Crop Protection 77 (2015) 139e147
Contents lists available at ScienceDirect
Crop Protection journal homepage: www.elsevier.com/locate/cropro
Impact of Fusarium graminearum inoculum availability and fungicide application timing on Fusarium head blight in wheat Anna N. Freije, Kiersten A. Wise* Purdue University, 915 West State Street, West Lafayette, IN 47907, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 April 2015 Received in revised form 15 July 2015 Accepted 16 July 2015 Available online xxx
Fusarium head blight (FHB) of wheat (caused by Fusarium graminearum Schwabe (telemorph Gibberella zeae (Schwein.) Petch) is considered one of the most economically important diseases on wheat in the United States. Currently, farmers rely heavily on fungicides applied at early anthesis or Feekes Growth Stage (FGS) 10.5.1 to protect their crop from this disease. Field trials were conducted at the Agronomy Center for Research and Education in West Lafayette, IN during the 2012e2013 and 2013e2014 growing seasons to determine the impact of post-anthesis fungicide timing in conjunction with initial infection by F. graminearum on development of FHB and deoxynivalenol (DON) in soft red winter wheat. To achieve this, each experimental plot within a treatment was inoculated and received a fungicide application on the same day beginning at FGS 10.5.1 (anthesis), and continuing each day for anthesis þ1, 3, 5, 7, 9, and 11 days. The fungicide prothioconazole þ tebuconazole was applied at 475 mL/ha, and experimental plots were inoculated with macroconidia of F. graminearum on the same day as each fungicide application. Results indicate that fungicide applications made up to 11 days post-anthesis may be useful in reducing FHB and DON in wheat when inoculum becomes present near the time of application. © 2015 Published by Elsevier Ltd.
Keywords: Fusarium head blight Fungicide Application timing DON Triticum aestivum Fusarium graminearum
1. Introduction Fusarium graminearum Schwabe [telemorph Giberella zea (Schweinitz) Petch] is the primary causal agent of Fusarium head blight (FHB) of wheat (Triticum aestivum L. ssp. aestivum) in the United States (Goswami and Kistler, 2004). This fungus infects wheat heads during anthesis, causing salmon to white colored “tombstone” kernels to form in lieu of healthy grain (Sutton, 1982). The fungus also produces several mycotoxins, including deoxynivalenol (DON), which is known to inhibit protein synthesis in eukaryotes, making it harmful to humans and other mammals (O'Donnell et al., 2000). Although FHB began as a sporadic problem in the U.S. since the mid 1920's, it re-emerged as a disease of great economic importance after a series of epidemics in the mid 1990's (McMullen et al., 1997, 2012). The pathogen is also able to infect corn (Zea mays L.), another important crop in the Midwestern U.S., causing the disease Gibberella ear rot (Sutton, 1982). Currently, FHB is considered the disease of greatest concern to wheat cultivation in the U.S. (Bockus et al., 2010).
* Corresponding author. Dept. of Botany and Plant Pathology, Purdue University, 915 West State Street, West Lafayette, IN 47907, USA. E-mail address: [email protected]
(K.A. Wise). http://dx.doi.org/10.1016/j.cropro.2015.07.016 0261-2194/© 2015 Published by Elsevier Ltd.
No single management practice will completely suppress FHB. Typical integrated pest management (IPM) strategies for FHB include planting wheat after soybean instead of after corn, using moderately resistant wheat cultivars, and applying fungicide at zy, 1995; Willyerd et al., 2011). The beginning anthesis (Mesterha most effective fungicides currently labeled for use against F. graminearum on wheat are the demethylase inhibitor (DMI) triazoles, prothioconazole and tebuconazole, trade name Prosaro (Bayer CropScience LP, Research Triangle Park, NC), and metconazole, trade name Caramba (BASF, Research Triangle Park, NC; FRAC, 2011; Paul et al., 2008; Wise, 2014). All of the active ingredients in these fungicides are sterol biosynthesis inhibitors (SBI) and are members of fungicide resistance action committee (FRAC) group G1: SBI class 1: DMI fungicides (FRAC code 3; FRAC, 2011). Both products are broad-spectrum fungicides and are also used to control foliar diseases in wheat (Wise, 2014). The DMI triazole fungicides inhibit sterol biosynthesis in fungal membranes by inhibiting the action of enzyme C14-demethylase on C-14 in lanosterol. This is a necessary step in the biosynthesis of ergosterol, and its inhibition leads to the buildup of fatty acids and ergosterol precursors in the fungal cells, resulting in abnormal growth patterns and inhibition of growth (Kӧ;ller, 1992; Schnabel and Jones, 2001; Siegel, 1981). The DMI triazole fungicides are also partially systemic, meaning they can
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
penetrate the plant and move within its tissue, but they are unable to enter the xylem (Mueller and Bradley, 2008; Siegel, 1981). Proper application timing and techniques are essential for optimum fungicide efﬁcacy. Current recommendations state that fungicide should be applied at early anthesis, or Feekes Growth Stage (FGS) 10.5.1, the time at which 50% of the primary tillers in a ﬁeld have 50% of their anthers extruding (Large, 1954). In winter wheat there are several limitations to meeting this optimal application timing. Winter wheat may produce many tillers and therefore beginning anthesis on different heads may be spread over several days. Because of this, the ﬂowering period of a single plant can extend up to two weeks. Spraying fungicide at beginning anthesis of the primary tiller will not coincide with beginning anthesis for the secondary tillers. Weather has a great inﬂuence on spore production and infection by F. graminearum. When ﬂowering of individual heads within a ﬁeld is staggered over a week or more, some heads will be more vulnerable to infection than others. Likewise, if fungicide timing is truly critical, the variability of ﬂowering wheat heads within a ﬁeld will impact the efﬁcacy of a fungicide application. Rain can also pose an obstacle to spraying at precisely FGS 10.5.1 due to the inability of spray equipment to enter a wet ﬁeld. Several studies have also determined that the optimum application timing for FHB suppression and DON reduction may be different (Yoshida et al., 2012; Yoshida and Nakajima, 2010). Finally, it is important that fungicide applications do not violate the fungicide pre-harvest interval, the amount of time that must pass between the ﬁnal fungicide application and harvest. If fungicide is applied after the optimum timing and weather conditions favor a prompt harvest, it is possible that the 30-day pre-harvest interval for commonly applied fungicides will not be met. Several studies have demonstrated that fungicide applications can reduce FHB and DON levels when applied up to 6 days past FGS 10.5.1 and that DON may be reduced by applications made up to 20 days after anthesis (DAA) (D'Angelo et al., 2014; Hart et al., 1984; Yoshida et al., 2012). However, all of these studies have focused on the effect of post-anthesis fungicide applications when inoculum became available at FGS 10.5.1. Since inoculum availability is primarily inﬂuenced by environmental conditions, it is likely that infection does not always occur precisely at FGS 10.5.1. Also, because anthesis can last approximately 2 weeks, it is important to determine the efﬁcacy of fungicide use during this entire period of host susceptibility. Therefore, the objective of this study was to determine the impact of fungicide timing, in conjunction with initial infection by F. graminearum, on FHB and DON up to 11 days post-anthesis in soft red winter wheat. 2. Materials and methods Field studies were conducted in two ﬁeld seasons, 2012e2013 and 2013e2014, at Purdue's Agronomy Center for Research and Education (ACRE) in West Lafayette, Indiana.
