/13 A+4 5/27/13 A+6 5/29/13

aField experiments conducted between 2011 and 2013 at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio (WO11, WO12, and WO13); OARDC Western Agricultural Research Station near South Charleston, Ohio (SC11 andSC13); and at the Illinois Council for Food and Agricultural Research facility (IL12 and IL13). bPlots were either artificially inoculated or naturally infected. Artificial inoculation was done by spray- applying ascospore and/or macroconidia suspensions of Fusarium graminearum of densities ranging from 2-14 x 104 spores/ml using backpack sprayers. Applications were made at mid- or late-anthesis on the dates listed. cFungicide treatments consisted of the application of 19% prothioconazole + 19% tebuconazole (TEBU+PROT) or 8.6% metconazole (METC) at anthesis (A, Feekes growth stage 10.5.1), or at 2, 3, 4, 5 or 6 days after anthesis (A+2, A+3, A+4, A+5 and A+6, respectively), on the dates indicated.

Table 2.2 Summary statistics from linear mixed model analyses of data from field experiments conducted in multiple location- year to evaluate the effects of post-anthesis fungicide treatments (TRT) on Fusarium head blight index (IND), incidence (INC), Fusarium damaged kernel (FDK) and deoxynivalenol (DON) in soft red winter wheat as influenced by wheat cultivar (CV), fungicide active ingredient (FUN) and inoculum density (INOC)

INDc INCc FDKc DONc Experimenta Factorsb F valued Pd F value P F value P F value P WO11 CV 10.04 0.087 3.99 0.184 47.56 0.020 39.40 0.024 TRT 5.47 0.006 3.74 0.025 28.06 <0.001 31.77 <0.001 TRT*CV 0.70 0.601 0.09 0.985 2.04 0.137 1.23 0.336 INOC 0.23 0.875 0.21 0.890 1.42 0.246 0.04 0.987 CV*INOC 0.71 0.550 1.32 0.277 1.04 0.380 0.78 0.510 TRT*INOC 0.46 0.931 0.60 0.831 1.40 0.190 0.73 0.716 TRT*CV*INOC 1.05 0.421 1.23 0.285 1.02 0.443 1.19 0.314 WO12 CV 55.58 <0.001 68.92 <0.001 … …e … … TRT 6.33 0.001 5.44 0.002 … … … … 37 TRT*CV 0.83 0.512 0.73 0.577 … … … … INOC 0.62 0.627 1.06 0.438 … … … … CV*INOC 3.28 0.031 3.09 0.039 … … … … TRT*INOC 1.65 0.129 1.62 0.141 … … … … TRT*CV*INOC 1.74 0.096 1.61 0.131 … … … … WO13 FUN 3.75 0.057 3.51 0.098 0.38 0.558 1.94 0.168 TRT 53.07 <0.001 45.60 <0.001 27.34 <0.001 11.90 <0.001 TRT*FUN 1.73 0.154 2.02 0.103 0.27 0.895 1.45 0.228 INOC 1.82 0.243 1.86 0.238 0.23 0.875 2.50 0.155 FUN*INOC 0.58 0.627 2.04 0.187 1.30 0.346 0.78 0.509 TRT*INOC 1.05 0.417 1.35 0.214 0.24 0.995 0.90 0.554 TRT*FUN*INO C 0.96 0.496 0.98 0.478 1.41 0.184 1.70 0.086 SC11 TRT 24.63 <0.001 15.01 <0.001 8.95 <0.001 11.06 <0.001 SC13 TRT 5.93 <0.001 3.76 0.001 … … 5.06 <0.001 INOC 4.57 0.038 0.26 0.656 … … 0.66 0.421 TRT*INOC 0.86 0.567 0.91 0.529 … … 1.76 0.104 IL12 TRT 4.24 <0.001 1.92 0.025 0.78 0.729 1.67 0.061 IL13 TRT 3.06 0.001 1.52 0.126 1.40 0.176 1.41 0.174

Continued

Table 2.2: Continued

aField experiments conducted between 2011 and 2013 at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio, under artificial inoculation (WO11, WO12, and WO13); at the OARDC Western Agricultural Research Station near South Charleston, Ohio, under natural infection (SC11) and artificial inoculation (SC13); and at the Illinois Council for Food and Agricultural Research facility under artificial inoculation (IL12 and IL13). bTRT = treatments consisting of an untreated check, an application made at anthesis (Feekes growth stage 10.5.1), plus one or more post-anthesis treatments (2, 3, 4, 5 or 6 days after anthesis); CV = wheat cultivars with different levels of resistance to FHB (Hopewell, susceptible and Malabar, moderately susceptible); INOC = concentration of Fusarium graminearum inoculum used to spray inoculate plots; and FUN = DMI fungicides 8.6% metconazole and 19% prothioconazole + 19% tebuconazole. For the IL12 and IL13 experiments, treatments consisted of various combinations of TRT x INOC x FUN. For SC11 and SC13 treatments consisted of various combinations of TRT x FUN. cIND = Fusarium head blight index (mean proportion of diseased spikelets per spike); INC = Fusarium head blight incidence (mean proportion of diseased spikes); FDK = Fusarium damaged kernels (percentage of visually shriveled, small, and discolored kernels), DON = deoxynivalenol content of harvested grain (ppm). dF = F statistic and P = probability (level of significant) based on mixed model analyses of arcsine-transformed IND, INC and 38 FDK and log-transformed DON data. eMissing because data were not collected, or in the case on DON, levels were too low (<0.06 ppm) for evaluation of treatment effects.

Table 2.3 Summary statistics from linear mixed model analyses of post-anthesis treatments effects on Fusarium head blight index (IND) and deoxynivalenol content of harvested grain (DON), showing F statistics and P values for comparisons between treatments or treatment groups

INDc DONc Experimenta Contrastb Diffd F value (P)e Diff F value (P) WO11 Check vs. A 5.48 16.70 (<0.001) 0.79 76.80 (<0.001) Check vs. A+2 5.07 14.06 (0.002) 0.77 84.30 (<0.001) Check vs. A+4 4.99 11.94 (0.003) 0.79 84.58 (<0.001) Check vs. A+6 4.66 10.21 (0.006) 0.77 70.24 (<0.001) A vs. A+2 -0.41 0.11 (0.741) -0.02 0.17 (0.681) A vs. A+4 -0.49 0.40 (0.537) 0.00 0.19 (0.671) A vs. A+6 -0.82 0.79 (0.386) -0.02 0.15 (0.707) WO12 Check vs. A 0.69 8.33 (0.007) …f … Check vs. A+2 1.00 23.79 (<0.001) … … Check vs. A+4 0.54 4.74 (0.037) … … Check vs. A+6 0.75 10.61 (0.003) … … A vs. A+2 0.30 3.96 (0.055) … … A vs. A+4 -0.17 0.55 (0.463) … … A vs. A+6 0.06 0.14 (0.713) … … WO13 Check vs. A 1.94 44.41 (<0.001) 0.12 0.33 (0.325) Check vs. A+2 3.03 130.64 (<0.001) 0.95 32.53 (<0.001) Check vs. A+4 3.07 137.31 (<0.001) 0.77 13.27 (<0.001) Check vs. A+6 3.04 142.07 (<0.001) 0.76 16.93 (<0.001) A vs. A+2 1.09 22.71 (<0.001) 0.83 26.35 (<0.001) A vs. A+4 1.13 25.54 (<0.001) 0.65 9.55 (0.003) A vs. A+6 1.09 27.62 (<0.001) 0.65 12.56 (<0.001) SC11 Check vs. A 8.10 59.36 (<0.001) 4.58 43.3 (<0.001) Check vs. A+2 10.31 119.88 (<0.001) 4.80 52.42 (<0.001) Check vs. A+5 9.41 90.0 (<0.001) 3.48 16.38 (<0.001) A vs. A+2 2.21 10.52 (0.004) 0.22 0.43 (0.517) A vs. A+5 1.32 3.18 (0.088) -1.10 6.41 (0.019) SC13 Check vs. A 5.69 28.72 (<0.001) 0.93 10.45 (<0.001) Check vs. A+2 5.88 36.24 (<0.001) 1.05 16.20 (<0.001) Check vs. A+4 4.66 17.32 (<0.001) 1.05 14.78 (<0.001) Check vs. A+6 4.72 16.65 (<0.001) 1.11 14.98 (<0.001) A vs. A+2 0.32 0.75 (0.389) 0.14 0.98 (0.329) A vs. A+4 -1.06 1.56 (0.218) 0.12 0.68 (0.414) A vs. A+6 -0.98 1.39 (0.243) 0.16 0.92 (0.344) IL12 Check vs. A 22.50 41.78 (<0.001) nsg ns Check vs. A+2 25.91 57.98 (<0.001) ns ns Check vs. A+3 25.78 58.71 (<0.001) ns ns Check vs. A+4 25.13 48.94 (<0.001) ns ns Check vs. A+6 21.78 34.54 (<0.001) ns ns A vs. A+2 3.41 1.99 (0.163) ns ns A vs. A+3 3.28 2.15 (0.147) ns ns A vs. A+4 2.63 0.42 (0.518) ns ns A vs. A+6 -0.72 0.52 (0.475) ns ns IL13 Check vs. A 10.30 15.2 (<0.001) ns ns Check vs. A+2 8.69 10.90 (0.002) ns ns Check vs. A+4 13.38 28.32 (<0.001) ns ns Check vs. A+6 14.19 31.43 (<0.001) ns ns A vs. A+2 -1.62 0.61 (0.437) ns ns A vs. A+4 3.07 2.69 (0.107) ns ns A vs. A+6 3.88 3.93 (0.053) ns ns

39 Continued

Table 2.3: Continued aField experiments conducted between 2011 and 2013 at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio, under artificial inoculation (WO11, WO12, and WO13); at the OARDC Western Agricultural Research Station near South Charleston, Ohio, under natural infection (SC11) and artificial inoculation (SC13); and at the Illinois Council for Food and Agricultural Research facility under artificial inoculation (IL12 and IL13). bCheck = untreated check, A = treatment applied at 50% anthesis (Feekes 10.5.1); A+2, A+3, A+4, A+5 and A+6 = treatments applied at 2, 3, 4, and 6 days after anthesis, respectively. cIND = Fusarium head blight index (mean proportion of diseased spikelets per spike) and DON = deoxynivalenol content of harvested grain (ppm). dDiff = difference between treatment means on the raw data scale. Contrasts (e.g. differences) were constructed using LS Means based on transformation of the original data, however, for presentation, differences of original means are shown. eF = F statistic and P = probability (level of significant) based on mixed model analyses of arcsine-transformed IND and log-transformed DON data. fNot compared because mean DON in the untreated check was less than 1 ppm gNot compared because the effect of treatment was not statistically significant.

