FY16 USWBSI RECOMMENDED PROJECT(S) PROPOSAL · FY16 USWBSI RECOMMENDED PROJECT(S) PROPOSAL NOTE: This document will be retained in the USWBSI’s NFO COVER PAGE Instructions: Update

FY16 USWBSI RECOMMENDED PROJECT(S) PROPOSAL NOTE: This document will be retained in the USWBSI’s NFO

COVER PAGE

Instructions: Update any contact information that is not correct, sign cover page, and attach to your FY16 USWBSI Individual Project Proposal(s)

USWBSI Grant Title: Modeling The Effects of Weather on FHB And DON and Developing Robust Strategies to Minimize Losses.

Principal Investigator (PI): Pierce Paul PI’s Institution: Ohio State University

PI’s Address: Department of Plant Pathology 1680 Madison Ave. Wooster, OH 44691

PI’s E-mail: [email protected] PI’s Phone: 330-263-3842

PI’s Fax: 330-263-3841 Fiscal Year (FY): 2016

ARS Agreement Number: 59-0206-4-018 Institution’s Indirect Cost Rate

for the USWBSI: 5% DC

USWBSI’s FY16 Total Recommended Amount:

$ 55,571

______________________________________03/11/2016__ Principal Investigator’s Signature Date

USWBSI Project ID

USWBSI Research Category USWBSI Project Title

PI Requested Amount

USWBSI‘s Recommended

Amount

FY16-IM-013 MGMT Efficacy and Curative Effects of Fungicides for FHB and DON Management in Ohio.

$ 28,465 $ 29,847

FY16-DE-022 MGMT Functional Analysis for Getting Better Weather-based Predictors of Fusarium Head Blight.

$ 10,920 $ 10,920

FY16-PA-030 MGMT Risk-based Fungicide Decision-making for FHB and DON Management in Wheat.

$ 14,804 $ 14,804

USWBSI’s FY16 Total Recommended Amount $ 55,571

FY16 USWBSI INDIVIDUAL PROJECT PROPOSAL

FY16 USWBSI Individual Project Proposal

PROJECT SUMMARY PAGE Principal and Co- Investigator(s):

Principal Investigator: Pierce Paul Institution: Ohio State University

Co-Investigator #1: Laurence V Madden Institution: Ohio State University

Co-Investigator #2: Jorge David Salgado Institution: Ohio State University

Co-Investigator #3: Institution:

Project Title 1: Efficacy and Curative Effects of Fungicides for FHB and DON Management in Ohio. PROJECT SUMMARY

The overall goal of this round of the MGMP-CP is to develop more robust guidelines to provide producers with additional and more effective options for managing FHB and DON. The specific objectives of this proposed research project are to:

1. Determine the efficacy and economics of integrating pre- and post-anthesis fungicide applications and cultivar resistance to minimize losses due to FHB and DON.

2. Investigate the curative effect of Prosaro® and Caramba® on FHB and DON. Results from these experiments will improve our understanding of the efficacy and economics of more robust integrated strategies for FHB and DON management. In addition, invaluable information pertaining to the mode of action of Prosaro and Caramba and insights into their effects on infection and symptom development will be gained from this project. To accomplish the goal and objectives of this project, field and growth chamber/greenhouse experiments will be conducted during the 2015-2016 and 2016-2017 growing seasons. In all cases, the experimental design will be a randomized complete block, with a split-plot arrangement of treatment factors. There will be four replicate blocks. For Objective 1, fungicide treatment programs, consisting of 1) an untreated check, 2) Prosaro at anthesis; 3) Prosaro at anthesis followed by Caramba 4 days later; 4) Caramba at anthesis followed by tebuconazole 4 days later; 5) Proline at anthesis followed by tebuconazole 4 days later; and 6) an untreated, non-inoculated check, will be applied to plots of susceptible, moderately susceptible, and moderately resistant cultivars. For Objective 2, for both field and growth chamber experiments, 14 fungicide treatments consisting of preventative (before infection) and curative (after infection) applications of Prosaro and Caramba, will be evaluated. These treatments will be applied to plots of a moderately resistant and a susceptible cultivar in the field, and evaluated under point and spray inoculations in the growth chamber. For both objectives, FHB, DON and yield will be quantified, and efficacy will be defined in terms of percent control relative to the untreated susceptible check. Statistical models will be fitted to all data to determine the main and interaction effects of treatment programs and cultivar, and to model the temporal change in efficacy in response to curative applications. Finding from these experiments will be used to develop new, and improve existing, management guidelines for FHB, providing stakeholders with more options and greater flexibility in terms of fungicide application programs for minimizing losses caused by FHB and DON.



PLAN OF WORK

Rationale and Significance: As stated in the MGMT ACTION PLAN and outlined in the standard protocol, since current management guidelines such as a single fungicide application at anthesis is generally not sufficient to control FHB under highly favorable conditions, further research is needed to evaluate other options and more robust approaches for minimizing the impact of this disease. During the FY14 and FY15 rounds of the MGMT-CP, we evaluated the effects of post-anthesis applications of Prosaro and Caramba on FHB and DON and found that applications made up to 6 days after anthesis effectively reduced both the disease and the toxin (D’Angelo et al. 2014; Salgado et. al. 2014a). This will allow producers greater flexibility in terms of fungicide timing for FHB management, particularly when adverse conditions prevent applications from being made at anthesis. The success of “late” or “post-anthesis” fungicide applications raised several questions worthy of further investigations, including: 1) what is the efficacy and economic benefit of following an anthesis application with a second application after anthesis for FHB/DON management?; 2) is the response to late application due to a curative effect of Prosaro and Caramba?; 3) is the response due to protection of late-flowering primary tillers and/or secondary tillers that were not flowering at the time of the anthesis application?; 4) how late can Prosaro and Caramba be applied and still effectively reduce FHB and DON? Question 1 will be addressed as part of the MGMT-CP, and we propose a second set of complimentary experiments to address questions 2 and 4. Answers to these questions will be invaluable for future efforts to integrate late applications into FHB management programs. Demethylation inhibitor (DMI) fungicides are generally used as protectants for FHB management in wheat to prevent infection of the spikes during anthesis. However, previous research on the effects of DMI fungicides on fungal diseases of wheat and other crops suggested that this class of fungicides can also be used in a curative manner, since post-infection treatments helped to reduce sporulation, colonization, and symptom development (Gachomo et al. 2009, Han et al. 2006; Mangin-Peyrard and Pepin 1996; Schöfl and Zinkernagel 1997). Do Prosaro and Caramba have similar effects on FHB when applied after infection with F. graminearum? Although the standard recommendation is for these fungicides to be applied preventatively at 50% anthesis, the growth stage at which the wheat crop is most vulnerable to infection (Andersen 1948), results from recent studies showed that an application of Caramba or Prosaro at 2, 4, or 6 days after anthesis was just as effective as an anthesis application for FHB and DON control (Bradley et al. 2009; D’Angelo et al. 2014; Salgado et al. 2014a). Applications made at 4-6 days after anthesis reduced FHB and DON by an average of 62 and 50%, respectively, relative to the untreated check. The corresponding levels of control for the anthesis applications were 56 and 50%. Based on previous research conducted using a sister fungicide, Tilt (Boyacioglu et al. 1992), the authors speculated that Caramba and Prosaro may have a curative effect on FHB that results in effective control after infection occurs. Another hypothesis was that the so-called late applications protected secondary tiller that were not yet at anthesis at the time of the anthesis applications, but had reached anthesis by the time the “late” applications were made. It is difficult to effectively determine and quantify the curative effect of Prosaro and Caramba based on average FHB and DON responses from trials conducted under field conditions, since this requires spikes to be inoculated/infected at a known growth stage (anthesis in this case), and then treated at regular intervals after infection. In standard field experiments, plot average FHB, and particularly, DON responses to fungicides are based on spikes at a range of growth states at the time of treatment application, since anthesis varies among tillers within the same plot. We propose to use systematic sampling of spike at different growth stages in field experiments, along with cohorts of spikes at know growth stages in controlled-environment experiments to evaluate the curative effect of Prosaro and Caramba. We hypothesize that i) efficacy (percent control of FHB and DON relative to the untreated check) will decrease in a non-linear manner as the time between infection and application increases; ii) the magnitude of the difference in efficacy between preventative and curative treatments will vary with baseline levels of FHB and DON. This proposal falls under Goals 1 (“Develop integrated management strategies for FHB and mycotoxins that are robust to conditions experienced in production fields”) and 2 (“Help develop and validate the next generation of management tools for FHB/DON control”) of the MGMT Action Plan and addresses the following Research Needs: “Evaluating the flexibility of fungicide