the experiment. Factorial level one refers to the presence or absence of fungicide given an inoculation with F. graminearum. Level two refers to the timing, or day that fungicide and inoculum were applied relative to the beginning of anthesis (FGS 10.5.1). Anthesis was deﬁned as the ﬁrst day that 50% of the primary tillers across the ﬁeld were extruding 50% of their anthers. The application (both fungicide and inoculum) occurring at anthesis was given a designation of day 0. Applications occurring after anthesis were designated as the number of days after anthesis (DAA), with six applications occurring at 1, 3, 5, 7, 9, or 11 DAA. In 2013, anthesis occurred on May 25. Each plot was designated as an experimental unit and one of the seven application timings (0e11 DAA) was randomly assigned to each experimental unit within a replicate. Each plot received a maximum of one fungicide application corresponding to the randomly assigned day after anthesis. The inoculated, non-fungicide treated plots served as controls within each application time. Experimental plots were 2.1 m wide and 6.1 m long with a 1.5 m wide alley between each plot. Border plots of the same size were established between experimental plots to prevent the effects of inoculum and/or fungicide drift during treatment applications. Border plots were planted with cultivar INW0803 at a seeding rate of 3.4 106 seeds/ha. 2.2. 2014 Field experiment Plots were established on October 15, 2013 with soft red winter wheat cultivar P25R47 at a seeding rate of 3.4 106 seeds/ha using a Great Plains Drill. The previous crop was corn. The ﬁeld was disked four times prior to planting. Fertilizer, in the form of DAP (at a rate of 100.8 kg/ha) and nitrogen (at a rate of 107.3 kg/ha), was applied on September 2, 2013 and March 28, 2014 respectively. Weeds were controlled by hand prior to anthesis. Due to a harsh winter that led to winter kill of wheat plants, the healthiest 21 plots in each replication, from a total of 30, were selected for use in the experiment prior to treatment randomization. The experimental design in 2014 consisted of a randomized complete block with a 3 7 factorial arrangement and was replicated four times within the experiment. Factorial level one refers to the presence or absence of fungicide given an inoculation with F. graminearum. A naturally infected, non-fungicide treated control was added in 2014, which served as a means of evaluating the baseline level of disease in naturally infected plots alongside inoculated plots. Level two refers to the timing, or day that fungicides and inoculum were applied relative to the beginning of anthesis (FGS 10.5.1). Anthesis was deﬁned as above, and the application (both fungicide and inoculum) occurring at anthesis was given a designation of day 0. Treatments occurred on 1, 3, 5, 8, 9, or 11 days after anthesis (DAA). In 2014, FGS 10.5.1 occurred on May 28. Treatment applications scheduled for 7 DAA were moved to 8 DAA due to rain. Experimental and border plots were established as described for the 2013 experiment (2.1).
2.1. 2013 Field experiment 2.3. Inoculum preparation Plots were established on October 12, 2012 with soft red winter wheat cultivar P25R47, which is moderately susceptible to FHB. Seed was drilled at a seeding rate of 5.0 106 seeds/ha using a Great Plains drill into soil that had been disked and ﬁeld cultivated after a crop of corn. Fertilizer in the form of diammonium phosphate (DAP) was applied at 100.8 kg/ha on September 19, 2012, followed by an application of potash at 336 kg/ha on September 25 and an application of urea at 224 kg/ha on March 21, 2013. Weeds were controlled by hand prior to anthesis. The experimental design was a randomized complete block with a 2 7 factorial arrangement, and was replicated four times within
Macroconidia inoculum of F. graminearum was prepared in the laboratory prior to ﬁeld inoculation. Isolates used had been stored on corn kernels at 80 C storage prior to use. We used a mix of isolates collected in Indiana each year to simulate typical local inoculum pressure. In 2013 the isolates 09INDecaturF3S1, 09INDecaturF1S1, and 10INSWS2U112 were selected for inoculation. In 2014, isolates 09DecaturF3S1, 10INSWS2U112 and 13INHunt600NPH5 were used in the ﬁeld trial. Each isolate was screened for virulence on wheat in a greenhouse prior to use in these trials. Isolates were grown on full strength potato dextrose
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
agar (PDA) amended with ampicillin at 0.05 mg/mL. After approximately 1e2 weeks, a single plug of each isolate was transferred into separate Erlenmeyer ﬂasks containing sterile mung bean broth according to the protocol described by Bai and Shaner (1996), with the following alterations: Beans were added to water (at 95e99 C) and left to soak for 10 min before the broth was divided into 200 mL ﬂasks with 100e150 mL of broth per ﬂask. Flasks were plugged with pieces of cotton wrapped in cheesecloth and covered with aluminum foil before autoclaving. Inoculated ﬂasks of broth were placed on a shaker plate (model 15000-1, VWR Scientiﬁc, Randor, PA) until the concentration of the macroconidia was greater than 50,000 conidia/mL (~3 weeks). We measured the concentration of macroconidia with a hemacytometer. The ﬁnal inoculum suspension was created by combining equal parts (by macroconidia count) of broth from each isolate before being diluted to 50,000 macroconidia/mL with water. Inoculum was kept at 4.4 C until use. 2.4. Fungicide application The DMI triazole fungicide prothioconazole þ tebuconazole (Prosaro; Bayer CropScience, Research Triangle Park, NC) was applied at the recommended rate of 475 mL/ha (Bayer CropScience) to all plots receiving a fungicide application in both years of the experiment. Preference non-ionic surfactant and anti-foaming agent (AgriSolution, LLC) was included in