Table 2.4 Log of the response ratio, percent control and corresponding statistics for the effect of fungicide treatments on Fusarium head blight index and deoxynivalenol (DON) in soft red winter wheat

Effect Sizec Mean percent controld a b Response Treatment L̅ SE(L̅ ) ClL ClU Z P C̅ ClL CLU INDEX A vs. CK -0.82 0.113 -1.04 -0.60 -7.27 <0.001 55.85 44.94 64.60 A+2 vs. CK -1.18 0.155 -1.48 -0.88 -7.62 <0.001 69.29 58.38 77.34 A+4 vs. CK -0.96 0.142 -1.24 -0.68 -6.76 <0.001 61.61 49.32 70.92 A+6 vs. CK -0.97 0.112 -1.19 -0.75 -8.65 <0.001 61.96 52.63 69.45 A vs. A+2 -0.36 0.097 -0.55 -0.17 -3.74 <0.001 30.45 15.84 42.53 A vs. A+4 -0.14 0.083 -0.30 0.02 -1.69 0.091 13.06 -2.24 26.06 A vs. A+6 -0.15 0.075 -0.30 0.00 -1.98 0.048 13.83 0.12 25.66 DON A vs. CK -0.69 0.149 -0.98 -0.39 -4.61 <0.001 49.61 32.56 62.35 A+2 vs. CK -0.77 0.138 -1.04 -0.50 -5.58 <0.001 53.58 39.18 64.56 A+4 vs. CK -0.74 0.117 -0.97 -0.51 -6.34 <0.001 52.43 40.14 62.20 A+6 vs. CK -0.65 0.101 -0.85 -0.46 -6.45 <0.001 48.02 36.57 57.40 41 A vs. A+2 -0.08 0.076 -0.23 0.07 -1.07 0.284 7.87 -7.03 20.69 A vs. A+4 -0.06 0.079 -0.21 0.09 -0.72 0.469 5.59 -10.34 19.22 A vs. A+6 0.03 0.086 -0.14 0.20 0.36 0.718 -3.17 -22.23 12.92

a Fusarium head blight index (proportion of diseased spikelets per spike) and deoxynivalenol (DON) from harvested grain (ppm) b Fungicide treatments applied at anthesis (A) or at 2 (A+2), 4 (A+4), or 6 (A+6) days after anthesis. CK = untreated check c L̅ = mean log of the response ratio between each treatment and the check and between post-anthesis and anthesis treatments; SE( L̅ ) = standard error of L̅ ; Z = standard normal test statistic; P = significance level; CIU and CIL = upper and lower limits of the 95% confidence interval around L̅ . d C̅ = mean percent control as estimated from L̅ as C̅ = [1-exp(L̅ )]•100; CIU and CIL = upper and lower limits of the 95% confidence interval around C̅ .

Fig. 2.1 Fusarium head blight index (mean proportion of diseased spikelets per spike) for different treatments applied to soft red winter wheat (SRWW) in field experiments conducted at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio in 2011 (A) 2012 (B) and 2013 (C); at the OARDC Western Agricultural Research Station near South Charleston, Ohio in 2011 (D) and 2013 (E); and at the Illinois Council for Food and Agricultural Research facility in 2012 (F) and 2013 (G). Treatments were Check = Untreated control (CK), A = fungicide treatment (8.6% metconazole [METC] or 19% tebuconazole + 19% prothioconazole [TEBU+PROT]) applied at 50% anthesis and A+2, A+3, A+4, A+5, and A+6 = applications made at 2, 3, 4, 5, and 6 days after anthesis, respectively. Malabar and Hopewell = moderately resistant and susceptible SRWW cultivars. Bars represent means across 4 to 24 experimental units (i.e., across the levels of the other treatment factors in the study) and their corresponding standard error.

14 3.5 12 A WO11 Check 3.0 B WO12 Check Malabar Malabar 10 Hopewell 2.5 Hopewell 8 2.0 6 1.5 4 1.0 2 0.5 0 0.0 CK A A+2 A+4 A+6 CK A A+2 A+4 A+6 5 16 C WO13 Check 14 D SC11 Check 4 TEBU+PROT TEBU+PROT 12 METC METC 3 10 8 2 6 4 1 2 0 0 CK A A+2 A+4 A+6 CK A A+2 A+5 12 50 Check Check 10 E SC13 F IL12 TEBU+PROT 40 TEBU+PROT METC 8 METC 30 6 Fusarium head blight index (%) index blight head Fusarium 20 4

2 10

0 0 CK A A+2 A+4 A+6 CK A A+2 A+3 A+4 A+6 40 G IL13 Check 30 TEBU+PROT METC

0 CK A A+2 A+4 A+6 Figure 2.1

Fig. 2.2 Deoxynivalenol content of harvested grain for different treatments applied to soft red winter wheat (SRWW) in field experiments conducted at the Ohio Agricultural Research and Development Center (OARDC) Snyder Farm near Wooster, Ohio in 2011 (A) and 2013 (B), at OARDC Western Agricultural Research Station near South Charleston, Ohio in 2011 (C) and 2013 (D) and at the Illinois Council for Food and Agricultural Research facility in 2013 (E). Treatments were Check = untreated check (CK), A = fungicide treatment (8.6% metconazole [METC] or 19% tebuconazole + 19% prothioconazole [TEBU+PROT]) applied at 50% anthesis and A+2, A+4, A+5, A+6 = applications made at 2, 4, 5, and 6 days after anthesis, respectively. Malabar and Hopewell = moderately resistant and susceptible SRWW cultivars. Bars represent means across 4 to 24 experimental units (i.e., across the levels of the other treatment factors in the study) and their corresponding standard error.

2.0 3.0 1.8 A WO11 Check B WO13 Check 1.6 Malabar 2.5 TEBU+PROT Hopewell METC 1.4 2.0 1.2 1.0 1.5 0.8 0.6 1.0 0.4 0.5 0.2 0.0 0.0 CK A A+2 A+4 A+6 CK A A+2 A+4 A+6 10 3.0 C SC11 Check D SC13 Check 8 TEBU+PROT 2.5 TEBU+PROT METC METC 2.0 6 1.5 4 1.0

2 0.5

Deoxynivalenol (ppm) Deoxynivalenol 0 0.0 CK A A+2 A+5 CK A A+2 A+4 A+6 12 E IL13 Check 10 TEBU+PROT METC 8

0 CK A A+2 A+4 A+6

Figure 2.2

Chapter 3

Efficacy of Post-Anthesis Fungicide Application against Fusarium Head Blight and Deoxynivalenol in Soft Red Winter Wheat

3.1 INTRODUCTION

Fusarium head blight (FHB), a fungal disease of wheat and other small grain crops, is caused predominantly by Fusarium graminearum in North America. FHB, or head scab as it is commonly known, occurs in virtually all major wheat-growing regions of the world, causing millions of dollars in grain yield and quality losses as a result of floret sterility, production of small, shriveled, light-weight kernels, and contamination of grain with mycotoxins such as deoxynivalenol (DON) and nivalenol (Windels 1999).

Consumption of mycotoxin contaminated grain may lead to compromised immune functions and causes vomiting, feed refusal, reproductive disorders, and low weight gain, especially in non-ruminant animals (Rocha et al. 2005). FHB and DON also affect the milling and baking quality of flour produced from contaminated grain (Dexter et al.

1996), and since DON is heat-stable and water-soluble, it may persist in foods throughout the cooking process (Rotter et al. 1996). Because of these and other concerns, the maximum allowable limit for DON established by the United States Food and Drug

Administration is 2 ppm in harvested grain and 1 ppm in finished wheat products

destined for human consumption. Grain lots with DON above 2 ppm are either rejected

completely or priced down at grain elevators (McMullen et al. 1997).

Quinone Outside Inhibitors (QoI) and Demethylation Inhibitors (DMI) are

effective fungicides against many wheat diseases. While the DMIs are highly

recommended for FHB and DON control, the QoIs are not. This is due, in part, to the fact that fungicides belonging to the DMI group of compounds have been shown to be more effective against FHB and DON than QoIs and other fungicide chemical groups (Magan et al. 2002, Pirgozliev et al. 2002). Paul et al. (2008) and Willyerd et al. (2012) conducted

quantitative syntheses of data from FHB management trials to evaluate the efficacy of

DMI fungicides and observed that, depending on the level of disease and DON and the

resistance and market class of the cultivars treated, a single anthesis application of 41% prothioconazole (Proline 480 SC; Bayer CropScience, Research Triangle Park, NC), 19%

tebuconazole + 19% prothioconazole (Prosaro 421 SC; Bayer CropScience, Research

Triangle Park, NC) or 8.6% metconazole (Caramba 90 SL; BASF Corporation

Agricultural Products, Research Triangle Park, NC) provided up to 70% reduction of

FHB and DON.

In addition to being relatively less effective than DMIs, QoI are not recommended

for FHB and DON management because some members of this group of fungicides have

been shown to increase DON in wheat grain, especially when applied at anthesis in

limited studies (Zhang et al. 2009). However, the specific set of conditions under which

DON increases in response to QoIs are still largely unknown. For instance, it is unclear

whether DON response to QoIs is consistent across active ingredients in this class of

fungicides; whether the response is influenced by application timing and weather

47 conditions; or whether DON increase is associated with an increase in fungal colonization of the grain. Furthermore, there is conflicting information in the literature regarding the effects of QoIs on DON, with corresponding contrasting explanation for the observed results (Pirgozliev et al. 2002, Simpson et al. 2001, Edwards et al. (2001). Moreover results from some studies suggest that DMI fungicides may also be associated with mycotoxin (DON and nivalenol [NIV]) increase in wheat grain under certain conditions

(Simpson et al. 2001).

Simpson et al. (2001) used polymerase chain reaction (PCR) assays to evaluate the effects of DMI and QoI fungicides on FHB-causing pathogens on wheat in the United

Kingdom, and reported differential control of species within the pathogen complex. In particular, they observed that azoxystrobin (applied as Amistar) selectively controlled non-toxin-producing Microdochium species, but allowed the proliferation of toxin- producing Fusarium species. The also observed high levels of DON per unit biomass of

F. culmorum in azoxystrobin treated plots and concluded that this fungicide likely stimulated toxin production. Zhang et al. (2012) also used a PCR-based assay and visual symptoms to evaluate the effects of DMI and QoI fungicides on FHB, DON, and F. graminearum DNA under inoculated field conditions in China, and came to similar conclusions to those reported by Simpson et al. (2001) regarding the likely effects azoxystrobin on DON. Contrastingly, however, based on results from greenhouse inoculated studies with both F. graminearum and F. culmorum, Pirgozliev et al. (2002) determined that applications of azoxystrobin significantly reduced both FHB and DON levels compared to the untreated check, and found no evidence to support this QoI fungicide increasing DON levels in harvested grain. They further concluded based on

relationships between Fusarium biomass and DON that the effect of fungicides on DON

was directly related to their effects (or lack thereof) on the amount of trichothecene-

producing Fusarium species present in the grain. Edwards et al. (2001) came to a similar

conclusion and hypothesized that fungicides affected the FHB/DON relationship by

altering Fusarium biomass and not by changing the rate of toxin synthesis.