application timing within the context of integrated management strategies and “Developing economic analyses of effective integrated management strategies used alone and in combination”. Research Material and Methods Description and sequence of proposed experiments: Field and growth chamber/greenhouse experiments will be conducted in Wooster, Ohio during the 2015-2016 and 2016-2017 growing seasons to investigate the efficacy and curative effects of fungicide programs that include post-infection applications on FHB and DON in moderately resistant and susceptible soft red winter wheat cultivars. Plots will be established on the Snyder research farm near the OARDC campus in fields previously planted with oats or soybeans and managed following standard agronomic practices in Ohio. In all cases, the experimental design will be a randomized complete block (RCB), with a split-plot arrangement of treatment factors. There will be four replicate blocks. Objective 1 (STANDARD PROTOCOL) – Determine the Efficacy and Economics of Integrating Pre- and Post-anthesis Fungicide Applications and Cultivar Resistance to Minimize Losses due to FHB and DON. In this experiment, cultivar will be the whole-plot and fungicide treatment the sub-plot. Strips of four SRWW cultivars (Hopewell - susceptible standard, mid-season, and moderate to high yield potential; Bromfield – moderately susceptible, mid-season, and moderate to high yield potential; Malabar - moderately resistant, mid-season, and moderate to high yield potential; and Truman - standard moderately resistant, late maturing and low to moderate yield potential) will be divided into 5 x 20-ft sub-plots to which fungicide treatments will be randomly assigned. Fungicide treatment programs will consist of: 1) an untreated check; 2) Prosaro at anthesis; 3) Prosaro at anthesis followed by Caramba 4 days later; 4) Caramba at anthesis followed by tebuconazole 4 days later; 5) Proline at anthesis followed by tebuconazole 4 days later; and 6) an untreated, non-inoculated check (See Fig 1 of the standard protocol for details). Strips (5-ft wide) of the cultivar Truman will be planted between adjacent whole- and sub-plots to minimize interplot interference. Treatments will be applied, plots inoculated, FHB and foliar disease intensity rated, and FDK, DON, yield and test weight data collected as described in the STANDARD PROTOCOL. Objective 2 - Investigate the curative effect of Prosaro and Caramba on FHB and DON. A field and a growth chamber experiment will be conducted to accomplish this objective. For the field experiment, cultivar (Hopewell and Malabar) will be used as whole-plot and a factorial arrangement of fungicide and application timing as sub-plot. Each whole-plot will consist of 15 sub-plots (5 x 20-ft) to which an untreated check and 14 fungicide treatments will be randomly assigned (Fig 1). Prosaro (6.5 fl oz/A + 0.125% induce) or Caramba (13.5 fl oz/A + 0.125% induce) will be applied to separate sub-plots either at anthesis (the preventative treatments), or at 2, 4, 6, 8, 15 or 21 days after anthesis (the curative treatments). Note: Due to pre-harvest restrictions, the latter two applications will not be of practical value, but will be useful for characterizing and modeling the temporal change in efficacy. Strips (5-ft wide) of Truman will again be planted between plots to minimize interplot interference. Fifty spikes that reach anthesis (with anthers extruded) at the time of the preventative application and another 50 (on secondary tillers and/or late-developing primary tillers) not at anthesis will be tagged in each plot, rated for scab as described in the STANDARD PROTOCOL, hand-harvested and threshed, and grain samples will be rated for FDK and assayed for DON. A second experiment similar to that described above will be conducted under controlled conditions (greenhouse and temperature-controlled growth chamber). Seeds of Hopewell will be planted in trays, vernalized, transplanted to cones, and then managed as previously described (Andersen et al. 2014, 2015). The experimental design will be a RCB, with a split-plot arrangement of inoculation method (point and spray) and a factorial arrangement of fungicide and application timing (described above) as sub-plot. Cohorts of wheat plants at anthesis will serve as the blocking factor. Fungicide treatments will be applied using a hand-held sprayer at rates and volumes comparable to that used in the field experiment (Andersen et al. 2014). At 24 hours after the anthesis (preventative) treatments are applied, approximately 50% of the spikes (20-50) in each experimental unit will be spray-inoculated and the other 50% will be point-inoculated as previously described (Andersen et al. 2014, 2015). Immediately after inoculation, plants will be placed in a mist chamber and subjected to 24 hours of mist (90 sec



of mist every 10 min, for 12 h during each 24-hour period) to enhance infection. Post-infection (curative) treatments will then be applied accordingly, beginning with the 2-day post-infection treatments, which will be applied shortly after plants are removed from the mist chamber and allowed to dry. Results expected: We expect to see differences in all measures of FHB, yield, and grain quality among cultivars and among fungicide programs in all experiments. For Objective 1, two-treatment programs will likely have lower levels of FHB and DON and higher grain yield and test weight than the standard single-treatment program. Among the two-treatment programs, program 3 (Prosaro at anthesis followed by Caramba four days later) will likely be the most effective, but programs with Folicur (programs 4 and 5) will likely be the most economically beneficial. For Objective 2, we expect curative treatments applied between 2 and 8 days after infection to have comparable efficacy (based on percent control relative to the untreated check) to the preventative treatment, but efficacy will decline as treatments are delayed beyond 8 days. We further anticipate that comparable results will be observed between Prosaro and Caramba for all measures of FHB, yield and grain quality. Data analysis: For all experiments, the untreated check and the standard anthesis treatments will serve as references against which other fungicide programs will be compared. For Objective 1, linear mixed model analyses of variance (Littell et al. 2006) and meta-analyses (Madden and Paul 2011) of the pooled data from the MGMT-CP will be used to quantify the effects of cultivar, fungicide program, and their interaction on FHB incidence and index, yield, test weight, FDK, and DON as described in the STANDARD PROTOCOL. Summary results from these analyses will be used along with grain prices, price discounts, fungicide application costs, and estimates of yield loss due to wheel tracks to determine the economic benefit of single- and two-treatment fungicide programs, both alone and in combination with genetic resistance, for FHB/DON management (Salgado et al 2014b). Linear mixed models will also be used to analyze data collected under Objective 2, and contrasts will be incorporated into the models to test for linear, logarithmic, or other nonlinear decline in percent control as curative applications are delayed, and to test for the effect of cultivar on the rate of decline. Application of results/technology transfer: Results from these experiments will improve our understanding of the efficacy and economics of more robust integrated strategies for FHB and DON management. In addition, invaluable information pertaining to the mode of action of Prosaro and Caramba and insights into their effects on infection and symptom development will be gained from this project. These findings will be used to develop new, and improve existing, management guidelines for FHB, providing stakeholders with more options and greater flexibility for minimizing losses caused by FHB/DON. Results will be made available to growers, ag-dealers, county extension educators, researchers, and others in the wheat industry through extension presentations, scientific meetings, and peer-reviewed publications. In addition, data from these experiments will be used to advance the development and validation of FHB and DON risk assessment models. Possible pitfalls and limitations: Results from field experiments will depend on the weather. If the weather is entirely unfavorable for FHB, then no disease will develop regardless of treatment. Conversely, if weather is extremely favorable for infection, there may be no significant main effects of any treatment. Artificial inoculations and the use of cultivars with different maturities and levels of resistance will help to minimize the chance of either of these eventualities occurring. Results from previous inoculated studies showed that even when weather conditions were unfavorable for natural infection, FHB still developed in inoculated plots, allowing for treatment comparison. For the controlled-environment experiments, synchronizing crop development and maturity is always a challenge. However, over the years the Paul lab. has learned that this problem can be minimized by conducting greenhouse experiments during the cooler months of the year and by thinning plants and cycling them between greenhouse and growth chambers set at different temperatures at different growth stages. Timeline of Proposed Events Field experiments

o Planting Fall 2015 and 2016 o Treatment application and inoculation Spring 2016 and 2017



o Disease assessment Spring 2016 and 2017 o Harvest and FDK and DON analyses Summer 2016 and 2017 o Data analysis and preparation of reports Fall 2016 and 2017

Controlled-environment experiments o Planting and vernalization Late summer 2016 and 2017 o Transplant and maintenance of plants Fall 2016 and 2017 o Inoculum preparation Fall 2016 and 2017 o Treatment application and data collection Late-Fall 2016/17 – Early-Spring 2017/18 o Data analysis and preparation of reports Late-Fall 2016/17 – Early-Spring 2017/18

Prepare manuscripts for publication Spring/Summer/Fall 2018

Figure 1 - An example field plan of the FHB integrated manage trial described under Objective 2 to evaluate the curative effects of Prosaro and Caramba on FHB. Shade/no shade represents cultivar (susceptible Hopewell and moderately resistant Malabar), the whole-plot, and numbers represent fungicide treatment programs (the sub-plot): 1 = untreated check, 2-8 = Prosaro at anthesis and 2, 4, 6, 8, 15 and 21 days after infection, respectively, and 9-15 = Caramba at anthesis and 2, 4, 6, 8, 15 and 21 days after infection, respectively References to project description: 1. Andersen, A. L. 1948. The development of Gibberella zeae head blight of wheat. Phytopathology 38:595-611. 2. Andersen, K. F., Madden, L. V., and Paul, P. A. 2015. Fusarium head blight development and deoxynivalenol

accumulation in wheat as influenced by post-anthesis moisture patterns. Phytopathology 105:210-219. 3. Andersen, K. F., Morris, L., Derksen, R.C., Madden, L.V., and Paul, P. A. 2014. Rainfastness of

prothioconazole+tebuconazole for Fusarium head blight and deoxynivalenol management in soft red winter wheat. Plant Dis. 98:1396-1406.

4. Boyacioglu, D., Hettiarachchy, N. S. and Stack, R. W. 1992. Effect of three systemic fungicides on deoxynivalenol (vomitoxin) production by Fusarium graminearum in wheat. Can. J. Plant Sci. 72: 93-101.

5. Bradley, C. A., Adee, E. A., Ebelhar, S. A, Grybauskas, A. P., Hollingsworth, C. R., Kirk, W. W., McMullen, M. P., Milus, E. A., Osborne, L. E., Ruden, K. R., and Young, B. G. 2009. Application timings of Caramba and Prosaro foliar fungicides for management of FHB and DON. Page 34 in: Proc. 2009 National Fusarium



Head Blight Forum. S. Canty, A. Clark, J. Mundell, E. Walton, D. Ellis, and D. Van Sanford, eds. Orlando, FL.

6. D’Angelo, D. L., Bradley, C. A., Ames, K. A., Willyerd, K. T., Madden, L. V., and Paul, P. A. 2014. Efficacy of fungicide applications during and after anthesisagainst Fusarium head blight and deoxynivalenol in soft red winter wheat. Plant Dis. 98:1387-1397.