each application at 0.125% v/v to improve fungicide coverage. Fungicide was applied using a hand-held backpack sprayer and spray boom with 4 Teejet 8001 nozzles spaced 48 cm apart and at a 90 angle to the boom. The boom was powered by compressed CO2 set at 276 kPa and was held approximately 25 cm above the plot during application. A total volume of 187 L/ha was applied to each treated plot. Fungicides were applied in the morning on each treatment day. 2.5. Inoculum application In 2013, inoculum was applied to experimental plots with a handheld 1.5 m wide boom ﬁtted with 4 Teejet 8002 nozzles spaced 48 cm apart, powered by compressed CO2. The boom was calibrated to deliver 190 L/ha, at 276 kPa for a total volume of 300 mL/plot. In 2014, the handheld boom was ﬁtted with Teejet 8001 nozzles, with all other factors consistent with 2013 applications. Plots were inoculated in early evening on each treatment day approximately 8 h after fungicide application. 2.6. Disease assessment FHB incidence and index were assessed on June 14 in 2013, and on June 18 in 2014, corresponding to FGS 11.1, the milky kernel stage, 10 days after the last treatment application. Assessments were made on 4 arbitrarily-selected groups of 25 tillers from each plot. FHB incidence was determined by counting the heads with FHB symptoms in each group of tillers (Stack and McMullen, 2011). Disease index, also known as disease severity, as deﬁned by Paul et al. (2005), was determined using a visual rating scale to estimate the percentage of total head area with FHB symptoms across the 25 tillers in each sample (Stack and McMullen, 2011). In 2013 and 2014, foliar diseases such as leaf rust, (Puccinia triticina), Septoria tritici blotch, (Septoria tritici), and Stagonospora leaf blotch (Stagonospora nodorum) were noted in the trials, but were not present at levels to impact results (data not shown). 2.7. Post-harvest assessments The middle 1.75 m of each plot was harvested on July 9, 2013
and July 16, 2014 respectively, with a small plot Kincaid 8XP combine. Harvest was 36 days and 31 days after the last treatment, in 2013 and 2014, respectively, thereby meeting the 30 day preharvest interval speciﬁed by the fungicide manufacturer. Percent kernel moisture, test weight, and yield were obtained for each plot and used to calculate the adjusted yield (kg/ha) at 13.5% moisture (Hellevang, 1995). In addition to yield, an arbitrary sample of approximately 2.3 kg of harvested grain was obtained from each plot at harvest. Postharvest analyses were performed on subsamples of this grain. A Key-mat Model 946 Seed Counter (Key-mat Equipment Company Inc., St. Charles, IL), adjusted for wheat kernel size, was used to obtain samples of 1000 kernels that were subsequently weighed to obtain the 1000 kernel weight. Percent Fusarium damaged kernels (FDK), was visually assessed for each plot using a percentage based visual scale created by Jones and Mirocha (1999). We ground kernels from each plot into a rough powder using a Romer Series II Mill (Romer Labs, Inc., Union, MO) until we had a volume of approximately 4 L of wheat-meal total. A 20 g subsample of this wheat-meal was then used for a deoxynivalenol (DON) assay. The mill was thoroughly vacuumed out between each sample to avoid cross-contamination. DON analysis was performed using a DON3 QuickTox kit (EnviroLogix, Portland, ME) catalog number AQ 204 BG in 2013 and a DON3 QuickTox kit, catalog number AQ 254 BG in 2014. The DON3 kit used in 2013 had a detection limit of 5.0 mg/g and the 2014 kit had a detection limit of 12.0 mg/g. Analyses were performed according to the instructions provided in the kit and DON levels were obtained for each sample unit using the QuickScan (Environlogix, Portland, ME) system. In 2013, if DON levels exceeded 5.0 mg/g, the sample was re-tested using another 20 g subsample. Samples were diluted 2-fold after the extraction step and buffer was added according to kit instructions. The resulting DON value was doubled to attain the ﬁnal measurement. In 2014, no dilutions were required. A subsample of grain from each plot in each year was also sent to the University of Minnesota to validate DON levels using gaschromatography, mass-spectrometry (GC-MS). 2.8. Statistical analysis Due to differences in experimental design by year, trials were analyzed separately. All data analyses were performed using the PROC MIXED procedure of SAS 9.3 (SAS Institute Inc., Cary, NC). Since FHB index is a factor of both FHB incidence and FHB severity and is typically the unit used to quantify FHB, only FHB index is reported. Tests of homogeneity were performed on all variables. FHB index was arcsine-square root transformed to attain homogeneity of variance. A BoxeCox regression analysis on the postharvest data (DON, FDK, and 1000 kernel weight) indicated that a log transform was appropriate to use on the 2013 and 2014 FDK values in order to achieve homogeneity of variance. All dependent variables (FHB index, FDK, 1000 kernel weight, DON and adjusted yield, hereafter referred to as ‘yield’) were subjected to analysis of variance (ANOVA) to test for signiﬁcant differences among fungicide treatments and inoculum timing. The interaction between fungicide treatment level and inoculum timing was also tested. Least squares means (LSM) tests were performed for signiﬁcant variables using PROC MIXED with a Kenward-Roger correction for adjusting degrees of freedom and estimation of random effects. Treatment, application timing, and their interaction were treated as ﬁxed effects and replication was treated as the random effect. A separate residual variance was estimated for each treatment level. A TukeyeKramer adjustment was used in the comparisons of the Least-squares means for all ﬁxed effects. Fixed effects were considered signiﬁcant at P < 0.05. All values are recorded as back transformed mean estimates.