It is clear from the studies cited above that in-planta evaluation of the effects of

QoI fungicides on DON is complicated by the fact that DON contamination is often positively correlated with FHB symptom intensity (Paul et al. 2006 and 2007) and grain colonization with F. graminearum (Sneller et al. 2010, Pirgozliev et al. 2002, Edwards et al. 2001). Consequently, high DON levels in QoI-treated plots may be due in part to correspondingly high levels of FHB and/or grain colonization, resulting from the relatively poor efficacy of this group of fungicides against this disease and pathogen

(Magan el al. 2002, Pirgozliev et al. 2002, Simpson et al. 2001). The confounding effect of FHB and the highly variable relationship between FHB and DON may lead to spurious results in these types of studies, which could in part explain the aforementioned contrasting findings regarding the effects of QoIs on DON. Therefore, the confounding effects of FHB and grain colonization need to be taken into account when evaluating the effects of fungicides on DON. This could be done either at the time of designing the experiments or when analyzing the data.

Despite concerns about QoIs for controlling FHB and DON, these fungicides are very widely used for controlling foliar diseases of wheat. Therefore, their effect on FHB and DON needs to be quantified. The objectives of this study were to i) determine the effects of two QoI active ingredients (22.9% azoxystrobin [Quadris]; Syngenta Crop

Production INC and 23.6% pyraclostrobin [Headline]; BASF Cooperation Agricultural

Products), relative to untreated checks and DMI (19% tebuconazole + 19%

prothioconazole) reference treatments, on DON contamination of wheat grain and spikes,

ii) determine whether the effect of QoIs on FHB index and DON was influenced by

application timing, and iii) determine whether the relationship among index, DON, and

Fusarium damaged kernels varied with active ingredient and application timing. Multiple

field and greenhouse experiments were conducted under different conditions, and treatment effects were evaluated based on mixed model analyses of variance and covariance.

3.2 METHODOLOGY

Establishment and design of experiments. Field experiment. During 2011 and

2013, field experiments were conducted at the Ohio Agricultural Research and

Development Center Snyder farm near Wooster, OH, to evaluate the effects of two QoI

fungicide active ingredients (azoxystrobin and pyraclostrobin), applied at three different

growth stages, on FHB development, DON, and F. graminearum biomass in wheat. Plots

of FHB susceptible wheat cultivar Hopewell were established on 25 Sept 2010 and 17

Oct 2012, using a Kincaid planter, calibrated to plant at a seeding rate of 4 x 106 seeds/ha.

The experimental design was a randomized complete block with four replicates. Each experimental unit was 1.5 x 6 m, with 1.5 m-wide strips of Truman planted between adjacent experimental units to minimize inter-plot interference.

Greenhouse experiment. During winter 2012-2013, two greenhouse experiments were conducted at the OARDC, Wooster, OH. In August and December 2012, fungicide treated seeds of susceptible wheat cultivars Hopewell and Cooper were planted in 508 x

635 mm flats containing autoclaved silt loam. Five rows of approximately 50 seeds per row were planted in each flat, resulting in a total of 250 individual plants per flat. Flats were then transferred to a greenhouse and allowed to germinate and grow for 14 days.

Individual flats were fertilized with 15 g of Osmocote (14-14-14 N-P-K, Scott, Inc.

Marysville, OH), before being moved to a vernalization chamber, with a temperature range of 0-3.5oC, for 8-10 weeks. During that time, seedlings were watered twice a week.

After vernalization, plants were acclimated in the greenhouse for 48-72 hours, before being individually transplanted to conetainers (Stuewe and Sons, Inc. Corvallis, OR) filled with autoclaved silt loam. Transplants were maintained in the greenhouses, watered twice a day, and fertilized weekly with a mixture of 20-20-20 N-P-K. Between Feekes growth stages 5 and 8, plants were staked, thinned to 2-3 primary tillers per cone, and then transferred to temperature- and light-controlled growth chambers (23oC and 70% relative humidity, with a photoperiod of 16 hours light [50% fluorescent, 50% incandescent light) to stimulate uniform plant growth and spike development. The experimental design was a randomized complete block, and each experiment consisted of

600 plants, with 6 replicates of 10 plants per treatment. Cohorts of plants at anthesis served as the blocking factor.

Fungicide treatment application. For both the field and greenhouse experiments, treatments consisted of the applications of 23.6% pyraclostrobin (Headline, BASF

Cooperation Agricultural Products) and 22.9% azoxystrobin (Quadris, Syngenta Crop

Production INC), two QoI fungicides, and 19% prothioconazole + 19%tebuconazole, a

DMI fungicide, at different crop growth stages. The DMI fungicide treatments were included as references for comparison with the QoI treatments. In both field and greenhouse experiments, the three fungicides were applied at Feekes growth stages (GS)

8 (flag leaf emergence), GS 10 (boot), and GS 10.5.1 (anthesis), for a total of 9 unique fungicide x timing treatment combinations, plus an untreated check. Fungicides for FHB and DON are usually applied at GS 10.5.1, but an earlier application is typically made for foliar disease control.

All treatments were applied at the label-recommended rates of 438 ml/ha for

Headline, 658 ml/ha for Quadris, and 475 ml/ha for Prosaro. The nonionic surfactant,

Induce (Helena Chemical Co. Collierville, TN), was added to each treatment at a rate of

0.125% v/v, and field applications were made using either tractor mounted or backpack

sprayers, with booms fitted with Twinjet XR8002 or paired XR8001 nozzles, mounted at

an angle (45°) forward and backward, and calibrated to apply at a rate of 187L/ha at 207

kPa. In the greenhouse, fungicides were applied using a 500-ml handheld spray bottles, at

rates comparable to those used in the field. Fungicides were applied until run-off to

ensure that individual heads were completely and evenly covered.

Inoculum preparation and inoculation. Inoculum was prepared using 10

aggressive isolates of Fusarium graminearum collected from naturally infected wheat

fields in Ohio. Isolates were first grown on Komada selective media, and then transferred

to either mung bean agar (MBA) or carrot agar (CA) for macroconidia and ascospore

production, respectively. MBA plates were incubated for 7-10 days at room temperature

under ultraviolet light, with a 12 hour photoperiod, after which macroconidia were

harvested by adding sterile deionized water and gently dislodging spores with a rubber

policeman. For ascospore production, mycelial mat was removed from CA plates after 4

days of incubation to stimulate ascospore production, and cultures were then re-incubated

for an additional 10-14 days until mature perithecia were produced. Ascospores were

harvested as described above and the concentrations of both spore types were determined

using a hemacytometer (Reichert-Bright Line, Hausser Scientific. Horsham, PA 19044).

The final concentration was adjusted by adding sterile deionized water plus 2% tween 20.

Equal volumes and concentrations of ascospores and macroconidia were mixed prior to

53 inoculation. For field experiments, plots were spray-inoculated with a concentration of

100,000 spores/ml using a backpack sprayer (R&D Sprayers, Opelousas, LA), fitted with

3 Teejet Flat Fan nozzles on a 3 m-long boom, and calibrated to apply approximately 600 ml of inoculum to each plot. All inoculations were done approximately 24-36 h after the anthesis fungicide treatments were applied.

In the greenhouse, the macroconidia + ascospore mixture was prepared a minimum of 7 days in advance and stored under refrigerated conditions (2-8C). In the first experiment, a spray inoculation method was utilized, whereas in the second, spikes were inoculated using a single floret or point inoculation technique. Spray inoculation methods are used to mimic the natural infection process and point inoculations forego the physical protective layers of the plant (the lemma and palea) and assume an infection event has occurred. For both experiments, all spikes were inoculated at early to mid anthesis, when extruding anthers were distinctly yellow. For the spray inoculation of single florets experiment, a 50,000 spore/ml suspension was prepared on the day of inoculation and applied to plants isolated in an enclosed area with minimum air current in order to minimize cross contamination. Two sprays were applied to each individual spike using a spray bottle held approximately 30 cm away from the spike and at a 45° angel directed downward. A total of approximately 150-200 µl of inoculum was applied per spike. For the single floret inoculation experiment, a 10,000 spore/ml suspension was prepared on the day of inoculation and applied to a spikelet located in the middle of the spike. The glume and lemma were gently separated, and 10 µl of the inoculum was pipetted directly into the floret. After inoculation, all plants in both experiments were immediately moved into a mist chamber set to deliver 30 sec of mist every 5 min for 72

h. These conditions ensured a >90% relative humidity. Plants were then transferred back

to a greenhouse bench where they remained for the duration of the experiment (average

temperature 27°C).

Data collection and analysis. Disease Assessment and DON quantification. In all

experiments, FHB intensity was rated approximately 3 wks after anthesis, during the soft

dough stage of crop development (Feekes 11.2). FHB index was visually estimated as the

mean percentage of diseased spikelets per spike, and FHB incidence as the percentage of

diseased spikes out of all spikes evaluated. For the field experiments, mean plot index

and incidence were estimated by rating visual symptoms in five clusters of twenty spikes,

arbitrarily selected per plot (for a total of 100 spikes). In the greenhouse, all 10 spikes in

each experimental unit were visually scored for FHB and mean index and incidence were

estimated as described (Paul et al 2005, Stack and McMullen 1998). At harvest, a sample

of grain from each field plot was used to estimate Fusarium damaged kernels as the percentage of diseased kernels (small, shriveled, and discolored) in the sample.

A sample of grain from each field plot and entire spikes from each experimental unit in the greenhouse were ground using a coffee grinder (Braun Aromatic KSM 2 B,

Gillette Commercial, Boston, MA), or a laboratory mill (Laboratory Mill 3033, Perten

Instruments, Springfield, IL) for DON analysis. A sub-sample of the ground grain/spike was then sent to the US Wheat and Barley Scab Initiative mycotoxin testing laboratory at the department of plant pathology, University of Minnesota, St. Paul, MN for DON quantification using gas chromatography-mass spectrometry.

Data analysis. Data from all experiments were analyzed using PROC GLMMIX in SAS (SAS, Cary, NC) to determine the effects of treatment on all measured responses.