7. Gachomo, E.W., Dehne, H.W., and Steiner, U. 2009. Efficacy of triazoles and strobilurins in controlling black spot disease of roses caused by Diplocarpon rosae. Annals of Applied Biology 154: 259-267.

8. Han, Q.M., Kang, Z.S., Buchenauer, H., Huang, L.L., and Zhao, J. 2006. Cytological and immunocytochemical studies on the effects of the fungicide tebuconazole on the interaction of wheat with stripe rust. Journal of Plant Pathology 88:263-271.

9. Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfinger, R. D., and Schabenberger, O. 2006. SAS System for Mixed Models. SAS Institute, Cary, NC.

10. Madden, L. V., and Paul, P. A. 2011. Meta-analysis for evidence synthesis in plant pathology: An overview. Phytopathology 101:16-30.

11. Mangin-Peyrard, M. and Pepin, R. 1996. Encasement of Erysiphe graminis haustoria after treatment with bromuconazole. Pestic. Sci. 46:121-130.

12. Salgado, J. D., Ames, K., Bergstrom, G., Bradley, C., Byamukama, E., Cummings, J., Dill-Macky, R., Friskop, A., Gautam, P., Kleczewski, N., Madden, L., Milus, E., Nagelkirk, M., Ransom, J., Ruden, K., Wegulo, S., Wise, K., and Paul, P. A. 2014. Best FHB Management Practices: A 2014 Multi-State Project Update. In: S. Canty, A. Clark, N. Turcott, and D. Van Sanford, Proceedings of the 2014 National Fusarium Head Blight Forum (pp. 31). East Lansing, MI/Lexington, KY: U.S. Wheat & Barley Scab Initiative.

13. Salgado, J. D., Madden, L. V., and Paul, P. A. 2014. Efficacy and economics of integrating in-field and harvesting strategies to manage Fusarium head blight of wheat. Plant Dis. 98:1407-1421.Schöfl, U. A. and Zinkernagel, V. 1997. A test method based on microscopic assessments to determine curative and protectant fungicide properties against Septoria tritici. Plant Pathol. 46:545-556.

Facilities and Equipment: The PI have access to the research farm space and equipment (e.g. sprayers, plot planters, combine harvesters, grain grinding apparatus etc) necessary to conducted the research proposed herein.



FY16 BUDGET JUSTIFICATION PAGE

Project 1 Title: Efficacy and Curative Effects of Fungicides for FHB and DON Management in Ohio.

Principal Investigator: Pierce Paul

USWBSI’s FY16 Total Recommended Amount: $ 29,847

Instructions: Complete all applicable sections below. If budget category is not applicable, leave blank. NOTE: All amounts

must be rounded to the nearest whole number. A. Direct Labor (salaries and wages): List below the number and titles of personnel, percentage of time/total hours to be devoted to the grant, and rates of pay. Please list according to category/subcategory and include the amount requested for each sub category (i.e. Post Doc, Research Technician, Undergraduate Students, etc.) below next to ‘$’ and the total amount requested for the category (PI/PD, Other Professional Personnel, Support Personnel) in column on the right.

Total $ Requested per

Category PI(s)/PD(s): Funds are being requested for half a month (5.6%) salary support for the PI/PD

$5,018.00

Other Professional Personnel (Post Docs, Specialists (non-tenured faculty), and other administrative professionals):

$12,307.00

Post Doc: Funds are being requested to support salaries for the post-doc/research associate (32%) who, under the supervision of the PD, will continue to coordinate the collection, organization, and analysis of data from the MGMT-CP, and contribute to the establishment and maintenance of experiments described under Objectives 1 and 2.

$11,538.00

Other Administrative Professionals: $

Support Personnel (research technicians, students (graduate and undergraduate), and temporary employees): $ Research Technician(s):

$

Graduate Student(s):

$

Undergraduate Student(s):

$

Temporary Employee(s):

$

B. Fringe Benefits: For each category of personnel, list below the fringe rates, etc. Include the amount requested for each subcategory (i.e. Post Doc, Research Technician, Undergraduate Students, etc.) below next to ‘$’ and the total amount requested for the category (PI/PD, Other Professional Personnel, Support Personnel) in column on the right.


Category PI(s)/PD(s): 29.1% of PI/PD salary

$1,460.00


$4,591.00

Post Doc: 37.3% of post-doc salary

$4,591.00

Other Administrative Professionals:

$


$


$


$


$



D. Nonexpendable Equipment: List below equipment items, relevance to proposed research and dollar amounts. Include cost per item

Total $ Requested

$

E. Materials and Supplies (M/S): Provide below as much detail and specificity as possible for all materials and supplies associated with proposed research. Materials and Supplies should be described in detail e.g., chemical reagents, computer paper and supplies, glassware, lumber, etc. under each sub category (Field, Greenhouse, Laboratory and Other). Include total amount per sub category below next to ‘$’ and total amount requested for M/S in column on the right (i.e. Total $ Requested).

Total $ Requested

Field:

$ $

Greenhouse:

$

Laboratory:

$

Other:

$

F.1. Domestic Travel (DT): List below proposed trips individually and describe their purpose in relation to the grant. Also provide dates, destination, and number of travelers where known. Include total amount per sub category below next to ‘$’and total amount requested for DT in column on the right.

Total $ Requested

Research Related (e.g. travel to research plots): $ $2,221.00

Non-Research Related (i.e. professional meetings): FHB Forum: We are requesting funds (partial support) for the graduate student and PI to attend the 2015 FHB forum to communicate research findings.

$2,221.00

Other Conferences/Meetings: $

F.2. Foreign Travel (FT): List below proposed trips individually and describe their purpose in relation to the grant. Also provide dates, destination, and number of travelers where known. Include total amount per sub category below next to ‘$’ and total amount requested for FT in column on the right.

Total $ Requested

Research Related (e.g. travel to research plots): $ $

Non-Research Related (i.e. professional meetings): $

G. Publications Costs/Page Charge: Provide below an estimated number of papers, total pages, and total cost.

Total $ Requested

Funds are being requested to cover the cost (page charges and reprints) of publishing at least two peer-reviewed papers in either Plant Disease or Phytopathology.

$2,000.00

H. Computer (ADPE) Services/Costs: Provide below the type of service and total cost. Total $

Requested $



I. Other Direct Costs (ODC): Under each relevant sub category below, enter a brief description, and basis for the estimate (i.e. individual fee rate/price). Include total amount per sub category below next to ‘$’ and total amount requested for ODC in column on the right.

Total $ Requested

Equipment/Facility/Land Rental and User Fees: $ $ Laboratory Animal Fees: $ Service/Maintenance Contracts: $ U.S. based Winter Nurseries: $ International Nurseries: $ Double Haploids: $ Other Analyses/Services: $ Communication (postage, shipping, fax, long distance phone): $ Photocopying: $ Sub Contracts: $ Tuition Remission: $ Other (describe): $

K. Indirect Costs (IDC): Provide below your Institution’s approved Indirect Cost (IDC) rate for USWBSI/USDA-ARS grants. Total $ for IDC 5% IDC $1,380.00

M. Small Business Act – SBIR Fee: The SBIR fee is a Congressional mandated fee charged to all ARS/USWBSI grants and is applicable to all non-ARS PIs. The rate for FY16 is 3.0% of total USWBSI recommended amount (Formula: Step 1 - USWBSI Recommended Amount/1.030; Step 2 - Result from Step 1 should be subtracted from USWBSI Recommended Amount to obtain the SBIR Fee).

SBIR Fee Amount

3.0% SBIR Fee $870.00



INDIVIDUAL PROJECT PROPOSAL BUDGET PAGE PROJECT 1: Efficacy and Curative Effects of Fungicides for FHB and DON Management in Ohio.

Instructions: Insert values from corresponding budget justification into this budget form. ORGANIZATION AND ADDRESS

Ohio State University Department of Plant Pathology 1680 Madison Ave. Wooster, OH 44691

USDA-ARS GRANT NO:

59-0206-4-018

Funds RequestedPRINCIPAL INVESTIGATOR:

Pierce Paul A. Salaries and Wages

1. PI(s)/PD(s) ................................................................................................................... $5,018

2. Other Professional Personnel (e.g. Post-Docs, Specialists and other professional personnel) .................................................................................................................... $12,307

3. Support Personnel (e.g. research technicians, students (graduate and undergraduate), secretarial-clerical and temporary employees) ....................................................................

Total Salaries and Wages .......................................................................... $17,325

B. Fringe Benefits (If charged as Direct Costs) ......................................................................... $6,051

C. Total Salaries, Wages, and Fringe Benefits (A plus B) ................................................. $23,376

D. Nonexpendable Equipment ............................................................................................

E. Materials and Supplies .................................................................................................... F. Travel

1. Domestic .....................................................................................................................

2. Foreign ........................................................................................................................

$2,221

G. Publication Costs/Page Charges ..................................................................................... $2,000

H. Computer (ADPE) Costs .................................................................................................

I. All Other Direct Costs ....................................................................................................

J. Total Direct Costs (C through I) ........................................................................................ $27,597K. Indirect Costs ....................................................................................................................................