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
Pearson's correlation tests were run on the untransformed values of dependent variables as deemed appropriate based upon the results of the ANOVA. 3. Results The scale of replication (block) effect relative to other random effects was non-signiﬁcant in both years of the trials, and therefore the effect of block is not included in further discussion. Weather in 2013 was more conducive to disease than in 2014 (Figs. 1 and 2). In 2013, temperatures remained in the ideal range for F. graminearum infection (20e25 C) for the majority of anthesis, and there were 6 rain events for a total accumulation of 72.4 mm. Weather in 2014 was warmer and drier than in 2013. Precipitation totaled only 6.6 mm (excluding June 5 where precipitation level information is missing) and over the course of this trial only 3 days had temperatures that fell between 20 and 25 C, with 7 days experiencing temperatures over 30 C. 3.1. Effect of fungicide treatment and inoculum application timing on FHB index In 2013, there was no interaction effect between fungicide treatment and inoculum timing on FHB index (the time F. graminearum inoculum became available to wheat heads) (Table 1). Both fungicide treatment and inoculum timing signiﬁcantly affected FHB index (Table 1). Inoculum applied 9 days after anthesis (DAA) resulted in a lower FHB index level than inoculum applied at 0 DAA (P ¼ 0.034; Fig. 3). In 2014, there was a signiﬁcant interaction between fungicide treatment and inoculum timing for FHB index (Table 1). Therefore, differences in LSMs were evaluated to determine the effect of fungicide treatment at each inoculum timing and inoculum timing within each fungicide treatment. The only effect of fungicide within inoculum timing was found at 0 DAA, where FHB index was lower in naturally infected, non-fungicide treated plots than in inoculated, non-fungicide treated plots (Fig. 3). Within inoculated, nonfungicide treated plots, inoculum applied at 0 DAA resulted in signiﬁcantly higher FHB index than all other inoculum timings with the exception of 3 DAA. Inoculum timing did not affect FHB index in inoculated, fungicide-treated plots or in naturally infected, nonfungicide treated plots. 3.2. Effect of fungicide treatment and inoculum application time on deoxynivalenol The effect of fungicide treatment on DON did not vary with
Fig. 2. Daily maximum temperatures (solid line) and precipitation (bars) during the course of the experiment in 2014. Inoculation timing zero (0) corresponds to May 28, 2014. Weather data were retrieved from the Agronomy Center for Research and Education (ACRE) and Indiana State Climate Ofﬁce, iClimate.org. aOn Application day 8, daily high temperature was retrieved from accuweather.com and precipitation data are missing.
inoculation timing and was itself signiﬁcant (Table 1). Fungicide treatment reduced DON levels in inoculated, fungicide treated plots in 2013 compared to inoculated, non-fungicide treated plots at 3, 7, and 9 DAA. Average DON levels in inoculated, non-fungicide treated plots were 1.46 times higher than DON levels in inoculated, fungicide treated plots in 2013 (Fig. 4). A spike in DON levels occurred at 3 DAA in the inoculated, non-fungicide treated plots but not in the inoculated, fungicide treated plots (Fig. 4). This increase coincided with a heavy rain event within one day of inoculation. There was a signiﬁcant interaction between inoculation timing and fungicide treatment in 2014 (Table 1). DON levels were signiﬁcantly higher in inoculated, non-fungicide treated plots than in naturally infected, non-fungicide treated plots at 0 and 5 DAA. Fungicide reduced DON in inoculated, fungicide treated plots by 48.3% compared to inoculated, non-fungicide treated plots. Average DON levels in non-fungicide treated plots were 1.45 times higher than DON levels in inoculated, non-fungicide treated plots, and 1.25 times higher than in naturally infected, non-fungicide treated plots. Naturally infected, non-fungicide treated plots had average DON levels that were only 1.17 times greater than inoculated, fungicide treated plots. Within inoculated, non-fungicide treated plots, DON levels were signiﬁcantly higher when inoculation was at anthesis compared to 11 DAA. Within inoculated, fungicide-treated plots, those inoculated at 0 DAA developed higher levels of DON compared to those inoculated at 9 and 11 DAA, and plots inoculated at 1 DAA had signiﬁcantly higher DON levels than those inoculated at 11 DAA. Two spikes in DON levels were observed within the inoculated, non-fungicide treated plots, one at anthesis and another at 5 DAA. Both of these spikes corresponded to rain events within 1 day of the plots being inoculated. 3.3. Effect of fungicide treatment and fungicide and inoculation timing on 1000 kernel weight
Fig. 1. Daily maximum temperatures (solid line) and precipitation (bars) during the course of the experiment in 2013. Inoculation time zero (0) corresponds to May 24, 2013. Weather data were retrieved from the Agronomy Center for Research and Education (ACRE) and Indiana State Climate Ofﬁce, iClimate.org.