Prior to analysis, data from the two field experiments were pooled, and FHB incidence,

index, and FDK data were arcsine-square-root-transformed, while DON data were log-

transformed to stabilize variances. For the greenhouse studies, the point- and spray-

inoculated experiments were kept separate, and the analyses were also performed on

arcsine-square-root and log-transformed index and DON data, respectively. In all case

block (Time in the case of the greenhouse experiments) and year in the case of the fiels

experiment were treated as random effects, whereas, fungicide treatment was the fixed

effect. Contrast and lsmestimate statements in GLIMMIX were employed to compare each Qol treatment with the untreated check and the DMI anthesis reference treatment.

Since FHB visual symptoms (such as IND and FDK) are known to have a direct but highly variable relationships with DON contamination, the effects of fungicide

treatments on the latter response could be confounded by differences in index and FDK

among experimental units. It is informative to determine the extent to which DON is

affected by treatment at a fixed level of disease. Therefore, before comparing mean

logDON among fungicide treatments, the means were adjusted for the effect of FDK for

the field experiments and for the effect of index for the greenhouse experiments. This

was accomplished by adding FDK (or index) as a continuous covariate to the linear

mixed models used to analyze the effect of fungicide treatments on logDON, and then

using the FDK/logDON (or index/logDON) relationship to estimate and compare mean

logDON among treatments at a common level of FDK or index.

3.3 RESULTS

Fusarium head blight index and deoxynivalenol levels. In both field and greenhouse experiments, mean FHB index (IND, mean proportion of diseased spikelets per spike) and DON varied among active ingredient and among application timing for any given active ingredient (Fig. 3.1). Averaged across the two years (2011 and 2013), mean IND ranged from 1.1% for plots treated with Prosaro at anthesis (P_10.5.1) to 6.5%

for the untreated check, the two reference treatments. Means for all other treatments fell

within the range of the two reference treatments. Mean DON levels in the two reference

treatments, averaged across the two year, were 0.85 and 2.23 ppm, for P_10.5.1 and the

untreated check, respectively. With the exception of Prosaro at Feekes 10 (P_10) and

Feekes 10.5.1 (P_10.5.1), all other treatments had higher mean DON than the check (Fig.

3.1B).

The 2012-2013 studies under greenhouse conditions displayed much wider ranges

of disease index and DON between the reference treatments than observed in the field

experiments. For instance, average IND levels for the reference treatments in spray-

inoculated plants ranged from 5.3 to 59.3%, and 7.3 to 32.7% in point inoculated plants

(Fig. 3.1C and 3.1E). The wider ranges of disease in the controlled-environment studies

may be attributed to the fact that plants were hand selected, and spikes were individually inoculated at the appropriate growth stage (Feekes 10.5.1), and then subjected to intermittent mist for 72 hours to enhance infection. As was the case in the field, all other

57 treatments had mean index within the range of the reference treatments, with only one instance of mean IND in a Feekes 8 (H_8) treatment being higher than the check (Fig.

3.1E). The range of mean DON between the reference treatments for the point- inoculation experiments was 0.80 to 15.69 ppm; for spray inoculation, the range was 0.01 to 18.12 ppm (Fig. 3.1D and 3.1F). In spray inoculated plants, some QoI treatments reached mean levels of up to 9.65 ppm greater DON than the untreated inoculated check.

For point inoculations, DON levels were as high as 10.34 ppm greater in some of the QoI treatments than in the check.

Although wider IND and DON ranges were seen under controlled conditions than in the field, very similar trends in terms of disease and toxin responses to application timing were observed in all greenhouse and field experiments (Fig. 3.1). For both of the

QoI fungicides as well as the DMI fungicide, there was an overall reduction in IND and

DON between the early application times and the anthesis treatments. The field data showed reductions in IND from Feekes 8 to Feekes 10.5.1 applications for Prosaro,

Headline, and Quadris (Fig. 3.1A). Greenhouse data showed similar IND (Fig. 3.1C and

3.1E) reductions from Feekes 8 to Feekes 10.5.1 for spray as well as point inoculations.

When comparing fungicide timing within fungicide AI’s, the highest levels of DON were seen in Feekes 8 and 10 treatments, and the lowest levels tended to be in the anthesis treatments. In the greenhouse spray-inoculated experiment, DON levels ranged from 0.28 to 23.9 ppm, 5.71 to 27.77 ppm, and 0.01 to 13.95 ppm between application timing for

Headline, Quadris, and Prosaro, respectively. Ranges were similar in the point-inoculated experiment. In the field, similar trends were seen; however, DON ranges were much

narrower (2.7 to 3.2 ppm, 2.84 to 4.15 ppm, and 0.85 to 2.51 ppm for Headline, Quadris,

and Prosaro, respectively).

Fungicide effects on Fusarium head blight and deoxynivalenol. Results from

linear mixed model analyses of data from the field experiments showed that the random

effect of year was not statistically significant for any of the measured responses (P = 0.29 for logDON, 0.25 for arcIND and 0.44 for arcFDK), based on standard normal tests.

Therefore, all reported field results are based on analyses of the pooled data. For both field and greenhouse experiments, the effect of fungicide treatment was statistically significant (P ≤ 0.05) for arcIND, and for arcFDK for the field experiment. All anthesis treatments, regardless of active ingredient (P_10.5.1, H_10.5.1, and Q_10.5.1) and the

P_10 treatment (Prosaro at Feekes 10) led to a significant reduction of arcIND relative to the check (Table 3.1). However, with the exception of P_10.5.1 and P_10 none of the other treatments resulted in a significant reduction of arcFDK compared to the check. All treatment had significantly higher arcIND and arcFDK than the other reference treatment,

P_10.5.1 (Table 3.1). The only exception was for P_10 which was not significantly different from P_10.5.1 for arcIND.

In both greenhouse studies, similar results to those observed in the field were seen in the greenhouse in terms of treatment effects of arcIND. As was the case in the field experiments, all anthesis treatments, regardless of active ingredient, reduced arcIND relative to the check for both the point- and spray-inoculated experiments (Table 3.2).

However, contrary to what was observed in the field, several of the early applications of all three active ingredients, with the exception of the H_8 and Q_10 treatment for the

point-inoculation experiments and the Q_8 treatment for both experiments, had

significantly lower arcIND than the check.

Results from pairwise comparisons between each treatment and the DMI

reference treatment (P_10.5.1) were also similar to those observed in the field, with all

but two of the treatments (P_10 and H_10.5.1) having significantly higher arcIND than

P_10.5.1 treatment. FDK was not quantified in the greenhouse due to the limited

availability of harvested kernels to effectively rate this response.

Results from the covariance analyses showed that the FDK x logDON interaction

for the field experiment and IND x logDON interactions for the greenhouse experiments

were not statistically significant (P > 0.05). Therefore, common-slope (b) models of the

form logDON = ai + bX (X = FDK or IND, the covariate) were fitted to each of the three

data sets, with separate intercepts (ai) for each treatment. The heights of the fitted

regression lines, i.e. the intercepts or logDON at a fixed level of FDK or IND, were then

compared among treatments using the lsmestimate statement in GLIMMIX. Summary statistics for those comparisons are reported in Tables 3.1 and 3.2 as adjusted DON

(DON_adj). These adjustments were based on the mean for each covariable. For the field experiment, the difference in logDON at a fixed level of FDK (adjusted logDON) varied among treatments, with all treatments (except for P_10.5.1) having numerically, but not always statistically, higher adjusted logDON than the check (Table 3.1 and Fig. 3.2).

Both QoI anthesis treatments (H_10.5.1 and Q_10.5.1) as well as the Q_10 treatments

(Quadris at Feekes 10) had significantly higher adjusted logDON than the check.

Similarly, many of the QoI applications (except for H_10.5.1) in both greenhouse experiments had higher, but not significantly different, adjusted logDON compared to the

60 untreated check. In fact, contrary to what was observed in the field, for both spray- and point-inoculation experiments, the difference in adjusted logDON between the H_10.5.1 and P_10.5.1 (DMI reference treatment and industry standard for FHB and DON control) treatments was not significantly different (Table 3.2, P = 0.443 and P = 0.104 respectively, for spray and point inoculations).

3.4 DISCUSSION

This study was conducted largely to revisit the question of deoxynivalenol (DON)

increase in response to Quinone Outside Inhibitor (QoI) fungicides. QoIs have not been

recommended for Fusarium head blight management since some members of this group

have been reported in limited studies to increase DON in wheat grain following

treatment. Most previous studies only evaluated a single QoI active ingredient when

applied at a single growth stage, and few evaluated the confounding effect of FHB

symptoms when evaluating DON response. Here we compared two QoIs at three

different growth stages, and adjusted for the effect of FHB index (FHB) and Fusarium

damaged kernels before evaluating and comparing treatment effects of DON. Comparing regression lines among treatments provided a better indication of expected DON levels given the amount of visual symptoms. In the field studies, FDK/DON relationships were examined, and IND/DON relationships were evaluated in greenhouse studies. Both in the field and in the greenhouse, all treatments applied at anthesis (10.5.1) significantly reduced IND relative to the check, but only the DMI significantly reduced FDK. Relative to the DMI anthesis treatment (the industry standard for FHB control), all QoI-based treatments, including those applied at anthesis, showed inferior performance in the field, confirming previous reports of relatively lower efficacy of QoIs against FHB when compared to DMIs (Edwards et al. 2001, Pirgozliev et al. 2002).

Both in the field and in the greenhouse, fungicide treatment did not have an effect

of the rate of DON increase per unit increase in FDK or IND, the slope of the FDK/DON

or IND/DON relationships, suggesting that there was no evidence for greater or lesser

change in DON contamination for any particular treatment as FDK or IND increased (or

decreased). However, the height of the regression line, the level of DON at a fixed level

of disease, varied significantly among treatments. In the field, all of the treatments

resulted in higher DON than the 19% prothioconazole + 19% tebuconazole treatment

(P_10.5.1), the industry standard. However, both QoI anthesis treatments (23.6%

pyraclostrobin and 7% azoxystrobin applied at Feekes growth 10.5.1; the H_10.5.1 and

Q_10.5.1, respectively), as well as the 7% azoxystrobin treatment applied at boot stage

(Q_10) had significantly higher regression lines, (i.e. higher adjusted DON), than the

untreated check and the 19% tebuconazole + 19% prothioconazole anthesis (P_10.5.1) reference treatments. Although the former two treatments provided a significant reduction in IND (visual symptoms), there is sufficient evident here to suggest that they both increased DON in the field. These results are consistent with those reported by

Ruske et al. (2003) and Blandino et al. (2009) who also found disproportionately high amounts of DON compared to visual disease in response to QoIs.