Rate: 5% DC Base:

$1,380

L. ARS Award Amount - Total Direct and Indirect Costs (J plus K) ................................ $28,977

M. Small Business Act – SBIR Fee (3.0% for FY16) ..................................................................... ARS will deduct this amount prior to the award. N/A for ARS Scientists. $870

USWBSI FY16 Total Recommended Amount ................................................................................... $ 29,847

NAME AND TITLE (Type or Print) SIGNATURE DATE

Principal Investigator Pierce Paul

03/11/2016



PROJECT SUMMARY PAGE

Project Title 2: Functional Analysis for Getting Better Weather-based Predictors of Fusarium Head Blight. PROJECT SUMMARY

The overall project goal is to create better models for predicting Fusarium head blight (FHB). The objectives are to (i) identify periods within weather time series that are significantly different between FHB epidemics and non-epidemics, (ii) create variables summarizing those identified periods, (iii) use the summary variables in new logistic regression models for predicting FHB epidemics, (iv) compare the predictive performances of new models with the performances of the currently deployed models, and (v) replace the current models with the newer versions after they have been field-tested. The expected outcome is a more accurate FHB predictive system available on the Fusarium Head Blight Prediction Center (FHBPC) website. Towards this end, we have an existing data matrix of 865 observations of FHB epidemics and non-epidemics, linked to hourly weather time series (dew point, temperature, dew point depression, atmospheric pressure, relative humidity and vapor pressure deficit) spanning the September of year prior to anthesis through 60 days post-anthesis. Continued collaboration with the Integrated Management Coordinated Project (IM-CP) will further expand the data matrix so that as many growing conditions as possible are covered. We propose using an approach called functional data analysis to determine periods within each time series that are significantly different between epidemics and non-epidemics. These identified periods will be summarized (e.g. mean, number of hours above a threshold), and the resulting summary variables used as inputs to logistic regression models for predicting FHB epidemics. Logistic regression models are more amenable to scaling up for deployment on the FHBPC site. The new models will be compared against the current models in terms of predictive performance on the set of 865 observations, and field-tested during the 2017 growing season. We expect new models will be ready for use on the public site in time for the 2018 growing season. Growers, crop consultants, farm managers and other stakeholder interest groups will have more accurate forecasts of FHB epidemics, which will better inform the decisions they have to make related to FHB disease control and mitigation.

Principal and Co- Investigator(s):


Co-Investigator #1: Laurence V. Madden Institution: Ohio State University





Title Functional analysis for getting better weather-based predictors of Fusarium head blight Introduction Models for Fusarium head blight (FHB) are the driving force behind the Fusarium Head Blight Prediction Center (FHBPC; http://www.wheatscab.psu.edu/), which is used in 30 U.S. states where FHB and deoxynivalenol contamination are most damaging to wheat and barley. The models have evolved since originally published some 12 years ago (De Wolf et al. 2003). The underlying model engine is based on classical logistic regression, which scales well to the national level given that it is fed by a rapidly updated grid of weather-based predictors. Since 2003, the model search has been expanded to additive logistic regression (Shah et al. 2013) and boosted regression trees (Shah et al. 2014). These more advanced predictive algorithms yielded novel insights into the relationship between weather and FHB epidemics, which were incorporated into revised logistic regression equations more amenable for general usage within the FHBPC. The first models (De Wolf et al. 2003) were crafted with just 50 observations (unique site-year-variety combinations). Cooperation with the Integrated Management Coordinated Project (IM-CP) expanded the data set to 527 observations by 2009, and these latter data were used for the additive logistic regression and boosted regression tree efforts mentioned above (Shah et al. 2013, Shah et al. 2014). During FY14-15, our continued interaction with the IM-CP expanded the existing data set to 865 observations, which is a 64% increase from 2009. The significant expansion of the data set from 2003 through 2014 attests to the collaborative strength of the IM-CP. A notable advantage of the current data set is the inclusion of 85 observations from 2012 and another 69 observations from 2013. The 2012 and 2013 growing seasons were of particular interest because they represented striking deviations from the norm. The 2012 U.S. wheat growing season was one of the warmest on record, with wheat heading weeks earlier than normal in many states. By contrast, the 2013 growing season was cooler than normal, delaying wheat crop development in most areas of the Midwest. The FHB observations from 2012 and 2013 will expand the range of conditions used in model training, which should lead to more robust predictions. All models our research group has produced to date (De Wolf et al. 2003, Shah et al. 2013, Shah et al. 2014) have used weather-based predictors within a relatively narrow 30-day period: 15 days before anthesis to 15 days after anthesis. We had restricted the predictor space to this narrow range of the weather time series partly because earlier research had suggested good signals (i.e. predictors of FHB) existed close to anthesis (De Wolf et al. 2003, Moschini et al. 2001), and partly because collaborator-supplied weather data did not typically extend further than 15 days pre-anthesis. Later research showed that signals may exist much further out in the pre-anthesis period (Kriss et al. 2010). Predictors representing weather within the 30-day period were created by carving up the period into smaller windows (e.g. 5-, 7-, 10-, or 14-day for example). This ‘windowing’ approach is not without precedent (Coakley and Line 1982, Kriss et al. 2010), and was motivated by capturing the signal between weather and FHB epidemics within narrow time frames while filtering out the noise in the time series. There are some caveats to windowing, however. The start and end times of each window were set arbitrarily, and therefore windows do not necessarily coincide with the start or end of a signal. Hence a signal present in a weather time series may be ‘broken’ by the arbitrariness of the window definition (start, stop and window length). Another limitation was that windows were exactly the same for each of the basic weather variables (temperature, dew point, relative humidity), which assumed that signals associated with each of these variables are aligned temporally. We have re-created the weather time series (dew point, temperature, temperature-dew point depression, relative humidity, vapor pressure deficit) associated with each of the 865 observations by sourcing curated time series data through functions in the Mathematica software package (Wolfram



Research, Champaign, IL); Landschoot et al. (2013) used the same approach. Retrieving weather time series in this way gave us access to a wide range of geo-referenced university, federal, and citizen reporting stations, and removed the burden on collaborators for sourcing weather data files themselves. The new protocol also made for easier quality control, data cleaning and error checking on our end. We have added atmospheric pressure time series to the weather data matrix (a variable not considered previously; low pressure is often associated with rainfall). The new time series data span the growing season from the September of the year prior to anthesis (for winter wheat) to 60 days post-anthesis. In discovering relationships between weather and FHB epidemics, we propose moving away from the paradigm of carving up the weather time series into discrete windows. Each time series is really a set of sample points (e.g. hourly) from a smooth, underlying continuous process. That is, each time series (for temperature, dew point, relative humidity and pressure) can be represented as a smooth function. We propose to use functional data analysis (Ramsay and Silverman 2005) to model an entire time series in relation to FHB epidemics. Therefore, we will be moving away from the arbitrary discreteness of windows in finding signals. Our primary goal for FY16-17 is to use functional data analysis procedures to identify periods within each time series that are the most predictive of FHB epidemics. The objective is to create weather-based predictors (reflecting the identified periods) to be used in revised logistic regression models amenable for large-scale deployment via the FHBPC website. A secondary goal is to expand the data matrix by including observations and weather from 2014 onwards. The objective is to enlarge the observational space, so that models will be trained to recognize as wide a variety of conditions as possible that may be experienced. Rationale & Significance The predictive models on the FHBPC site are actively used by growers, farm managers and crop consultants. The forecasts and FHB Alerts stemming from the models influenced disease management decisions on 3 million acres of wheat and barley in 2014, according to a user survey (E. De Wolf unpublished). Users were asked to put a dollar value on the worth of the forecasts and alerts to their farm businesses. The median estimate was $15,000 per user, which translates to $65 million annually across all active users. Yet, the models are still regarded by the scientific and user communities as not yet possessing sufficient predictive power (Skelsey and Newton 2015). For example, a 07/06/2015 posting to the Scab Blog (www.scabusa.org/blog), giving an update from MN, pointed out that “The model is trending a little conservatively for the moderately susceptible varieties”. Users are calling for further improvements to the predictive models. This proposal falls within the FHB Management (MGMT) research area. Specifically, the proposed research is within the FY16-17 MGMT Research Categories research area of “the refinement and deployment of disease prediction and forecasting models, and disease management decision tools”. It addresses the FY16-17 MGMT Research Priority No. 2: “Help develop and validate the next generation of management and mitigation tools for FHB and mycotoxin control”. Research Materials and Methods Our primary question is: are there periods in the pre-anthesis and early post-anthesis weather time series associated with FHB epidemics? And if so, are they sufficiently distinct between epidemics and non-epidemics to be captured in summary variables useful for logistic regression? Our hypothesis is that such distinctions in the weather time series do exist, and that they are large enough to be captured by summary variables that distinguish FHB epidemics from non-epidemics. The current FHB observational matrix contains 629 non-epidemic and 236 epidemic cases; or put another way, about 27% of the observations are of epidemics. All resistance classes are represented: Very Susceptible (126), Susceptible (378), Moderately Susceptible (166), and Moderately Resistant