In 2013, there were no interaction effects between fungicide treatment and inoculum timing on 1000 kernel weight. Fungicide signiﬁcantly increased 1000 kernel weight, but inoculation timing had no effect (Table 1). In 2014, there was a signiﬁcant interaction between inoculation timing and fungicide treatment for 1000 kernel weight (Table 1). However, this interaction is largely explained by the signiﬁcant difference between inoculated, non-fungicide treated plots and both inoculated, fungicide-treated and naturally infected, nonfungicide treated plots at (0 DAA; P ¼ 0.022 and 0.064 respectively). Across all inoculation timings, 1000 kernel weight of
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
Table 1 Two-way analysis of variance from the (A) 2013 and (B) 2014 ﬁeld experiments. Inoculum application time and fungicide treatment are treated as main effects. Type three tests of ﬁxed effects were performed on the raw data, with the exception of FDK, which was log transformed, and FHB index which was arcsine square-root transformed to increase the homogeneity of variance. The degrees of freedom (df) are represented as numerator, denominator. A)
F value P value df
Inoculum timing 6, 34.7 3.26 Fungicide treatment 1, 34.7 5.26 Inoculum timing* 6, 34.7 1.04 Fungicide treatment B)
Inoculum timing Fungicide treatment Inoculum timing* Fungicide treatment
6, 50.4 2.83 2, 36.4 1.77 12, 44 3.04
a b c d e f
0.012f 0.028 0.415
Yieldc F value P value
6, 30.3 0.90 1, 30.3 76.93 6, 30.3 2.15
0.019 0.185 0.003
F value P value
0.510 6, 34.8 1.20 <0.0001 1, 34.8 67.70 0.077 6, 34.8 1.14
DON P value df
6, 52.2 7.34 2, 36.3 17.90 12, 44.4 3.24
F value P value
0.329 6, 37.6 1.18 <0.0001 1, 37.6 17.55 0.362 6, 37.6 0.93
Yield F value P value
1000 Kernel Weighte df
FDK F value P value df
<0.0001 6, 59.9 0.72 <0.0001 2, 39.3 10.19 0.002 12, 47 0.90
0.631 0.0003 0.557
P value 0.193 <0.0004 0.883
1000 Kernel Weight
F value P value
6, 58.2 1.99 2, 37.7 10.93 12, 56 0.78
0.337 6, 38.7 1.53 <0.0002 1, 38.7 15.19 0.483 6, 38.7 0.39
0.082 0.0002 0.668
6, 54.2 2, 39.6 12, 46.5
0.94 20.79 2.34
0.476 <0.0001 0.019
FHB index was calculated from visual disease ratings taken 21 days after anthesis. DON as measured in mg/g quantiﬁed from a post-harvest grain sample. Yield (kg/ha) was adjusted for moisture content (13.5%) prior to analysis. FDK, the percent of kernels damaged by Fusarium graminearum, was visually estimated post-harvest from a 40 mL grain sample from each experimental plot. One thousand kernel weight (1000 kernel weight) as measured in grams (g) was determined by enumerating 1000 kernels and weighing them. Bolded P-value indicates signiﬁcance at a ¼ 0.05.
inoculated, fungicide-treated plots was signiﬁcantly higher than inoculated, non-fungicide treated plots. Additionally, 1000 kernel weight of inoculated, fungicide-treated plots was higher than naturally infected, non-fungicide treated plots (P ¼ 0.0004), and 1000 kernel weight of inoculated, non-fungicide treated plots was lower than naturally infected, non-fungicide treated plots (P ¼ 0.0106). 3.4. Effect of fungicide treatment and fungicide and inoculation timing on FDK Fungicide treatment signiﬁcantly reduced FDK in 2013 and 2014 and did not vary depending on inoculation timing. Statistically
similar levels of FDK developed when plants were inoculated from anthesis to 11 DAA (Table 1 and Fig. 5). 3.5. Effect of fungicide treatment and fungicide and inoculation timing on yield There were no interaction effects between fungicide treatment and inoculum timing on yield in 2013 or 2014 (Table 1). In both years, fungicide treatment signiﬁcantly increased yield, but inoculum timing was not signiﬁcant overall. Statistically similar yields were observed when plots were inoculated from 0 DAA to 11 DAA (Fig. 6). In 2013, fungicide treatment signiﬁcantly increased yield at 1, 3, and 5 DAA compared to yield in inoculated, non-fungicide
Fig. 3. Effect of inoculum application time and fungicide treatment on Fusarium head blight (FHB) index in (A) 2013 and (B) 2014. Error bars represent the upper and lower limits of the standard error of the mean values based on least squares means estimations. aInoculum was applied at a total volume of 300 mL/plot at 50,000 Fusarium graminearum macroconidia/mL. bThe fungicide prothioconazole þ tebuconazole was applied at 475 mL/ha with 0.125% v/v of a non-ionic surfactant.
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
Fig. 4. Effect of inoculation and fungicide treatment timing on deoxynivalenol (DON, as measured in mg/g in (A) 2013 and (B) 2014. Error bars represent the upper and lower limits of the standard error of the mean values based on least squares means estimations. aInoculum was applied at a total volume of 300 mL/plot at 50,000 Fusarium graminearum macroconidia/mL. bThe fungicide prothioconazole þ tebuconazole was applied at 475 mL/ha with 0.125% v/v of a non-ionic surfactant at FGS 10.5.1.
treated plots treated on those same days. In 2014, plots that were treated with fungicide had a signiﬁcantly higher yield than both inoculated, non-fungicide treated plots and naturally infected, nonfungicide treated plots.
3.6. Correlations In 2013, DON was not signiﬁcantly associated with either FHB index or FDK in inoculated, non-fungicide treated plots across all
Fig. 5. Effect of fungicide treatment and fungicide and inoculation timing on Fusarium damaged kernels (FDK), reported as percent visually damaged kernels, in (A) 2013 and (B) 2014. Error bars represent the upper and lower limits of the standard error of the mean values based on least squares means estimations. aInoculum was applied at a total volume of 300 mL/plot at 50,000 Fusarium graminearum macroconidia/mL. bThe fungicide prothioconazole þ tebuconazole was applied at 475 mL/ha with 0.125% v/v of a non-ionic surfactant.
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
Fig. 6. Effect of _and fungicide treatment and fungicide and inoculation timing on yield, adjusted to 13.5% grain moisture, in (A) 2013 and (B) 2014. Error bars represent the upper and lower limits of the standard error of the mean values based on least squares means estimations. aInoculum was applied at a total volume of 300 mL/plot at 50,000 Fusarium graminearum macroconidia/mL. bThe fungicide prothioconazole þ tebuconazole was applied at 475 mL/ha with 0.125% v/v of a non-ionic surfactant.at FGS 10.5.1.
inoculation timings (Table 2). However, in inoculated, fungicidetreated plots, DON was positively associated with FHB index, but the r-value was relatively weak (Table 2). In 2014, DON was positively associated with both FHB index and FDK (Table 3). The association between FDK and DON was stronger than the association between FHB index and DON (Table 3). In inoculated, fungicide treated plots, no association was observed between FHB index and DON, but FDK and DON were positively associated.