As with the field study, in the greenhouse, DON and adjusted DON for the reference treatment (P_10.5.1) was lower than for most of the other treatments. However, the results observed in the field in terms of the differences in heights of the regression lines between anthesis QoI treatment and the check were somewhat different in the greenhouse. The height of the regression lines (and corresponding adjusted logDON) for many of the treatments was similar to, and not significantly different from, the control

line. In both greenhouse experiments, the Q_10 and Q_10.5.1 treatments consistently had two of the highest DON regression lines, but these were not significantly different from the check, suggesting the lack of significant effect on DON. However, these same treatments had significantly higher regression lines than DMI anthesis reference treatment (P_10.5.1). This was more than likely the result of the P_10.5.1 treatment significantly reducing DON rather than the Q_10 and Q_10.5.1 treatments increasing it in the greenhouse. Differences between results observed in the field and the greenhouse could be attributed to several factors, including the relatively longer anthesis-to-harvest window (45 days) in the field compared to the greenhouse (35 days). Ruske et al. (2003) and Blandino et al. (2009) hypothesized that a wider anthesis-to-harvest window (delayed senescence), likely gives the fungus more time to grow and colonize which could result in increased DON production. Differences between the greenhouse and the field could have also been due to the fact that FDK was used to adjust for disease in the field while

IND was used in the greenhouse. Paul et al. (2005) shows that FDK generally has a stronger relationship with DON and IND, which could have affected the results.

From this study, there was little evidence to suggest that Feekes 8-9 applications of QoI increased DON in wheat, suggesting that application timing may matter in terms of DON response to QoIs. This was probably because Feekes 8-9 is too far removed from anthesis, time of infection and initial DON production by the pathogen, and even further removed from the time of grain harvest. Consequently, whatever the effects of QoIs are on DON, whether it is greater colonization (Simpson et al. 2001, Zhang et al. 2012), delayed senescence (Ruske et al. 2001, Blandino et al. 2009), or increased DON production (Edwards et al. 2001, Pirgozliev et al. 2002), these likely would have been

reduced or ineffective due to natural degradation of the fungicide between flag leaf

emergence and infection. Even in the greenhouse where the time between Feekes 8 and

anthesis was much shorter (14 days) then in the field (28 days) the effects of Feekes 8

treatments on DON (adjusted for IND) was not significant, although numerically the

DON means were higher for GS8.

The consistently (and often significantly) lower levels of DON at a fixed level of index confirms that the DMI anthesis treatment (P_10.5.1) not only controlled DON by

controlling IND and FDK, it also reduced DON. That is, even adjusting for a fixed value

of disease, DON was still reduced. This could be due to reduced fungal colonization of

the grain or simply by reduction in toxin production by the fungus. The opposite could be

said for the QoI treatments with increased DON in the field. Elevated DON was not

merely due to the inferior efficacy of these treatments against FDK; they could cause an

increase in DON by allowing greater and longer colonization of the grain, promoting

increased DON production by the fungus or some other unknown mechanism. Both

increased colonization by toxin-producing Fusarium species and stimulation of DON

production in response to QoIs have been postulated as explanation for DON increase following QoI application (Edwards et al. 2001, Simpson et al. 2001, Zhang et al. 2009).

However, the real reason(s) for the positive DON responses to QoI, when they occur, still remains largely unknown. For instance, since previous studies have shown that early QoI applications may led to delayed senescence (and as a result, higher grain yields) (Ruske et al. 2003, Blandino et al. 2009), some speculate that delayed senescence in QoI treated plots provides more of an opportunity for Fusarium to colonize the grain and produce more toxins. However, using grain moisture as a surrogate for senescence in

our field experiments, we observed that all fungicide treatments, regardless of chemistry,

resulted in elevated grain moisture relative to the check (results not shown), with the highest levels of moisture in Feekes 10.5.1. Consequently adjusting for grain moisture did not affect the results. Other studies have attributed elevated levels of DON in QoI treatments to fungicides selectively controlling non-toxin producing microdochium species (Simpson et al. 2001, Zhang et al. 2012); however, in Ohio, other species do not predominate in the FHB species complex, so this would likely not be an explanation for the responses we observed.

QoI fungicides are generally not recommended for FHB and DON management. Our findings support this because P_10.5.1 was generally far better than the other treatments at reducing DON. The QoI fungicides, however, are recommended for the control of powdery mildew, Septoria leaf blight, Stagonospora leaf and glume blotch, rust and other foliar diseases. The fact that a single or a few QoIs were previous shown to be associated with increased DON is not sufficient justification to completely remove the class of chemistry from wheat disease manage programs altogether. These and other chemistries are needed in combination with the DMIs as part of good, responsible fungicide resistance management programs (Hobbelen et al. 2011). In this study we attempted to gain a more comprehensive look at the effect of QoIs on FHB and DON, and what could be driving increased DON levels without any apparent increase in visual disease, or an increase or no change in DON when disease symptoms decrease, on average. Our results suggest that there is still reason for concern in terms of elevated DON contamination levels when the two QoIs evaluated here are applied close to or at anthesis in the field, especially when conditions are conducive to F. graminearum infection and FHB

66 development such as those imposed here through inoculation and misting. However, our results also suggested that 23.6% pyraclostrobin and 7% azoxystrobin could continue to be used for early-season disease management with minimal risk of DON increase due to these treatments, even if conditions for FHB becomes favorable later in the season.

Table 3.1 Pairwise comparisons of untreated check and DMI reference treatments with QoI treatments from linear mixed model analyses of fungicide effects on Fusarium head blight index (IND), Fusarium damaged kernels (FDK), deoxynivalenol (DON), and Fusarium graminearum biomass (FBM) from wheat plots spray-inoculated at anthesis with a spore suspension of Fusarium graminearum under field conditions

INDb FDKb DONb DON_adjb Comparisonsa Diffc Pc Diff P Diff P Diff P H_8 vs. Check -1.08 0.719 -1.25 0.629 0.48 0.156 0.18 0.057 H_10 vs. Check -2.06 0.213 1.32 0.486 0.65 0.062 0.17 0.095 H_10.5.1 vs. Check -3.50 0.011 -0.38 0.882 0.98 0.010 0.30 0.003 Q_8 vs. Check -0.88 0.785 -1.75 0.460 0.61 0.174 0.19 0.052 Q _10 vs. Check -0.33 0.945 1.72 0.528 1.93 <0.001 0.44 <0.001 Q _10.5.1 vs. Check -3.77 0.004 -0.66 0.762 0.74 0.035 0.25 0.011 P_8 vs. Check -0.78 0.895 -3.50 0.191 0.29 0.499 0.15 0.116 68 P_10 vs. Check -4.54 <0.001 -6.31 0.010 -0.04 0.887 0.13 0.204 P_10.5.1 vs. Check -5.39 <0.001 -12.09 <0.001 -1.38 <0.001 -0.27 0.017 H_8 vs. P_10.5.1 4.31 <0.001 10.84 <0.001 1.85 <0.001 0.45 <0.001 H_10 vs. P_10.5.1 3.33 <0.001 13.41 <0.001 2.03 <0.001 0.44 <0.001 H_10.5.1 vs. P_10.5.1 1.89 0.006 11.72 <0.001 2.35 <0.001 0.57 <0.001 Q_8 vs. P_10.5.1 4.51 <0.001 10.34 <0.001 1.99 <0.001 0.46 <0.001 Q_10 vs. P_10.5.1 5.07 <0.001 13.81 <0.001 3.30 <0.001 0.71 <0.001 Q_10.5 vs. P_10.5.1 1.62 0.016 11.44 <0.001 2.12 <0.001 0.52 <0.001 P_8 vs. P_10.51 4.61 <0.001 8.59 <0.001 1.67 <0.001 0.42 <0.001 P_10 vs. P_10.5.1 0.85 0.117 5.78 0.005 1.34 <0.001 0.40 <0.001 Check Mean 6.49 … 18.72 … 2.23 … 0.85 … P_10.5.1 Mean 1.09 … 6.63 … 0.85 … 0.58 …

Continued

Table 3.1: Continued

aFungicide treatment and application times - 23.6% pyraclostrobin (438 ml/ha) at flag leaf emergence (H8), boot (H10), and anthesis (H10.5.1); 7% azoxystrobin (658 ml/ha) at flag leaf emergence (Q8), boot (Q10) and anthesis (Q10.5.1); 19% tebuconazole + 19% prothioconazole at flag leaf emergence (P8), boot (P10), and anthesis (P10.5.1). Check and P_10.5.1 = untreated and treated reference, respectively. bIND = Fusarium head blight index (mean proportion of diseased spikelets per spike), FDK = Fusarium damaged kernels (percentage of visually shriveled, light-weight, and discolored kernels), DON = deoxynivalenol content of harvested grain (ppm), and DON_adj = log-transformed DON adjusted for the effect FDK as the covariate. cDiff = difference between pairs of means for IND, FDK, DON, and log-transformed DON and P = probability (level of significant) based on linear mixed model analyses of variance for arcsine-transformed IND and FDK, log-transformed DON, and linear mixed model analysis of covariance for log-transformed DON with FDK as covariate.

Table 3.2 Pairwise comparisons of untreated check and DMI reference treatments with QoI treatments from linear mixed analyses of fungicide effects on Fusarium head blight index (Index) and deoxynivalenol for wheat spikes inoculated at anthesis with a spore suspension of Fusarium graminearum under controlled conditions

POINTb SPRAYc Comparisonsa INDc DONc DON_adjc IND DON DON_adjc Diffd Pd Diff P Diff P Diff P Diff P Diff P H_8 vs. Check 9.00 0.180 10.35 0.183 0.12 0.704 -19.75 0.019 5.83 0.488 0.29 0.460 H_10 vs. Check -13.25 0.024 1.19 0.975 0.51 0.109 -34.69 <0.001 -3.52 0.365 0.64 0.154 H_10.5.1 vs. Check -25.36 <0.001 -13.99 <0.001 -0.82 0.025 -56.58 <0.001 -17.85 <0.001 -0.77 0.169 Q_8 vs. Check -3.92 0.503 0.76 0.968 0.13 0.654 -13.50 0.111 9.65 0.847 0.32 0.401 Q _10 vs. Check -8.52 0.170 0.67 0.938 0.30 0.335 -21.75 0.011 1.93 0.917 0.60 0.136 Q _10.5.1 vs. Check -17.58 0.004 -5.29 0.265 0.28 0.387 -31.42 <0.001 -12.41 0.048 0.07 0.865 70 P_8 vs. Check -13.75 0.018 -8.48 0.022 -0.28 0.375 -34.79 <0.001 -4.17 0.331 0.62 0.171 P_10 vs. Check -19.92 0.001 -8.79 0.007 -0.21 0.539 -44.92 <0.001 -14.69 0.001 -0.16 0.749 P_10.5.1 vs. Check -25.34 <0.001 -14.89 <0.001 -1.31 0.001 -53.92 <0.001 -18.11 <0.001 -1.05 0.056 H_8 vs. P_10.5.1 34.34 <0.001 25.24 0.001 1.43 0.001 34.17 0.065 23.94 0.777 1.34 0.004 H_10 vs. P_10.5.1 12.09 0.008 16.08 <0.001 1.83 <0.001 19.23 0.001 14.59 <0.001 1.69 <0.001 H_10.5.1 vs. P_10.5.1 -0.02 0.913 0.90 0.152 0.49 0.104 -2.67 0.428 0.26 0.636 0.28 0.443 Q_8 vs. P_10.5.1 21.43 <0.001 15.65 <0.001 1.45 <0.001 40.42 <0.001 27.76 <0.001 1.37 0.005 Q_10 vs. P_10.5.1 16.82 0.001 15.56 <0.001 1.61 <0.001 32.17 <0.001 20.04 <0.001 1.65 <0.001 Q_10.5 vs. P_10.5.1 7.76 0.043 9.60 <0.001 1.60 <0.001 22.50 <0.001 5.70 <0.001 1.12 0.007 P_8 vs. P_10.51 11.59 0.011 6.41 <0.001 1.03 0.002 19.13 0.001 13.94 <0.001 0.89 0.019 P_10 vs. P_10.5.1 5.43 0.147 6.10 <0.001 1.11 0.001 9.00 0.098 3.43 0.009 1.67 <0.001 Check Mean 32.67 … 15.69 … 2.26 … 59.25 … 18.12 … 1.74 … P_10.5.1 Mean 7.33 … 0.8 … 0.94 … 5.33 … 0.01 … 0.69 …