(195). As mentioned in the Introduction, the weather time series associated with each of the 865 observations has already been compiled and checked for integrity. This new weather data matrix covers a much longer preanthesis period (for winter wheat, all the way back to the September of the prior year) than the weather used in previous efforts (Shah et al. 2013, Shah et al. 2014). In summary, we already have at our disposal six weather time series associated with each of the 865 FHB observations. Ohio State University will continue the task of collecting and curating the FHB observational data generated by the IM-CP program. Data from the 2014 growing season has already been processed. OSU will also expand their efforts to incorporate data generated by the "risk based FHB managed program" proposed this year. The new protocols we have in place will be used to identify, source (via Mathematica) and perform integrity checks on weather time series data for each new added FHB observation. We propose to use functional data analysis (FDA) techniques to understand the relationship between weather and FHB epidemics (Ramsay and Silverman 2005). FDA is a rapidly growing field with many applications (Morris 2015). The premise is that a relatively noisy time series (such as daily weather) can be represented by a smooth underlying function. This functional representation of an entire time series is then used as a single variable. Therefore instead of treating each individual measurement in a time series as a variable, we use an entire curve as the variable! This makes sense; a variable such as temperature is continuous in time, but in reality one takes readings only at discrete intervals, say every hour. FDA reconstructs that underlying continuity, and then asks whether differences exist among curves. Figure 1 shows the daily temperature readings for observation 20 in the data set and a smooth functional representation of that data. Once we have a suitable functional representation of a time series, we can ask whether such series differ between epidemics and non-epidemics; or, if any differences are moderated by factors such as cultivar resistance. To illustrate, Figure 2 shows the mean relative humidity functional curves for FHB epidemics and non-epidemics, where the means are calculated from the respective individual functional curves. The mean curves do appear to differ. The next question is whether these curves are significantly different (given the underlying variability in the individual curves), and if so, where exactly along the time series do these significant differences occur? The separation between the curves beginning about 40 days pre-anthesis concurs with previous findings of pre-anthesis relative humidity as a predictor of FHB epidemics (De Wolf et al. 2003, Shah et al. 2013, Shah et al. 2014). There are some immediate novel insights gleaned from this simple visualization. Relative humidity is on average increasing from about 40 days pre-anthesis, for both epidemics and non-epidemics. However, relative humidity is, on average, higher throughout this 40-day period for epidemics (and continues into the post-anthesis period). These types of insights were not possible with the discrete ‘windowing’ approach used earlier (Shah et al. 2013, Shah et al. 2014). After we have identified significant regions of difference between epidemics and non-epidemics in a time series, the next step is identifying some sort of summary variable (e.g. mean, number of hours exceeding a threshold) capturing the signal in a significant region. We expect to identify several such summary predictor variables, which could cover periods of variable length (compared to the fixed windows used until now), and in pre-anthesis regions beyond the 15-day period used in our earlier work (De Wolf et al. 2003, Shah et al. 2013, Shah et al. 2014). These summary variables will be used in logistic regression models for predicting FHB epidemics. Logistic regression models scale well for prediction over large geographic areas. One could use the entire functional as a predictor, but that is not amenable to wide-scale deployment as required by the FHBPC. This is a limitation of the proposed functional approach. We propose using a number of open source R packages (e.g. fda, fda.usc, refund) with functions for accomplishing our stated goals (Morris 2015). A pitfall, in our experience, is that the quality of open source code and documentation varies considerably. We therefore adopted a skeptical approach, and chose to test all code with examples designed to (i) make sure code works without error, (ii) does what



the authors say it will do, and (iii) flag any functions that do not appear to work as intended. We have compiled an extensive code file documenting checks of all functions we intend to use, and are at the point of beginning to use the above-mentioned R packages with the FHB data matrix; Figures 1 and 2 are examples of exploratory functional analyses. Another possible pitfall lies in the weather data matrix compilation itself. We used GPS coordinates of the field trial sites to identify weather stations closest to the fields. Sometimes the closest station did not have data for the time range we needed. We were then forced to go to the next closest station. The further away a station is from the field site, the more ‘noise’ that is introduced into the time series, more so for temperature and dew point (temperature – dew point depression, relative humidity and vapor pressure deficit are all estimated from temperature and dew point. Atmospheric pressure is more forgiving of distance from a field site. There were gaps in some of the time series. We used imputation methods designed for time series to fill the gaps, but the imputation process introduces some additional noise in trying to reproduce the missing data. The tentative schedule is as follows: May 2016 – November 2016 (i) Functional analysis of the FHB data matrix: identify periods in weather time series which are significantly different between epidemics and non-epidemics. (ii) Translate significant differences into summary variables for input into logistic regression. (iii) Evaluate the predictive performance of newly formulated logistic regression models on the observational data matrix (865 observations). (iv) Measure the performance of new models against the models now in use on the public FHBPC site. (v) Draft manuscript No. 1 reporting the findings of the functional analysis of the weather time series. (vi) Process new FHB observational data from the IM-CP program. November 2016 – December 2016 (i) Finalize abstract and poster presentation for the Scab Forum 2016. (ii) PI attends and presents poster at Scab Forum 2016. January 2017 – April 2017 (i) Finalize and submit manuscript No. 1 (Target journal(s): Phytopathology or Plant Disease). (ii) Interface with Penn State programmers to have the newly formulated models available on the researcher part of the FHBPC site. (iii) Alert the researcher community to the new models. (iv) Link new FHB observational data to weather time series. April 2017 – August 2017 (i) Solicit feedback from the researcher community on how the new models performed in their respective states. (ii) Collect model performance data during the 2017 growing season: are new models performing better than the currently deployed models? (iii) Revise and resubmit manuscript No. 1. September 2017 – November 2017 (i) Critically assess the performance of the new models during the 2017 growing season. Revise the models if necessary. (ii) Prepare and submit Manuscript No. 2 documenting the new models and their predictive performance. November 2017 – December 2017 (i) Finalize abstract and poster presentation for the Scab Forum 2017. (ii) PI attends and presents poster at Scab Forum 2017. January 2018 – March 2018 (i) Revise and resubmit Manuscript No. 2. (ii) Interface with Penn State programmers to deploy revised logistic regression models on the public FHBPC site for the 2018 growing season.



Figure 1. Daily mean temperature and a fitted smooth functional representation for observation 20.

Figure 2. Smooth functional representations of the mean daily temperature for FHB epidemics and non-epidemics. The vertical green line references 40 days pre-anthesis. References to Project Description Coakley, S.M., and Line, R.F. 1982. Prediction of stripe rust epidemics on winter wheat using statistical models. (Abstr.) Phytopathology 72:1006. De Wolf, E.D., Madden, L.V., and Lipps, P.E. 2003. Risk assessment models for wheat Fusarium head blight epidemics based on within-season weather data. Phytopathology 93:428-435. Kriss, A.B., Paul, P.A., and Madden, L.V. 2010. Relationship between yearly fluctuations in Fusarium head blight intensity and environmental variables: A window-pane analysis. Phytopathology 100:784-797.



Landschoot, S., Waegeman, W., Audenaert, K., Haesaert, G., and De Baets. B. 2013. Ordinal regression models for predicting deoxynivalenol in winter wheat. Plant Pathology 62: 1319-1329. Morris, J.S. 2015. Functional regression. Annual Review of Statistics and Its Application 2: 321-359. Moschini, R.C., Pioli, R., Carmona, M., and Sacchi, O. 2001. Empirical predictions of wheat head blight in the Northern Argentinean pampas region. Crop Science 41: 1541-1545. Ramsay, J.O. and Silverman, B.W. 2005. Functional Data Analysis. Second Edition. Springer New York, NY. Shah, D.A., Molineros, J.E., Paul, P.A., Willyerd, K.T., Madden, L.V., and De Wolf, E.D. 2013. Predicting Fusarium head blight epidemics with weather-driven pre- and post-anthesis logistic regression models. Phytopathology 103:906-919. Shah, D.A., De Wolf, E.D., Paul, P.A., and Madden, L.V. 2014. Predicting Fusarium head blight epidemics with boosted regression trees. Phytopathology 104:702-714. Skelsey P, and Newton A.C. 2015. Future environmental and geographic risks of Fusarium head blight of wheat in Scotland. European Journal of Plant Pathology 142: 133-47.




Project 2 Title: Functional Analysis for Getting Better Weather-based Predictors of Fusarium Head Blight.






Category PI(s)/PD(s):

$


$7,354

Post Doc: Partial Support for post-doc at OSU (0.20)

$7,354


$


$


$


$


$




$


$2,743

Post Doc: Fringe Rate 37.3%

$2,743


$


$


$


$


$




Total $ Requested

$


Total $ Requested

Field:

$ $

Greenhouse:

$

Laboratory:

$

Other:

$


Total $ Requested


Non-Research Related (i.e. professional meetings): FHB Forum: $ Other Conferences/Meetings: $


Total $ Requested




Total $ Requested

$


Requested $




Total $ Requested


K. Indirect Costs (IDC): Provide below your Institution’s approved Indirect Cost (IDC) rate for USWBSI/USDA-ARS grants. Total $ for IDC 5% IDC $505


SBIR Fee Amount

$318



INDIVIDUAL PROJECT PROPOSAL BUDGET PAGE

PROJECT 2: Functional Analysis for Getting Better Weather-based Predictors of Fusarium Head Blight. Instructions: Insert values from corresponding budget justification into this budget form. ORGANIZATION AND ADDRESS


USDA-ARS GRANT NO:

59-0206-4-018



1. PI(s)/PD(s) ...................................................................................................................

2. Other Professional Personnel (e.g. Post-Docs, Specialists and other professional personnel) .................................................................................................................... $7,354

3. Support Personnel (e.g. research technicians, students (graduate and undergraduate), secretarial-clerical and temporary employees) ....................................................................




D. Nonexpendable Equipment .............................................................................................

E. Materials and Supplies ....................................................................................................

F. Travel 1. Domestic .....................................................................................................................

2. Foreign .........................................................................................................................

G. Publication Costs/Page Charges .....................................................................................



J. Total Direct Costs (C through I) ........................................................................................ $10,097K. Indirect Costs .....................................................................................................................................