Table 2 Pearson's correlation tests for associations between Fusarium head blight (FHB) index, deoxynivaleol (DON), Fusarium damaged kernels (FDK), 1000 kernel weight (1000 KW), and yield within fungicide treatments from 2013. Relationship
Index e DON Index e FDKc Index e 1000 KW Index e Yieldd DON e FDK DON e 1000 KW DON e Yield FDK e 1000 KW FDK e Yield Yield e 1000 KW
Inoculum, No fungicide
0.203 0.286 0.278 0.494 0.078 0.261 0.157 0.076 0.316 0.424
0.299 0.141 0.153 0.008 0.693 0.181 0.424 0.702 0.102 0.025
0.449 0.342 0.154 0.320 0.273 0.081 0.016 0.083 0.036 0.212
0.017e 0.075 0.433 0.097 0.160 0.684 0.937 0.676 0.855 0.278
a FHB index was calculated from visual disease ratings taken 21 days after anthesis. b DON as measured in mg/g quantiﬁed from a post-harvest grain sample. c FDK, the percent of kernels damaged by F. graminearum, was visually estimated post-harvest from a 40 mL grain sample from each experimental plot. d Yield (kg/ha) was adjusted for moisture content (13.5%) prior to analysis. e Bolded P-value indicates signiﬁcance at a ¼ 0.05.
4. Discussion The results of this study indicate that the recommendation to apply fungicides precisely at beginning anthesis (FGS 10.5.1) to manage FHB could be modiﬁed to include applications up to 11 days after primary tillers reach anthesis when the spray coincides with initial inoculum becoming available to the wheat plant. Current recommendations for fungicide application for suppression of FHB in the Midwest are to apply fungicide when 50% of the primary tillers are beginning anthesis in order to protect as many highyielding wheat heads as possible from infection by F. graminearum. This is based on research that demonstrated that wheat is most susceptible to infection by F. graminearum from FGS 10.5.1 through FGS 11.2 (Andersen, 1948). This study demonstrates that post-anthesis applications of prothioconazole þ tebuconazole can reduce FHB index, DON, and FDK, and increase yield similarly to fungicide applications at beginning anthesis up to 11 DAA when inoculum is available to infect the plant. These results are consistent with D'Angelo et al. (2014) who found that fungicide applications up to 6 days post-anthesis consistently reduced DON and FDK levels when inoculum was applied at anthesis. Our recent ﬁndings indicate that post-anthesis fungicide applications are efﬁcacious for an additional 5 days beyond what has previously been described, when inoculum availability coincides with later growth stages. They also conﬁrm that winter wheat is susceptible to infection by F. graminearum for up to two weeks after initial anthesis (FGS 10.5.1), which is consistent with Del Ponte et al. (2007) who found that wheat could incur FHB and develop DON when inoculated as late as FGS 11.3 (hard dough). These ﬁndings can beneﬁt farmers who may ﬁnd it difﬁcult to spray wheat at beginning anthesis due to uneven ﬂowering across a ﬁeld and rain near ﬂowering that prevents driving machinery through a ﬁeld.
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
Table 3 Pearson's correlation tests for associations between Fusarium head blight (FHB) index, deoxynivaleol (DON), Fusarium damaged kernels (FDK), 1000 kernel weight (1000 KW), and yield within fungicide treatments from 2014. Relationship
Inoculum, No fungicide
Index e DON Index e FDKc Index e 1000 KW Index e Yieldd DON e FDK DON e 1000 KW DON e Yield FDK e 1000 KW FDK e Yield Yield e 1000 KW a b c d e
Natural infection, No fungicide
0.537 0.197 0.360 0.110 0.634 0.456 0.101 0.355 0.141 0.058
0.003 0.316 0.060 0.576 <0.001e 0.015 0.610 0.064 0.473 0.769
0.14298 0.03227 0.12550 0.05384 0.55565 0.43058 0.39759 0.38193 0.34001 0.04088
0.468 0.871 0.525 0.786 0.002 0.022 0.036 0.045 0.077 0.836
0.17512 0.26507 0.17203 0.03127 0.05363 0.20888 0.46894 0.01452 0.00746 0.10471
0.373 0.173 0.381 0.875 0.786 0.286 0.012 0.942 0.970 0.596
FHB index was calculated from visual disease ratings taken 21 days after anthesis. DON as measured in mg/g quantiﬁed from a post-harvest grain sample. FDK, the percent of kernels damaged by F. graminearum, was visually estimated post-harvest from a 40 mL grain sample from each experimental plot. Yield (kg/ha) was adjusted for moisture content (13.5%) prior to analysis. Bolded P-value indicates signiﬁcance at a ¼ 0.05.
Farmers should still target anthesis when applying prothioconazole þ tebuconazole, but if weather favors inoculum development post-anthesis, they may see similar levels of disease suppression when coinciding fungicide applications are made up to 11 days after anthesis. Although fungicide application did reduce DON levels in this study, mean DON levels in all treatments across both years exceeded 2 mg/g. This is the level at which growers typically begin to experience price dockages when selling grain. DON levels exceeded 2 mg/g even when weather conditions did not favor disease development and wheat was treated with fungicide up to 11 DAA. Therefore, even though fungicide applications may reduce DON, they may not be enough to prevent economic loss due to DON in all years. Yoshida et al. (2012) demonstrated that DON could be reduced by applying the fungicide thiophanate-methyl at 20 DAA, but this application timing would not be practical in winter wheat due to the relatively short period of time between anthesis and harvest (~30e40 days). Several studies have evaluated the effect of moderately resistant cultivars on DON levels, but no cultivars have yet been developed that are completely resistant to DON accumulation (Bai et al., 2001; Saldago et al., 2014). Therefore, although combinations of moderately resistant cultivars and fungicide applications are still the best way to reduce DON levels, additional strategies should be investigated to ﬁnd ways to further reduce DON in grain. Since complete control of DON is not yet possible, several strategies are employed to predict the impact of FHB on grain quality and attempt to determine ﬁnal DON levels prior to sale or use of grain. Two visual estimators of FHB severity used are FHB index and FDK grain quality assessment. In this study, several positive correlations were found between FHB index, FDK, and DON, but significant associations were not consistent between years. These results are similar to those from a meta-analysis conducted by Paul et al. (2005) who found that associations between disease variables were higher in years with lower disease intensity. Because visual FHB symptoms do not occur until approximately 5 days after infection, and are difﬁcult to distinguish from natural senescence due to ripening, it is not surprising that FHB index is not consistently associated with ﬁnal DON levels (Andersen, 1948; Cowger and Arellano, 2013). Our results further serve to demonstrate that FHB index and FDK are not consistently useful estimators of DON levels, and quantitative methods such as immunostrip DON quantiﬁcation kits, high performance liquid chromatography (HPLC), or an enzyme-linked immunosorbant assay (ELISA) should be used to assess DON prior to sale of grain.