Continued

Table 3.2: Continued

aFungicide treatment and application times - 23.6% pyraclostrobin (438 ml/ha) at flag leaf emergence (H_8), boot (H_10), and anthesis (H_10.5.1); 7% azoxystrobin (658 ml/ha) at flag leaf emergence (Q_8), boot (Q_10) and anthesis (Q_10.5.1); 19% tebuconazole + 19% prothioconazole at flag leaf emergence (P_8), boot (P_10), and anthesis (P_10.5.1). Check and P_10.5.1 = untreated and treated references, respectively. bPOINT = point inoculated wheat spikes and SPRAY = spray-inoculated wheat spikes cIND = Fusarium head blight index (mean proportion of diseased spikelets per spike), DON = deoxynivalenol content of harvested grain (ppm), and DON_adj = log-transformed DON adjusted for the effect FDK as the covariate. dDiff = difference between pairs of means for IND, DON, and log-transformed DON and P = probability (level of significant) based on mixed model analyses of variance for arcsine-transformed IND and log-transformed DON, and linear mixed model analyses of covariance for log-transformed DON with IND as covariate.

Table 3.3 Regression coefficients from linear mixed models analyses of relationship between Fusarium head blight index (IND) or Fusarium damaged kernels (FDK) as continuous covariates and log-transformed deoxynivalenol content of harvest wheat grain as fungicide treatments

Field Point Spray Interceptc se Intercept se Intercept se Factor H_8 0.9 0.12 1.59 0.379 1.19 0.504 H_10 0.88 0.133 1.98 0.256 1.55 0.45 H_10.5.1 1.01 0.122 0.65 0.217 0.13 0.413 Q_8 0.9 0.118 1.6 0.302 1.22 0.532

72 Q _10 1.15 0.129 1.76 0.278 1.51 0.496 Q _10.5.1 0.96 0.122 1.75 0.239 0.98 0.461 P_8 0.87 0.113 1.19 0.254 1.52 0.45 P_10 0.84 0.105 1.26 0.231 0.75 0.426 P_10.5.1 0.45 0.093 0.15 0.217 -0.15 0.415 Check 0.72 0.124 1.47 0.324 0.9 0.599 Slopec 0.023 0.005 0.038 0.008 0.03 0.007

Continued

Table 3.3: Continued

aField experiments conducted in 2011 and 2013 and spray- and point-inoculated greenhouse experiments. Spikes were spray- inoculated with a Fusarium graminearum macroconidia suspension containing 50,000 spores/ml or point-inoculated by pipetting 10 µl of a macroconidia suspension containing 10,000 spores/ml into the central spikelet of each spike. b aFungicide treatment and application times - 23.6% pyraclostrobin (438 ml/ha) at flag leaf emergence (H_8), boot (H_10), and anthesis (H_10.5.1); 7% azoxystrobin (658 ml/ha) at flag leaf emergence (Q_8), boot (Q_10) and anthesis (Q_10.5.1); 19% tebuconazole + 19% prothioconazole at flag leaf emergence (P_8), boot (P_10), and anthesis (P_10.5.1). Check and P_10.5.1 = untreated and treated references, respectively. cIntercepts (logDON when IND = 0) and slopes (the rate of increase in logDON per unit increase in FDK for the field experiment and IND for the greenhouse experiments) were estimated from linear mixed model covariance analysis.

Fig. 3.1 Mean Fusarium head blight index (IND, mean proportion of diseased spikelets per spike) (A, C, and E) and grain (B) and whole-spike (D and F) deoxynivalenol (DON) content of wheat spikes inoculated at anthesis with a spore suspension of Fusarium graminearum and treated with 23.6% pyraclostrobin (438 ml/ha) at flag leaf emergence (H_8), boot (H_10), and anthesis (H_10.5.1); 7% azoxystrobin (658 ml/ha) at flag leaf emergence (Q_8), boot (Q_10) and anthesis (Q_10.5.1); 19% tebuconazole + 19% prothioconazole at flag leaf emergence (P_8), boot (P_10), and anthesis (P_10.5.1). Check is the untreated control. Experiments were conducted under field (A and B) and controlled-environment conditions (C, D, E and F). Controlled-environment studies were either point- (C and D) or spray- (E and F) inoculated with F. graminearum. Each bar shows the mean of 6 replicates for the controlled-environment experiments and 4 replicates x 2 years for the field experiment, along with their corresponding standard error.

10 5 A B 8 4

6 3

4 2

2 1

0 0 80 35 CD 30 60 25

20 40 15

10 20

Deoxynivalenol (ppm) 5

Fusarium head blight index (%) 0 0 50 30 E F 25 40

20 30 15 20 10

10 5

0 0 8 0 .1 8 0 .1 8 0 .1 k 8 0 .1 8 0 .1 8 0 .1 k H_ _1 .5 Q_ _1 .5 P_ _1 .5 ec H_ _1 .5 Q_ _1 .5 P_ _1 .5 ec H 10 Q 10 P 10 Ch H 10 Q 10 P 10 Ch H_ Q_ P_ H_ Q_ P_

Figure 3.1

Fig. 3.2 Predicted relationship based on linear mixed model covariance analysis between Fusarium damaged kernels and log-transformed deoxynivalenol (DON) content of wheat grain (A) and between Fusarium head blight index (IND, mean proportion of diseased spikelets per spike) and log-transformed DON content of spike (B and C) inoculated at anthesis with a spore suspension of Fusarium graminearum and treated with 23.6% pyraclostrobin (438 ml/ha) at flag leaf emergence (H_8), boot (H_10), and anthesis (H_10.5.1); 7% azoxystrobin (658 ml/ha) at flag leaf emergence (Q_8), boot (Q_10) and anthesis (Q_10.5.1); 19% tebuconazole + 19% prothioconazole at flag leaf emergence (P_8), boot (P_10), and anthesis (P_10.5.1). Check is the untreated control. Experiments were conducted under field (A) and controlled-environment (B and C) conditions. Controlled-environment studies were either point- or spray- inoculated with F. graminearum.

1.8 1.6 A 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 5 10 15 20 Fusarium damaged kernel (%) 5 H_8 H_10 B 4 H_10.5.1 Q_8 Q_10 3

2 Q_10.5.1 P_8 1 P_10 P_10.5.1 0 Check Deoxynivalenol (log[DON+1]) Deoxynivalenol 0 10203040506070 Fusarium head blight index (%) 3.0 C 2.5

2.0

1.5

1.0

0.5

0.0

0 1020304050 Fusarium head blight Index (%) Figure 3.2 77

References

1. Adams, G.C., & Hart, L.P. (1989). The role of deoxynivalenol and 15- acetyldeoxynivalenol in pathogenesis of Gibberella zeae, as elucidated through protoplast fusions between toxigenic and non-toxigenic strains. Phytopathology, 79:404-408.

2. Andersen, K. F. 2013. Influence of rainfall patterns on the development of fusarium head blight, accumulation of deoxynivalenol and fungicide efficacy. M.S. thesis, Department of Plant Pathology, The Ohio State University, Columbus.

3. Anonymous. 1991. Conservation in the 1990 farm bill U.S. Dept. Agric. FACTA2.

4. Bai, G., & Shaner, G. (1994), Scab of Wheat: Prospects for control. Plant Disease, 78(8), 760-766.

5. Bai, G., & Shaner, G. (2004), Management on resistance in wheat and barley to Fusarium head blight. Annu. Rev. Phytopathology 42:135-161.

6. Bai, G., and Shaner, G. (1994). Scab of Wheat: Prospects for control. Plant Dis. 78: 760-766.

7. Bai, G., Desjardins, A. E., & Plattner, R. D. (2001). Deoxynivalenol-nonproducing fusarium graminearum causes initial infection, but does not cause Disease Spread in wheat spikes. Mycopathologia, 153(2), 91-98.

8. Bartlett, D. (2002), Review: The Strobilurin Fungicides. Pest Management Science, 58:649-662.

9. Bateman, G.L., Gutteridge, R.J., Gherbawy, Y., Thomsett, M.A., and Nicholson, P. (2007), Infection of stem bases and grains of winter wheat by F. graminearum and effects of tillage method and maize-stalk residues. Plant Pathology 56:604-615.

10. Becher, R. (2009), Adaptation of Fusarium graminearium to tebuconazole yielded descendants diverging for levels of fitness, fungicide resistance, virulence, and mycotoxin production. Phytopathology, 100:444-453. 78

11. Beyer, M., Klix, M., Klink, H., and Verrett, J.-A. 2006. Quantifying the effects of previous crop, tillage, cultivar and triazole fungicides on the deoxynivalenol content of wheat grain-a review. J. Plant Dis. Prot. 113: 241-246.

12. Beyer, M., Klix, M., Klink, H., Verrett, J.-A. (2005), Quantifying the effects of previous crop, tillage, cultivar and triazole fungicides on the deoxynivalenol content of wheat grain-a review. Journal of Plant Diseases and Protection, 113(6), 241-246.

13. Blandino, M., Pilati, A., & Reyneri, A. (2009). Effect of foliar treatments to durum wheat on flag senescence, grain yield, quality and deoxynivalenol contamination in North Italy. Field Crop Research, 114:214-222.

14. Boenisch, M.J., and Schäfer, W. (2011). Fusarium graminearum forms mycotoxin producing infection structures on wheat. BMC Plant Biol. 11:110.

15. Bullerman, L.B., and Bianchini, A. (2007), Stability of mycotoxins during food processing. Int. Journal of Food Microbiology, 119:140-146.