Rate: 5% DC Base: 10,097

$505


M. Small Business Act – SBIR Fee (3.0% for FY16) ..................................................................... ARS will deduct this amount prior to the award. NA for ARS Scientists. $318




03/11/2016



PROJECT SUMMARY PAGE Principal and Co- Investigator(s):


Project Title 3: Risk-based Fungicide Decision-making for FHB and DON Management in Wheat. PROJECT SUMMARY

The overall goal of this project is to facilitate the practical utilization of the web-based FHB risk assessment system for fungicide application decision-making. We propose to accomplish this through the following specific objectives:

1. Evaluate criteria for using the web-based risk assessment tool to make fungicide application decisions for FHB management.

2. Develop risk-based fungicide application guidelines for FHB management. To accomplish these objectives, data will be collected at multiple locations in Ohio as well as in other states through collaborations with Drs. Bowen at Auburn University, Chilvers at Michigan State University, Collins at The Pennsylvania State University, Cowger at USDA-ARS North Carolina State University, Smith at the University of Wisconsin-Madison, Bradley at the University of Kentucky, Wegulo at the University of Nebraska-Lincoln, Kelly at the University of Tennessee, Friskop at North Dakota State University, Mehl at Virginia Tech Tidewater AREC, Byamukama at South Dakota State University, Wise at Purdue University, Dill-Macky at the University of Minnesota, Kleczewski at University of Delaware, and Darby at the University of Vermont. Based on the population of studies across locations, a wide range of environments at anthesis is anticipated, and thus a range of predicted FHB risk scenarios. We anticipate that fungicide efficacy will vary among risk scenarios based on whether the application is made when risk is low, moderate, high, or consistently moderate-high. PIs will plant strips of wheat cultivars with different levels of FHB resistance on university research farms or in farmers’ fields in 17 states, and half of each strip will be treated with Prosaro at 6.5 fl. oz/acre at early anthesis and the other half left untreated. Using on the web-based risk tool, the risk of FHB will be determined at the time of each application, and each cultivar x flowering date x location combination will be assigned a code (A, B, C, or D) based on the predicted risk. If the risk of FHB is low when the fungicide is applied, the treatment will be coded as A (calendar-based application), B (moderate-risk based application) if the risk is moderate, C (high-risk based application) if the risk is high, and D (sustained-risk based application) if the risk is moderate-high on consecutive days leading up to anthesis. Percent FHB/FDK/DON control will be estimated for each cultivar-location and then averaged across predicted risk scenarios to identify the scenario with the highest overall mean efficacy. Using data from sprayed and unsprayed strips, separate true positive proportion (fungicide treatment under predicted risk scenario B or C or D), true negative proportion (no fungicide application under scenario A), false positive proportion (fungicide application under scenario A), and false negative proportion (no fungicide application under scenario B or C or D) will be estimated. The advantage of using the tool under each risk scenario, relative to never spraying or always spraying will be quantified. Results from the analyses will be used to identify risk scenarios under which an anthesis application of Prosaro is most effective and advantageous, and to develop specific criteria and guidelines for using the risk tool to make fungicide application decisions.

Co-Investigator #1: Laurence V. Madden Institution: Ohio State University Co-Investigator #2: Jorge David Salgado Institution: Ohio State University



PROJECT DESCRIPTION Title: Risk-based Fungicide Decision-making for FHB and DON Management in Wheat Introduction: The sporadic nature of FHB and its long latent period (defined here as the time it takes for symptoms to develop following infection) under field conditions make the task of deciding whether a fungicide application is warranted very difficult. Unlike many foliar diseases of wheat for which disease levels of the upper two leaves can be used as action thresholds for management decision-making (Paveley et al. 1997), FHB intensity cannot be used to help guide fungicide applications. Although infections occur primarily during anthesis (Anderson 1948), characteristic symptoms of the disease only develop during early-soft dough, approximately 18-21 days after infection. By the time symptoms are observed, it is too late to effectively control FHB and its associated mycotoxins, since the most effective fungicides are used as protectants and as such need to be applied before infection occurs. Moreover, due to pre-harvest fungicide restrictions, it is often illegal to apply fungicides during the soft dough growth stage. Consequently, producers have to make fungicide application decisions at anthesis, well before knowing whether scab will develop and grain will be contaminated with mycotoxins. An unwarranted application results in added production costs that reduce profit, particularly if grain prices are low, fungicide application cost is high, and scab and DON end up not being a concern. Conversely, failure to apply a fungicide when it is needed results is substantial grain yield and quality losses. Uncertainties as to whether FHB will develop, whether or not a fungicide should be applied for FHB, and whether it will be effective and economical when applied are potentials barriers to the adoption of this strategy for FHB management (McRoberts et al 2011, Gent et al 2011). In an effort to minimize some of these uncertainties, and to provide producers and crop advisors with a tool to help guide fungicide applications, an FHB risk prediction system was developed more than a decade ago (De Wolf et al 2003) and has been available for use as part of FHB management programs in over 30 US states where FHB is a constant threat (http://www.wheatscab.psu.edu). Thanks in part to data generated through the USWBSI-funded integrated management coordinated project, this tools has since been validated, refined, and upgraded based on more contemporary modeling approaches (Shah et al 2013, 2015). The most recent generation of models was deployed in 2015. Using the risk tool, stakeholders can simply select their state, wheat market class, cultivar resistant class, and flowering data, and the risk of FHB exceeding 10% severity (considered an epidemic) is estimated and depicted on the state map, with red indicating high risk, yellow moderate risk, and green low risk. These predictions are based on weather conditions during a 7-day window immediately before anthesis and the susceptibility of the cultivar to FHB. The tool correctly predicts the risk of scab with an accuracy of about 75%. Research continues to improve the models and increase the overall accuracy. However, there is also a layer of uncertainty associated with the practical utilization of the risk tool to make fungicide application decisions. For instance, common scenarios and questions presented by stakeholders, and even researchers, include: 1) “wheat is flowering and the map is yellow (moderate-risk predicted), should a fungicide be applied (or recommend, in the case of crop advisors) or should applications be made only when the map is red (high-risk predicted)?” 2) “wheat is flowering and the map is green (low-risk prediction), but it rained yesterday and it’s currently raining, should a fungicide be applied in case it continues to rain over the next few days?” To date, efforts to validate the risk tool have focused primarily on determining its prediction accuracy when presented with new observations (that were not used for model development), particularly from locations with known epidemics. However, less attention has been devoted to validating the system in terms of its benefit as a tool for making fungicide application decisions. We propose here to gather data from fungicide treated and untreated strips/plots of moderately resistant, moderately susceptible, and susceptible cultivars planted at more than 35 locations in 17 states with the object of 1) evaluating criteria for using the web-based risk assessment tool to make fungicide application decisions for FHB management and 2) developing risk-based fungicide application guidelines for FHB management.



Rationale and Significance: For reasons discussed above, the web-based risk tool is an important component of FHB management programs, and results from a recent survey of users of the system attest to its importance for decision-making. Together the risk tool and it associated alert system influenced FHB management decisions on 3 million acres of wheat and barley in 2014 (E. De Wolf personal communication). According to the survey, the median dollar value of the system was estimated at $15,000 per user. As efforts to refine the system and improve its accuracy continue, its perceived value and usefulness will likely continue to increase. However, the lack of specific criteria or guidelines for practical utilization of the system to make fungicide application decisions constitutes a knowledge gap that could affect the wellness of stakeholders to utilize the tool. For some disease forecasting systems, some cumulative measure of risk such as disease severity value (DSV, Madden et al 1978; Pitblado 1992) is often used as the trigger for deciding when to apply a fungicide, and once some pre-established DSV threshold is reached, the fungicide is applied. For the FHB risk tool, there is no such clear-cut threshold, instead risk is measured in terms of the probability of FHB index being ≥ 10%. Therefore, the red, yellow, and green colors on the risk map simple mean that the probability of index exceeding 10% is high, moderate and low, but in all cases, even when the map is green, FHB may still develop. The difference between red and yellow or yellow and green could simply result from the change of one or two favorable days in the 7-day pre-anthesis window used to predict risk and assign colors to the map (Fig 1). One may argue that users of the risk tool should, and we as extension specialists do recommend that they do, explore the system by selecting multiple flowering dates around the time of anthesis as a way of minimizing overreliance on a single-day risk prediction and determining whether the risk is consistently high, low, or moderate over multiple days. The Model Basics web link on the FHB prediction center (which most users likely never read) also advises against relying solely on the risk tool to make application decisions by stating that “The models are only one source of information available to help make management decisions. We strongly encourage you to consult with local extension specialists, and crop consultants to determine if fungicide applications are needed to suppress Fusarium head blight in your area”. Indeed, state specialists often post commentaries on the risk tool based on their experience and multiple sources of information, including model predictions, to help with decision-making. In reality, a user’s decision to apply (or recommend) a fungicide when the map is green, yellow, or red depends on several factors other than predictions made by the risk tool, including his or her perception of risk and level of risk aversion, the value of the wheat crop, and fungicide application cost (Gent et al. 2011; McRoberts et al 2011). However, we believe that a more formal evaluation of the cost-benefit of fungicide application decisions made under a range of risk scenarios will generate data invaluable for the decision-making process. This will be of great value to both those who use and recommend the system as a decision-making tool. However, such an evaluation would require the collection of data from treated and untreated fields/plots under a wide range of environments at anthesis, and consequently, a wide range of risk levels. Unfortunately, the available data from the USWBSI-funded Uniform Fungicide and Integrated Management Trials cannot be used for this purpose as most of those were collected from mist-irrigated and/or artificially inoculated experiments. The risk tool was developed to predict risk under natural not artificial conditions, so it would not be useful to evaluate it using data from systems designed to favor or force FHB development and DON contamination. The proposed work is of direct relevance to MGMT Research Priority 2: “Help develop and validate the next generation of management and mitigation tools for FHB and mycotoxin control.” Research Material and Methods: Hypothesis: The overall efficacy and benefit of applying Prosaro at anthesis for FHB and DON management will be influenced by the predicted risk scenario under which the application is made.