In both years of this study, DON levels increased in nonfungicide treated plots that were inoculated on the evening prior to rain. DON functions as a virulence factor in the colonization of wheat by F. graminearum, allowing hyphae to colonize the rachis node and move from spikelet to spikelet (Brown et al., 2010; Jansen et al., 2005). DON has been shown to be most active at the infection front and transcription of DON biosynthesis genes has been detected within 24 h of initial infection (Hallen-Adams et al., 2011). DON levels have also been shown to increase in wheat tissue up to 45 DAA and to be positively inﬂuenced by increased levels of moisture during grain ﬁll up to 30 DAA (Cowger and Arellano, 2013). In a series of models developed by Hooker et al. (2002) to predict DON levels in spring wheat in Ontario, Canada, precipitation that exceeded 3 mm, 3 to 6 DAA was one of the most important predictors for ﬁnal DON accumulation, and it was the most important predictor 7 to 10 DAA. In this study, plots treated with fungicide did not exhibit increases in DON levels, suggesting that the fungicide may reduce the impact of moisture on DON levels. Since DON is synthesized primarily at hyphal tips, and DMI triazole fungicides function by preventing the normal growth of hyphae through disruption of the ergosterol biosynthetic pathway (Kӧ;ller, 1992; Siegel, 1981), it is possible that the fungicide disrupts the advancing fungal hyphae and prevents an upsurge in DON production after a rain event. In this study, fungicide was applied approximately 8 h prior to inoculation. Therefore, we were primarily investigating the protective effect of prothioconazole þ tebuconazole on F. graminearum infection. DMI triazole fungicides have been shown to have both pre- and post-infection activity, meaning they function both to protect the plant from initial infection and to hinder further colonization of plant tissue after infection begins (Andersen et al., 2014; Ivic, 2010; Mueller and Bradley, 2008; Szkolnik, 1981). Postinfection activity has been demonstrated to be useful for 1e5 days, and is less likely to be adversely affected by rain than the protective effect once the fungicide has been absorbed into the plant tissue (Andersen et al., 2014; Ivic, 2010). The dual action of DMI fungicides is particularly important in extending the time of fungicide efﬁcacy after a spray due to the fact that infection can occur from FGS 10.5.1 (early anthesis) through FGS 11.2 (soft dough) (Andersen, 1948). Therefore, although this study primarily investigated the protective effect of fungicide, differences in FHB index between inoculated, fungicide treated plots compared to inoculated, non-fungicide treated plots indicate that fungicide likely also had post-infection activity on wheat heads that had been naturally infected from 1 to 5 days prior. This may indicate that the fungicide
A.N. Freije, K.A. Wise / Crop Protection 77 (2015) 139e147
was exhibiting some post-infection activity on natural infection that had occurred at anthesis, as well as protecting the wheat heads that were susceptible at the time inoculum was applied. In summary, this research indicates that post-anthesis fungicide applications can reduce FHB and DON, giving farmers ﬂexibility in application timing should weather, or other factors, delay application timing at anthesis. This research also demonstrates the importance of predicting when environmental conditions are conducive for F. graminearum spore production and infection near anthesis. With a wide timeframe of fungicide efﬁcacy and wheat susceptibility to infection, the improvement of current forecasting systems could help growers to optimize their fungicide application instead of applying solely at beginning anthesis. However, further research is necessary to determine if these results are consistent across wheat cultivars, fungicides, and environments. Research is also needed to evaluate the contribution of secondary tiller infection to DON levels and to assess whether alternative methods of assessing FHB index levels would lead to stronger associations between FHB Index and DON. Acknowledgments Funding for this project was provided in part by the United States Wheat and Barley Scab Initiative (USWBSI) (59-0206-4-017). We thank N. Anderson, J. Ravellette, and P. Romero for their technical assistance; D. Anderson, G. Carmona, R. Fister, K. Lewis, and K. Ruan for seasonal ﬁeld assistance; Y. Dong for help with deoxynivalenol analysis, and Y. Li for assistance in statistical analysis. We also thank P. Romero and G. Shaner for valuable comments and suggestions to improve the manuscript. References Andersen, A.L., 1948. The development of Gibberella zeae headblight of wheat. Phytopathology 38, 595e611. Andersen, K.F., Morris, L., Derksen, R.C., Madden, L.V., Paul, P.A., 2014. Rainfastness of prothioconazole þ tebuconazole for Fusarium head blight and deoxynivalenol management in soft red winter wheat. Plant Dis. 98, 1398e1406. Bai, G.H., Shaner, G., 1996. Variation in Fusarium graminearum and cultivar resistance to wheat scab. Plant Dis. 80, 975e979. Bai, G.H., Plattner, R., Desjardins, A., Kolb, F., 2001. Resistance to Fusarium head blight and deoxynivalenol accumulation in wheat. Plant Breed. 