16. Burlakoti, R., Estrada, R., Rivera, V., Boddeda, A., Secor, G., & Adhikari, T. (2007). Real-time PCR quantification and mycotoxin production of Fusarium graminearum in wheat inoculated with isolates collected from potato, sugar beet, and wheat. Phytopathology, 97(7), 835-841.

17. Carter, J.P. (2002), Variation in the pathogenicity associated with the genetic diversity of Fusarium graminearum. European Journal of Plant Pathology, 108:573- 583.

18. D’Mello, F., MacDonald, A., Postel, D., Dijksma, W., Dujardin, A., and Placinta, C. (1998) Pesticide use and mycotoxin production in Fusarium and Aspergillus phytopathogens. European Jorunal of Plant Pathogens, 104:741-751.

19. Dahmen, H., and Straub, T. 1992. Protective, curative, and eradicant activity of difenoconazole against Venturia inaequalis, Cercospora arachidicola, and Alternaria solani. Plant Dis. 76:774-777.

20. De Wolf, E., Madden, L., and Lipps, P. (2003), Risk assessment models for wheat Fusarium head blight epidemics based on within season weather data. Phytopathology, 93:428-435.

21. Del Ponte E. M., Fernandes J. M. C., and Bergstrom G. C. 2007. Influence of growth stage on Fusarium head blight and deoxynivalenol production in wheat. J. Phytopathol. 155:577-581.

22. Desjardins, A. E., Hohn, T. M., & McCormick, S. P. (1993). Trichothecene biosynthesis in fusarium species: Chemistry, genetics, and significance. Microbiological Reviews, 57(3), 595-604.

23. Dexter, J.E., Clear, R.M., and Preston, K.R. (1996), Fusarium head blight: Effect on the milling and baking of some Canadian wheats. Cereal Chemistry, 73:695-701.

24. Diekman, M.A., and Green, M.L. 1992. Mycotoxins and reproduction in domestic livestock. J. Anim. Sci. 70:1615-1627.

25. Dill-Macky, R., & Jones, R. (2000). The effect of previous crop residues and tillage on fusarium head blight of wheat. Plant Disease, 84(1), 71-76.

26. Dufault, N. S., De Wolf, E. D., Lipps, P. E., and Madden, L. V. 2006. Role of temperature and moisture in the production and maturation of Gibberella zeae perithecia. Plant Dis. 90:637- 644.

27. Edwards, S., Pirgozliev, S., Hare, M., & Jenkinson, P. (2001). Quantification of trichothecene-producing fusarium species in harvested grain by competitive PCR to determine efficacies of fungicides against fusarium head blight of winter wheat. Applied and Environmental Microbiology, 67(4), 1575-1580.

28. Edwards, S.G., and Godley, N.P. 2010. Reduction of Fusarium head blight and deoxynivalenol in wheat with early fungicide applications of prothioconazole. Food Addit. Contam. Part A 27:629-635.

29. Ellis, M. L., Broders, K. D., Paul, P. A., and Dorrance, A. E. 2011. Infection of soybean seed by Fusarium graminearum and effect of seed treatments on disease under controlled conditions. Plant Dis. 95:401-407.

30. Fernandez, M.R., Fernandes, J.M., and Sutton, J.C. (1993). Effects of fallow and of summer and winter crops on survival of wheat pathogens in crop residues. Plant Dis. 77:698-703.

31. Fernando, W.G., Paulitz, T.C., Seaman, W.L., Dutilleul, P., and Miller, J.D. (1997). Head blight gradients caused by Gibberela zeae from area sources of inoculum in wheat field plots. Phytopathology, 84(4):414-421.

32. Foround, N.A. (2012). Greenhouse studies reveal increased aggressiveness of emergent Canadian Fusarium graminearum chemotypes in wheat. Plant Disease, 96:1271-1279.

33. Fry, W.E. 1982. Principals of Plant Disease Management. Academic Press, New York. 80

34. Gale, L. R., Ward, T. J., Balmas V., and Kistler H. C. 2007. Population subdivision of Fusarium graminearum sensu stricto in the upper Midwestern United States. Phytopathology 97:1434-1439.

35. Gilbert, J., and Tekauz, A. 2000. Review: Recent developments in research on Fusarium head blight of wheat in Canada. Can. J. Plant Pathol. 22:1-8.

36. Gilbert, J., Fernando, W.G.D., 2004. Epidemiology and biological control of Gibberella zeae Fusarium graminearum. Can. J. Plant Pathol. 26:464–472.

37. Godoy, C.V. and Canteri, M.G. 2004. Efeitos protetor, curativo e erradicante de fungicidas no controle da ferrugem da soja causada por Phakopsora pachyrhizi, em casa de vegetação. Fitopatologia Brasileira 29:97-101. 38. Goswami, R., Kistler, C. (2004), Headed for disaster: Fusarium graminearum on cereal crops. Molecular plant pathology, 5(6), 515-525.

39. Haidukowski, M., Pascale, M., Perrone, G., Pancaldi, D., Campagna, C., and Visconti, A. 2005. Effect of fungicides on the development of Fusarium head blight, yield and deoxynivalenol accumulation in wheat inoculated under field conditions with Fusarium graminearum and Fusarium culmorum. J. Sci. Food Agr. 85:191.

40. Hallen-Adams, H. E., Wenner, N., Kuldau, G. A., and Trail, F. 2011. Deoxynivalenol biosynthesis-related gene expression during wheat kernel colonization by Fusarium graminearum. Phytopathology, 101(10),1091-1096.

41. Hart L. P., Pestka J. J., and Liu M. T. 1984. Effect of kernel development and wet periods on production of deoxynivalenol in wheat infected with Gibberella zeae. Phytopathology 74:1415-1418.

42. Hobbelen, P.H.F., Paveley, N.D., and Vane den Bosch, F. (2011), Delaying selection for fungicide sensitivity by mixing fungicides at a low and high risk of resistance development: A modeling analysis. Phytopathology 101:1224-1233.

43. Hogg, A.C. (2010) Monitoring Fusarium crown rot populations in spring wheat residues using quantitative real-time polymerase chain reaction. Phytopathology, 100:49-57.

44. Hollingsworth, C. R., Mottenberg, C. D., Wiersma, J. V. & Atkinson, L. M. (2008). Agronomic and economic responses of spring wheat to management of Fusarium head blight. Plant Disease, 92(9):1339-1348.

45. Hooker, D.C., H. Spieser and A.W. Schaafsma. 2004. “Effective Application of Fungicides on Wheat Heads: What’s the Best?” Page 330. U.S. Wheat and Barley Scab Initiative, Michigan State Univ., East Lansing, Mich.

46. Ichinoe, M., Kurata, H., Sugiura, Y., and Ueno, Y. (1983). Chemotaxonomy of Gibberella zeae with special reference to production of Trichothecenes and Zearalenone. Applied and Environmental Mircobiology, 46(6), 1364-1369.

47. Ivic, D. 2010. Curative and eradicative effects of fungicides. In Carisse, O. (ed.), Fungicides (pp. 3-22). InTech. DOI:10.5772/555.

48. Jia, H., Millet, B., Cho, S., Bilgic, H., Xu, W., Smith, K., and Muehlbauer, G. (2011) Quantitative trait loci conferring resistance to fusarium head blight in barley respond differentially to Fusarium graminearum infection. Functional Integrated Genomics, 11:95-102.

49. Kim, Y.S., Dixon, E.W., Vincelli, P., and Farman, M.L. 2003. Field resistance to strobilurin (Q0I) fungicides in Pyricularia grisea caused by mutations in the mitochondrial cytochrome b gene. Phytopathology, 93:891-900.

50. Kimura, M., Tokai, T., Takahashi-Ando, N., Ohsato, S., & Fujimura, M. (2007). Molecular and genetic studies of fusarium trichothecene biosynthesis: Pathways, genes, and evolution. Bioscience Biothchnology and Biochemistry, 71(9), 2105-2123.

51. Koch, H.J., Pringas, C., Maerlaender, B. (2006), Evaluation of environmental and management effects on Fusarium head blight infection and deoxynivalenol concentration in the grain of winter wheat. European Journal of Agronomy, 24:357- 366.

52. Lacey J., Bateman G. L., and Mirocha C. J. 1999. Effects of infection time and moisture on development of ear blight and deoxynivalenol production by Fusarium spp. in wheat. Ann. Appl. Biol. 134:277-283.

53. Leslie, J. F., Pearson, C. A. S., Nelson, P. E., and Toussoun, T. A. 1990. Fusarium spp. from corn, sorghum, and soybean fields in the central and eastern United States. Phytopathology 80:343-350.

54. Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfinger, R. D., and Schabenberger, O. 2006. SAS for Mixed Models, second ed. SAS Institute Inc.,Cary, NC. 55. Lori, G.A., Sisterna, S.J., Rizzo, I., Chidichimo, H. (2009), Fusarium head blight in wheat: Impact of tillage and other agronomic practices under natural infection. Crop Protection, 28:495-502.

56. Magan, N., Hope, R., Colleate, A. and Baxter, E. (2002), Relationship between growth and mycotoxin production by Fusarium species, biocides, and the environment. European Journal of Plant Pathology, 108:685-690.

57. Martin, R.A. & Johnston, R.A. (1982), Effects and control of fusarium diseases of cereal grains in the Atlantic Provinces. Canadian Journal of Plant Pathology, 4(2), 210-216.

58. McMullen, M., Jones, R., Gallenberg, D. (1997), Scab of wheat and barley: A re- emerging disease of devastating impact. Plant Disease, 81(12), 1340-1348.

59. McMullen, M., S. Halley, B. Schatz, S. Meyer, J. Jordahl, and J. Ransom. 2008. Integrated strategies for Fusarium head blight management in the United States. Cereal Research Communications 36:563-568.

60. Mesterhazy, A. 1995. Types and components of resistance to Fusarium head blight of wheat. Plant Breed. 114:377–386.

61. Miller, J. & Greenhalgh, R. (1985), Nutrient effects on the Biosynthesis of Trichothecenes and Other Metabolites by Fusarium graminearum. Mycologia,77(1), 130-136.

62. Miller, J. & Young, C. (1985). Deoxynivalenol In An Experimental Fusarium Graminearum Infection of Wheat. Canadian Journal of Plant Pathology, 7(2), 132- 134.

63. Milus, E.A., & Parsons, C.E. (1994), Evaluations of foliar fungicides for controlling Fusarium head blight of wheat. Plant Disease, 78:697-699.

64. Mueller, D. S., and Bradley, C. A. 2008. Field crop fungicide for the north central United States. North Central Integrated Pest management Center. 65. Osborne, L.E., Stein, J.M. (2007), Epidemiology of Fusarium head blight on small- grain cereals. International journal of food microbiology, 119:103-108.