Description and sequence of proposed experiments: Each PI will plant strips/plots of at least 2 wheat cultivars (one moderately resistant and another susceptible) at 1 to 5 locations across his/her state (Table 1). Strips will be established on university research farms or in farmers’ fields, and managed according to standard agronomic practices for each location. The dimensions of the strips/plots may vary among locations depending on available space, but should be at least 5 ft x 20 ft. Half of each strip of each cultivar will be treated with Prosaro at 6.5 fl. oz/acre at early anthesis (Feekes 10.5.1) and the other half will be left untreated. Applications will be made using a sprayer equipped with paired Twinjet or flat fan XR8001 or XR8002 nozzles, mounted at an angle (30-45o from the horizontal) forward and backward (or forward only) and calibrated to deliver at a rate of 10 to 20 gallons per acre. FHB risk will be evaluated at the time of each application, and each cultivar x flowering date x location combination will be assigned a code (A, B, C, or D) based on the predicted risk of FHB (Table 2, Fig 2). A single application will be made to each cultivar, but the calendar date of the application will depend on the maturity of the cultivar, and the code assigned to the treatment will be a function of local weather and cultivar susceptibility (key FHB risk factors/predictors). For instance, if the risk of FHB is low (the risk map is green) when the fungicide is applied, the treatment will be coded as A (calendar-based application); if the risk is moderate (map is yellow), the treatment will be coded as B (moderate-risk based application); if the risk is high (map is red), the treatment will be coded as C (high-risk based application); and D (sustained moderate-high-risk based application) if the risk is moderate-high on consecutive days leading up to anthesis. Since the risk of FHB (consequently, the code assigned to each treatment) will be evaluated separately for each cultivar at each location based on susceptibility and maturity/flowering date, applications made on the same calendar date may end up receiving different treatment codes, and conversely, applications made on different calendar dates may end up receiving the same treatment code. This will allow us to obtain replicates of treatment categories A, B, C, and D when the data is pooled. For instance, the hypothetical example in Figure 2 shows three scenarios in which the same susceptible cultivar (Fig 2A, C, and D) will be assigned treatment codes of A, B and C based on the fact that it was planted in different regions (under different FHB-influencing environment conditions) and/or reached anthesis on different dates, causing the risk of FHB to be low, moderate, and high. FHB incidence and severity will be rated on 100-200 spikes per plot/strip (10-20 arbitrarily-selected clusters of 10 spikes spread out across the strip) at the soft dough growth stage (Feekes 11.2). The presence and flag leaf severity (as a percentage) of foliar diseases will also be rated. For those strips established in fields previously planted with a Fusarium graminearum host crop, percent residue on the soil surface will be estimated. The flowering date of each cultivar and GPS coordinates of each location will be recorded. Plots will be harvested and grain yield and test weight determined. Two subsamples of grain from each plot will be pulled - one will be rated to determine percent Fusarium damaged kernels (FDK), and the other will be sent to one of the USWBSI-funded mycotoxin Testing Laboratories for DON analysis. Campbell Scientific units (Campbell Sci., Logan UT) or other onsite weather stations will be used to collect temperature, relative humidity, surface wetness, rainfall, wind speed, and solar radiation data at regular intervals from Feekes GS 7 (stem elongation) to harvest. Results Expected: Based on the population of studies across and within states, a wide range of environments at anthesis is anticipated, and thus a wide range of risk levels for FHB and corresponding FHB and DON levels. We anticipate that fungicide efficacy (percent control relative to the untreated check) will vary among risk scenario A, B, C and D (and possibly others as determined by the data). Data Analysis: Percent control of FHB index, FDK, and DON will be estimated for each cultivar x location combination and then averaged across predicted risk scenarios to determine which scenario results in the highest overall mean efficacy. In addition, using the data from sprayed and unsprayed plots for the different cultivars x location combination, the so-called: true positive proportion (TPP; fungicide treatment when predictor indicates high risk; scenario C); true negative proportion (TNP; no fungicide application and the predictor indicates low risk; scenario A); false positive proportion (FPP; fungicide application and the predictor indicates low risk); and



false negative proportion (FNP; no fungicide application when the predictor indicates high risk) will be estimated (McRoberts et al. 2011). Subcategories will also be created for the levels of, and duration of, risk such as described under scenarios B (moderate risk) and D (sustained moderate-high risk) above. Assuming that the cost (i.e., impact) of using the risk predictor is, in part, proportional to disease level (Paul et al 2006, 2009). We will then use the expected “cost” (impact) equation of Biggerstaff (2000) to quantify the average impact of using the predictor under a range of risk conditions (as defined by the prevalence of epidemics in any given area). Then, using the skill function of Briggs and Zeretzki (2008), we will quantify the relative advantage (or disadvantage) of using the predictor for different risk scenarios (B, C, D and possibly others as defined by the data), relative to never spraying or always spraying at anthesis (A). Application of results/technology transfer: Previous evaluations of the FHB risk tool were based on retroactive assessment of disease under different risk prediction levels (McRoberts et al. 2011; Shah et al. 2014). This will be the first proactive evaluation, with fungicide application (or not) linked to the predicted risk. Results from the analyses described above will be used to identify risk scenarios under which an anthesis application of Prosaro is most effective and advantageous. This information will then be used to develop more specific guidelines for using the risk tool to make fungicide application decisions. Results will be presented at extension and scientific meetings, and published in peer-reviewed journals. In addition, data from the untreated checks will be added to the database available for developing and validating FHB risk models. Possible pitfalls and limitations: The biggest concern with the proposed research is the inability to reproduce the different risk scenarios due to highly favorable or unfavorable weather conditions. However, this is unlikely to be the case across all locations. Moreover, having cultivars with different levels of resistance to FHB at each location will increase the likelihood of us seeing a range of predicted risks, since cultivar resistance is one on the main input variables used in the risk model. Timeline of Proposed Events: Winter Wheat Field Experiments

o Plant plots/strips Fall 2015 and 2016 o Treatment application and disease assessment Spring 2016 and 2017 o Harvest and FDK and DON analyses Summer 2016 and 2017

Spring Wheat Field Experiments o Plant plots/strips Spring 2016 and 2017 o Treatment application and disease assessment Summer 2016 and 2017 o Harvest and FDK and DON analyses Summer 2016 and 2017

Collect data from collaborators and assign risk categories Fall 2016 and 2017 Data analysis and preparation of reports Fall 2016 and 2017 Develop risk-based fungicide guidelines and manuscript preparation Spring 2018 References to Project Description: 1. Biggerstaff, B.J. 2000. Comparing diagnostic tests: A simple graphic using likelihood ratios. Statistics in

Medicine 19: 649-663. 2. Briggs, W.M., and Zaretzki, R. 2008. The skill plot: A graphical technique for evaluating continuous

diagnostic test. Biometrics 54: 250-256. 3. De Wolf, E.D., Madden, L.V., and Lipps, P.E. 2003. Risk assessment models for wheat Fusarium head blight

epidemics based on within-season weather data. Phytopathology 93:428-435. 4. Gent, D. H., De Wolf, E., and Pethybridge, S. J. 2011. Perceptions of risk, risk aversion, and barriers to

adoption of decision support systems and integrated pest management: An introduction. Phytopathology 101:640-643.

5. McRoberts, N., Hall, C., Madden, L.V., and Hughes, G. 2011. Perceptions of disease risk: From social construction of subjective judgments to rational decision making. Phytopathology 101: 654-665.



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

7. Paul, P. A., McMullen, M. P., Hershman, 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.

8. Paveley, N. D., Lockley, K. D., Sylvester-Bradley, R., and Thomas, J. 1997. Determinants of fungicide spray decisions. Pestic Sci. 49:379-388.

9. Shah, D.A., Molineros, J.E., Paul, P.A., Willyerd, K.T., Madden, L.V., and De Wolf, E.D. 2013. Predicting Fusarium head blight epidemics with weather-driven pre- and post-anthesis logistic regression models. Phytopathology 103:906-919.

10. Shah, D.A., De Wolf, E.D., Paul, P.A., and Madden, L.V. 2014. Predicting Fusarium head blight epidemics with boosted regression trees. Phytopathology 104:702-714.

11. Pitblado, R. E. 1992. The development and implementation of TOM-CAST: A weather-timed fungicide spray program for field tomatoes. Ministry of Agriculture and Food, Ontario, Canada.

12. Madden L., Pennypacker, S. P., and McNab, A. A. 1978. FAST, a forecast system for Alternaria solani on tomato. Phytopathology 68:1354-1358.

Facilities and Equipment: All PIs have access to the research farm space and equipment (e.g. sprayers, plot planters, combine harvesters, grain grinding apparatus etc) necessary to conducted the research proposed herein. Collaborative Agreements: The proposed research does not require contractual agreements among PIs. Drs. Bowen at Auburn University, Chilvers at Michigan State University, Collins at The Pennsylvania State University, Cowger at USDA-ARS North Carolina State University, Smith at the University of Wisconsin-Madison, Bradley at the University of Kentucky, Wegulo at the University of Nebraska-Lincoln, Kelly at the University of Tennessee, Friskop at North Dakota State University, Mehl at Virginia Tech Tidewater AREC, Byamukama at South Dakota State University, Wise at Purdue University, Dill-Macky at the University of Minnesota, Kleczewski at University of Delaware, and Darby at the University of Vermont have all agreed to conduct the experiment described above (see letters below).