120, 1e6. Bockus, W., Bowden, R.L., Hunger, R.M., Morrill, W.L., Murray, T.D., Smiley, R.W., 2010. Compendium of Wheat Diseases and Pests, third ed. APS Press, St. Paul, Minnesota. Brown, N.A., Urban, M., Van de Meene, A.M., Hammond-Kosack, K.E., 2010. The infection biology of Fusarium graminearum: deﬁning the pathways of spikelet to spikelet colonization in wheat ears. Fung. Biol. 7, 555e571. Cowger, C., Arellano, C., 2013. Fusarium graminearum infection and deoxynivalenol concentrations during development of wheat spikes. Phytopathology 103, 460e471. D'Angelo, D.L., Bradley, C.A., Ames, K.A., Willyerd, K.T., Madden, L.V., Paul, P.A., 2014. Efﬁcacy of fungicide applications during and after anthesis against Fusarium head blight and deoxynivalenol in soft red winter wheat. Plant Dis. 98, 1387e1397. Del Ponte, E.M., Fernandes, J.M.C., Bergstrom, G.C., 2007. Inﬂuence of growth stage on Fusarium head blight and deoxynivalenol production in wheat. J. Phytopathol. 155, 577e581. Fungicide Resistance Action Committee, 2011. Mode of Action of Fungicides. Online. www.frac.info/frac/publication/anhang/FRAC%20MoA%20Poster%202011_ﬁnal_ HR.pdf (accessed date 24.01.13.). Goswami, R.S., Kistler, H.C., 2004. Heading for disaster: Fusarium graminearum on cereal crops. Mol. Plant Pathol. 5, 515e525. Hallen-Adams, H.E., Wenner, N., Kuldau, G.A., Trail, F., 2011. Deoxynivalenol
biosynthesis-related gene expression during wheat kernel colonization by Fusarium graminearum. Phytopathology 101, 1091e1096. Hart, L.P., Pestka, J.J., Liu, M.T., 1984. Effect of kernel development and wet periods on production of deoxynivalenol in wheat infected with Gibberella zeae. Phytopathology 74, 1415e1418. Hellevang, K.J., 1995. Grain Moisture Content Effects and Management. North Dakota State University Extension Service. Online Publication/AE-905 (Revised). http://www.ag.ndsu.edu/extension-aben/documents/ae905.pdf. Access date: 10.11.14. Hooker, D.C., Schaafsma, A.W., Tamburic-Ilincic, L., 2002. Using weather variables pre and post-heading to predict deoxynivalenol content in winter wheat. Plant Dis. 86, 611e619. Ivic, D., 2010. Curative and eradicative effects of fungicides. In: Carisse, O. (Ed.), Fungicides. IntEch, pp. 3e22. http://dx.doi.org/10.5772/555. Jansen, C., Von Wettstein, D., Sch€ afer, W., Kogel, K., Felk, A., Maier, F.J., 2005. Infection patterns in barley and wheat spikes inoculated with wild-type and trichodiene synthase gene disrupted Fusarium graminearum. P. Natl. Acad. Sci. U. S. A. 102, 16892e16897. Jones, R.K., Mirocha, C.J., 1999. Quality parameters in small grains from Minnesota affected by Fusarium head blight. Plant Dis 84, 506e511. Kӧller, W., 1992. Antifungal agents with target sites in sterol functions and biosynthesis. In: Target Sites of Fungicide Action. CRC Press, Boca Raton, FL, pp. 119e207. Large, E.C., 1954. Growth stages in cereals illustration of the Feekes scale. Plant Pathol. 3, 128e129. McMullen, M., Bergstrom, G., De Wolf, E., Dill-Macky, R., Hershman, D., Shaner, G., Van Sanford, D., 2012. A uniﬁed effort to ﬁght an enemy of wheat and barley: Fusarium head blight. Plant Dis. 96, 1712e1728. McMullen, M., Jones, R., Gallenberg, D., 1997. Scab of wheat and barley: a reemerging disease of devastating importance. Plant Dis. 81, 1340e1348. zy, A., 1995. Types and components of resistance to Fusarium head blight Mesterha of wheat. Plant Breed. 114, 377e386. Mueller, D.S., Bradley, C.A., 2008. Field Crop Fungicides for the North Central United States. North Central Integrated Pest Management Center. O'Donnell, K., Kistler, H.C., Tacke, B.K., Casper, H.H., 2000. Gene genealogies reveal global phylogeographic structure and reproductive isolation among lineages of Fusarium graminearum, the fungus causing wheat scab. P. Natl. Acad. Sci. U. S. A. 97, 7905e7910. Paul, P.A., Lipps, P.E., Hershman, D.E., McMullen, M.P., Draper, M.A., Madden, L.V., 2008. Efﬁcacy of triazole-based fungicides for Fusarium head blight and deoxynivalenol control in wheat: a multivariate meta-analysis. Phytopathology 98, 999e1011. Paul, P.A., Lipps, P.E., Madden, L.V., 2005. Relationship between visual estimates of Fusarium head blight intensity and deoxynivalenol accumulation in harvested wheat grain: a meta-analysis. Phytopathology 95, 1225e1236. Saldago, J.D., Madden, L.V., Paul, P.A., 2014. Efﬁcacy and economics of integrating inﬁeld and harvesting strategies to manage Fusarium head blight of wheat. Plant Dis. 98, 1407e1421. Schnabel, G., Jones, A.L., 2001. The 14a-demethylasse (CYP51A1) gene is overexpressed in Venturia inaequalis strains resistant to myclobutanil. Phytopathology 91, 102e110. Siegel, M.R., 1981. Sterol-inhibiting fungicides: effects on sterol biosynthesis and sites of action. Plant Dis. 65, 986. Stack, R.W., McMullen, M.P., 2011. A Visual Scale to Estimate Severity of Fusarium Head Blight in Wheat. North Dakota State University Extension Services: Small Grains Publications, p. 1095. Online Publication. http://www.ag.ndsu.edu/pubs/ plantsci/smgrains/pp1095.pdf. Sutton, J.C., 1982. Epidemiology of wheat head blight and maize ear rot caused by Fusarium graminearum. Can. J. Plant Pathol. 4, 195e209. Szkolnik, M., 1981. Physical modes of action of sterol-inhibiting fungicides against apple diseases. Plant Dis. 65, 981e985. Willyerd, K.T., Li, C., Madden, L.V., Bradley, C.A., Bergstrom, G.C., Sweets, L.E., McMullen, M., 2011. Efﬁcacy and stability of integrating fungicide and cultivar resistance to manage Fusarium head blight and deoxynivalenol in wheat. Plant Dis. 96, 957e967. Wise, K.A., 2014. Diseases of Wheat: Fungicide Efﬁcacy for Controlling Wheat Diseases. Purdue Extension, Purdue University, West Lafayette, IN. BP-162-W. 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 contamination in wheat. Plant Dis. 96, 845e851. Yoshida, M., Nakajima, T., 2010. Deoxynivalenol and nivalenol accumulation in wheat infected with Fusarium graminearum during grain development. Phytopathology 100, 763e773.