66. Parnell, S., Van den Bosch, and Gilligan, C.A. (2006), Large scale fungicide spray heterogeneity and regional spread of resistant pathogen strains. Phytopathology, 96(5) 549-555.

67. Parry, D.W., Jenkinson, P., and McLeod, L. 1995. Fusarium ear blight (scab) in small grain cereals- a review. Plant Pathol. 44:207-238.

68. Paul, P. A., El-Allaf, S. M., Lipps, P. E., & Madden, L. V. (2004). Rain splash dispersal of gibberella zeae within wheat canopies in ohio. Phytopathology, 94(12), 1342-1349. 83

69. Paul, P. A., El-Allaf, S. M., Lipps, P. E., and Madden, L. V. 2005. Relationships between incidence and severity of Fusarium head blight on winter wheat in Ohio. Phytopathology 95:1049-1060.

70. Paul, P., Lipps, P., & Madden, L. 2005. Relationship between Fusarium head blight intensity and deoxynivalenol accumulation in harvested wheat grain. Phytopathology 95:1225-1236.

71. Paul, P., Lipps, P., and Madden, L. 2006. Meta-analysis of regression coefficients for the relationship between fusarium head blight and deoxynivalenol content of wheat. Phytopathology 96:951-961.

72. Paul, P., Lipps, P., Hershman, D., McMullen, M., Draper, M., & Madden, L. (2007). A quantitative review of tebuconazole effect on fusarium head blight and deoxynivalenol content in wheat. Phytopathology, 97(2), 211-220.

73. Paul, P., Lipps, P., Hershman, D., McMullen, M., Draper, M., and Madden, L. 2008. Efficacy of triazole-based fungicides for Fusarium head blight and deoxynivalenol control in wheat: A multivariate meta-analysis. Phytopathology 98:999-1011.

74. Paul, P.A., McMullen, M.P., Hershmann, D.E., and Madden, L.V. (2010), Meta- analysis of the effects of triazole-based fungicides on wheat yield and test weight as influenced by Fusarium head blight intensity. Phytopathology, 100:160-171.

75. Peraldi, A., Beccari, G., Steed, A., and Nicholson, P. (2011). Brachypodium distachyon: A new pathosystem to study Fusarium head blight and other Fusarium diseases of wheat. BMC Plant Biology 11:100.

76. Pirgozliev, S.R., Edwards, S.G., Hare, M.C., and Jenkinson, P. 2002. Effect of dose rate of azoxystrobin and metaconazole on the development of Fusarium head blight and the accumulation of deoxynivalenol (DON) in wheat grain. European Journal of Plant Pathology 108:469-478.

77. Pradini, A., Sigolo, S., Battilani, P., & Piva, G. (2008), Review of predictive models for Fusarium head blight and related mycotoxin contamination in wheat. Food and Chemical Toxicology, 47:927-931.

78. Pritsch, C., Muehlbauer, G., Bushnell, W., Somers, D., and Vance, C. (2000). Fungal development and induction of defense response genes during early infection of wheat spikes by Fusarium graminearum. Molecular Plant Microbe Interactions, 13(2), 159- 169.

79. Puri, K. and Zhong, S. (2010). The 3ADON Population of Fusarium graminearum Found in North Dakota Is More Aggressive and Produces a Higher Level of DON than the Prevalent 15ADON Population in Spring Wheat. Phytopathology, 100:1007-1014.

80. Rocha, O., Ansari, K., and Doohan, F.M. 2005. Effect of trichothecene mycotoxins on eukaryotic cells: A review. Food Additives and Contaminants, 22(4):369-378.

81. Rotter, B.A., Prelusky, D.B., Pestka, and J.J. 1996. Toxicology of deoxynivalenol (vomitoxin). J. Toxicol. Environ. Health 48:1-34.

82. Rudd, J.C., Horsley, R.D., McKendry, A.L., and Elias, E.M. (2001), Host plant resistance genes for Fusarium head blight: Sources, mechanisms, and utility in conventional breeding systems. Crop Sci. 41:620-627. 83. Ruske, R.E., Gooding, M.J., Jones, S.A. (2003) The effects of adding picoxystrobin, azoxystrobin, and nitrogen to a triazole programme on disease control, flag leaf senescence, yield and grain quality of winter wheat. Crop Protection, 22:975-987.

84. Schroeder, H.W., & Christensen, J.J. 1963. Factors affecting resistance of wheat to scab caused by Gibberella zeae. Phytopathology, 53:831-838.

85. Shah, D.A., Molineros, J.E., Paul, P.A., Willyerd, K.T., Madden, L.V., and DeWolf, E.D. 2013. Predicting Fusarium head blight epidemics with weather-driven pre- and post-anthesis logistic regression models. Phytopathology 103:906-1019.

86. Simpson, D., Weston, G., Turner, J., Jennings, P., & Nicholson, P. (2001). Differential control of head blight pathogens of wheat by fungicides and consequences for mycotoxin contamination of grain. European Journal of Plant Pathology, 107(4), 421- 431.

87. Sneller, C., Paul, P., Guttiere, M. (2010), Characterization of resistance to Fusarium head blight in an eastern U.S. soft red winter wheat population. Crop Science 50:123- 133.

88. Snijders, C. H. A., & Krechting, C. (1992). Inhibition of deoxynivalenol translocation and fungal colonization in Fusarium head blight resistant wheat. Canadian Journal of Botany, 70(8), 1570-1576.

89. Snijders, C.H.A. and Van Eeuwijk, F.A. (1991), Genotype x strain interactions for resistance to Fusarium head blight caused by Fusarium culmorum in winter wheat. Theoretical and applied genetics 81:239-244.

90. Stack, R., & McMullen, M. (1998). A visual scale to estimate the severity of Fusarium head blight. Bull. PP-1095. North Dakota State Univ. Ext. Ser.,Fargo, ND

91. Summerell, B.A., Burgess, L.W., Klein, T.A., and Pattison, A.B. (1990), Stubble management and the site of penetration of wheat by Fusarium graminearum group 1. Phytopathology, 80:877-879. 92. Sutton, J. C. (1982). Epidemiology of wheat head blight and maize ear rot caused by Fusarium graminearum. Cannadian Journal of Plant Pathologh- Revue Canadienne De Phytopathologie, 4, 195-209.

93. Talas, F., Kalih, R., and Miedaner, T. (2012), Within-Field Variation of Fusarium graminearum Isolates for Aggressiveness and Deoxynivalenol Production in Wheat Head Blight. Phytopathology, 102:128-134.

94. Tanaka, T., Hasegawa, A., Yamamoto, S., Lee, U.S., Sugiura, Y., and Ueno, Y. (1988), Worldwide Contamination of Cereals by the Fusarium Mycotoxins Nivalenol, Deoxynivalenol, and Zearalenone. Journal of Agricultural and Food Chemistry 36:979-983.

95. Trail, F. (2005), Using genomics to understand the disease cycle of the Fusarium head blight fungus, Gibberella zeae (anamorph Fusarium granimearum). Canadian Journal of Plant Pathology, 27:486-498.

96. Trail, F., & Common, R. 2000. Perithicium development in Gibberella zeae, a light microscopy study. Mycologia, 92:130-138.

97. Walter, S., Nicholson, P., & Doohan, F.M. (2010), Research Review - Action and reaction of host and pathogen during Fusarium head blight disease. New Phytologist, 185:54-66.

98. Ward TJ, Clear RM, Rooney AP, O’Donnell K, Gaba D, Patrick S, Starkey DE, Gilbert J, Geiser DM, Nowicki TW (2008) An adaptive evolutionary shift in Fusarium head blight pathogen populations is driving the rapid spread of more toxigenic Fusarium graminearum in North America. Fungal Genet Biol 45:473–484.

99. Wegulo, S.N., Bockus, W.W., Hernandez Nopsa, J., De Wolf, E.D., Eskridge, K.M., Peiris, K.H.S., and Dowell, F.E. 2011. Effects of integrating cultivar resistance and fungicide application on Fusarium head blight and deoxynivalenol in winter wheat. Plant Dis. 95:554-560. 100. Weisz, R. (2011), Multiple mid-atlantic field experiments show no economic benefit to fungicide application when fungal disease is absent in winter wheat. Phytopathology, 101(3), 323-333.

101. Wiersma, J. J., and Motteberg, C. D. 2005. Evaluation of five fungicide application timings for control of leaf-spot diseases and Fusarium head blight in hard red spring wheat. Can. J. Plant Pathol. 27:25-37.

102. Willyerd, K.T., Li, C., and Madden, L.V. (2012), Efficacy and stability of integrating fungicide and cultivar resistance to manage Fusarium head blight and deoxynivalenol in wheat. Plant Disease 96:957-967.

103. Windels, C. (1999) Economic and social impacts of Fusarium head blight: changing farms and rural communities in the northern great plains.The American phytopathological society 91st annunal meeting presidential address. 90(1).

104. Windels, C. 2000. Economic and social impacts of Fusarium head blight: changing farms and rural communities in the Northern Great Plains. Phytopathology 90:17-21.

105. Yagen, B., & Joffe (1976), Scrrening of toxic isolates of Fusarium poae and Fusarium sporotrichioides involved in causing alimentary toxic aleukia. Applied and Environmental Microbiology, 32(3), 423-427.

106. Yoshida M., Kawada N., and Nakajima T. 2007. Effect of infection timing on Fusarium head blight and mycotoxin accumulation in open- and closed-flowering barley. Phytopathology 97:1054-1062.

107. Yoshida, M., Nakajima, T., Arai, M., Suzuki, F., & Tomimura, K. (2008). Effect of the timing of fungicide application on fusarium head blight and mycotoxin accumulation in closed-flowering barley. Plant Disease, 92(8), 1164-1170.

108. Yoshida, M., Nakajima, T., Tomimura, K., Suzuki, F., Arai, M., & Miyasaka, A. 2012. Effect of the timing of fungicide application on Fusarium head blight and mycotoxin accumulation in wheat. Plant Dis. 96:845-851. 109. Ypema, H. and Gold, R. (1999), Kresoxim-Methyl: Modification of a Naturally Occurring Compound to Produce a New Fungicide. Plant Disease, 83 (1), 4-18.

110. Yuen, G., & Schoneweis, S. (2007), Strategies for managing Fusarium head blight and deoxynivalenol accumulation in wheat. International Journal of Food Microbiology, 119:126-130.

111. Zhang, Y., Fan, P., Zhang, X., Chen, C., and Zhou, M. 2009. Quantification of Fusarium graminearum in harvested grain by real-time polymerase chain reaction to assess efficacies of fungicides on Fusarium head blight, deoxynivalenol contamination, and yield of winter wheat. Phytopathology 99:95-100. .