Table 1. List of participating states, locations/regions within states, and cultivars from which data will be collected State Locations (Counties) Cultivars (FHB Resistance, Maturity) Alabama Autauga (central), Baldwin (south), and

Limestone (north) Jamestown (MR, medium), SS 8641 (S, medium)

Delaware Newcastle Jamestown (MR) and Shirley (S) Indiana West Lafayette and Vincennes (2 reps), IN Malabar (MR, mid-season), Truman (MR, late) Hopewell (S, mid-season),

Bromfield (MS, mid-season) Kentucky/Illinois Western, KY; Southern, IL P25R32 (MS), BW5530 (MR), BW5228 (MR), P25R47 (S) Michigan East Lansing and Sanillac Co. Ambassador (S) and a moderately resistant cultivar TBD Minnesota Strathcona - northern Red River Valley,

Crookston - central Red River Valley, Fergus Falls - southern Red River Valley, Saint Paul - southern MN

Samson (FHB - 8, susceptible; DTH- 57.8, mid-season), Linkert (FHB - 5, moderately resistant; DTH - 58.2 mid-season), WB-Mayville (FHB - 7, moderately susceptible; DTH - 56.4 early), Faller (FHB - 4 moderately resistant; DTH - 61.8 late)

Nebraska Mead Everest (MR), McGill (S), Overley (S), and Overland (MR) North Carolina Wake NC-Yadkin (MR, late), USG 3438 (MS, late), and P26R20 (S, late) North Dakota Fargo, ND and a second location WB Mayville (S), Glenn (MR), Barlow (M) Ohio Wayne, Darke, Crawford, Wood,

Pickaway, and Clark Malabar (MR, mid-season), Bromfield (MS, mid-season), Truman (MR, late) Hopewell (S, mid-season), Cooper (S, early) on two planting dates

Pennsylvania Lancaster, Centre Malabar (MR, mid-season), Bromfield (MS, mid-season), Hopewell (S, mid-season)

South Dakota Brookings and Codington Grainfield (MR) Expedition (MR) Wesley (S) Tennessee Western TN Madison and Gibson,

possibly Middle TN – Robertson P25R32 (MS), BW5530 (MR)

Vermont Grand Isle, Vermont Co. Malabar (MR, mid-season), Truman (MR, late) Hopewell (S, mid-season), Bromfield (MS, mid-season)

Virginia Suffolk, Warsaw (Richmond Co.), and TBD

Shirley (S, late), Jamestown (MR, early)

Wisconsin Arlington, WI One MR and one S cultivar



Table 2 - Fusarium head blight risk-based Prosaro fungicide application programsa

Treatment code and description

Core

1. Untreated regardless of the risk prediction

2. Application of Prosaro at 6.5 fl oz/A + NIS at 0.125% at anthesis

A. Regardless of the risk prediction

B. If the risk is moderate (the risk map is yellow)

C. If the risk is high (the risk map is red) D. If the risk is moderate or high on two or more consecutive days immediately before the selected anthesis

date

Optional

3. At anthesis if the risk is moderate or high followed by a second application four days later if the risk continues to be moderate or high

aWith the data, it may be possible to associate other risk scenarios/patterns (such as the number of post-anthesis high-risk days, for instance) associated with fungicide efficacy (or lack thereof) against DON.

Fig. 1. Maps of Ohio showing the predicted risk of Fusarium head blight for two hypothetical scenarios A, high risk on a susceptible wheat cultivar planted in central or southern Wayne Co that reaching anthesis on June 10 and B, moderate risk on the same cultivar in the same region that reach anthesis two days later (June 12). Note (bar graphs): two high-RH days (81 and 78%) moved out of the 7-day pre-anthesis risk prediction window between June 10 and 12.



Fig. 2. Maps of Ohio showing the predicted risk of Fusarium head blight for 4 hypothetical scenarios A, high risk on a susceptible cultivar planted in north western Wayne Co and reaching anthesis on June 8 (treatment coded as C); B, low risk on a moderately resistant cultivar, planted in the same area and reaching anthesis on the same date (treatment coded as A); C, moderate risk on the same susceptible cultivar, reaching anthesis on the same date, but planted in central Hancock Co (treatment coded as B); and D, low risk on the same susceptible cultivar, reaching anthesis three days earlier (June 5) in Fayette Co (treatment coded as A).




Project 3 Title: Risk-based Fungicide Decision-making for FHB and DON Management in Wheat.







$


$

Post Doc:

$


$

Support Personnel (research technicians, students (graduate and undergraduate), and temporary employees): $11,760 Research Technician(s):

$

Graduate Student(s): Support stipend (50%) for a graduate student who, under the supervision of the PIs, will be responsible for conducting experiments at locations in Ohio, collect and analyze data from all locations, and prepare poster, reports, and manuscripts.

$11,760


$


$




$


$

Post Doc:

$


$

Support Personnel (research technicians, students (graduate and undergraduate), and temporary employees): $1,348 Research Technician(s):

$

Graduate Student(s): Benefits for graduate student at 11.4% of the salary.

$1,348


$


$




Total $ Requested

$


Total $ Requested

Field:

$ $

Greenhouse:

$

Laboratory:

$

Other:

$


Total $ Requested

Research Related (e.g. travel to research plots): Funds are being requested to help cover the cost of travel to research plots to apply treatments and collect data.

$580 $580

Non-Research Related (i.e. professional meetings): FHB Forum: $ Other Conferences/Meetings: $


Total $ Requested




Total $ Requested

$


Requested $




Total $ Requested


K. Indirect Costs (IDC): Provide below your Institution’s approved Indirect Cost (IDC) rate for USWBSI/USDA-ARS grants. Total $ for IDC 5% IDC $684


SBIR Fee Amount

3% SBIR fee $431



INDIVIDUAL PROJECT PROPOSAL BUDGET PAGE PROJECT 3: Risk-based Fungicide Decision-making for FHB and DON Management in Wheat.

Instructions: Insert values from corresponding budget justification into this budget form. ORGANIZATION AND ADDRESS


USDA-ARS GRANT NO:

59-0206-4-018



1. PI(s)/PD(s) ...................................................................................................................

2. Other Professional Personnel (e.g. Post-Docs, Specialists and other professional personnel) ....................................................................................................................

3. Support Personnel (e.g. research technicians, students (graduate and undergraduate), secretarial-clerical and temporary employees) .................................................................... $11,760




D. Nonexpendable Equipment .............................................................................................

E. Materials and Supplies .................................................................................................... F. Travel

1. Domestic .....................................................................................................................

2. Foreign .........................................................................................................................

$580

G. Publication Costs/Page Charges .....................................................................................



J. Total Direct Costs (C through I) ........................................................................................ $13,688K. Indirect Costs .....................................................................................................................................

Rate: 5% DC Base: 13,688

$685


M. Small Business Act – SBIR Fee (3.0% for FY16) ..................................................................... ARS will deduct this amount prior to the award. N/A for ARS Scientists. $431




03/11/2016



ADJUSTMENT SUMMARY PAGE



Institution: Ohio State University

Fiscal Year: 2016 USWBSI’s FY16 Total Recommended Amount: $ 55,571

ARS Grant Number: 59-0206-4-018

Instructions: Under each of the project titles listed below, please indicate the changes made to the pre-proposal that address the comments given by the Review Panel(s) and/or the Executive Committee (see table in “Letter of Instructions”). Use additional pages if necessary. Project Title 1: Efficacy and Curative Effects of Fungicides for FHB and DON Management in Ohio. None Project Title 2: Functional Analysis for Getting Better Weather-based Predictors of Fusarium Head Blight. None Project Title 3: Risk-based Fungicide Decision-making for FHB and DON Management in Wheat. None


I agree to send copies of any printed materials (e.g. brochures, extension publications, etc.) and/or electronic versions of communication materials or URL links to materials posted on the Web to the Networking & Facilitation Office of the U.S. Wheat & Barley Scab Initiative. ____________________________ 03/11/2016__ Principal Investigator Date


COMMUNICATION PLAN



Fiscal Year: 2016 USWBSI’s FY16 Total Recommended Amount: $ 55,571

ARS Grant Number: 59-0206-4-018

Instructions: Using the space below, describe in detail how you plan to communicate the results from this research to your stake-holders in the most effective way. Please describe your target audience (i.e. USWBSI Administration/members, industry, private growers, interest groups etc.) and the methods (i.e. written, electronic, oral, etc.) of communication you will use to communicate your results to your audience.

As is customary, results from the integrated management studies will be published in the annual proceedings of the USWBSI and on the USWBSI website. Results also will be delivered to researchers, growers, dealerships, county extension educators and others in the wheat industry by way of electronic newsletters, commentaries on the FHB forecasting website and Scab Alert, peer-reviewed publications, extension talks, and FHB management workshops. Results from the study designed to validate the forecasting system will be presented as posters and talks to researchers, stakeholders and students at the Scab Forum and at the annual meeting of the American Phytopathological Society. These finding will ultimately be published in peer-reviewed journals. Data generated through these studies will also be used to develop specific criteria and guidelines for using the risk tool to make fungicide application decisions. Results from the modeling effort will be used to update and refine the forecasting system, which continues to be delivered to more than 25 U.S. states. All improvements made to the existing system and models developed as a result of this project will be communicated to researches and others in the wheat industry by way of peer reviewed journal articles, UWBSI proceeding articles, presentations and posters at meetings, and extension workshops.

Documents

FY16 USWBSI RECOMMENDED PROJECT(S) PROPOSAL · FY16 USWBSI RECOMMENDED PROJECT(S) PROPOSAL NOTE: This document will be retained in the USWBSI’s NFO COVER PAGE Instructions: Update