| For 2020 and newer grants, please go to https://grants.ipmcenters.org/ |
|---|
|
|
| Home Current RFAs PD User Guide Projects Login |
|
Funded Project |
|
Funding Program:
Regional IPM Competitive Grants - Northeastern |
|
Project Title:
Predicting Inoculum Availability for Peach Scab: Development and Validation of a Forecasting Model |
Project Director (PD):
|
|
Lead State: NJ Lead Organization: Rutgers University |
| Research Funding: $20,000 |
|
Start Date: Jun-15-2007 End Date: Jun-14-2010 |
|
Pests Involved: scab, fungus, fungi |
|
Site/Commodity: peaches, nectarines |
|
Area of Emphasis: modeling, forecasting |
|
Summary:
This research project investigates the quantitative epidemiology of peach scab, caused by the plant pathogenic fungus Fusicladosporium carpophilum. Eastern states produce 62% of the total U.S. fresh production of peaches with an annual value of $179 million. Crop profiles for most peach growing states in this region, including NJ, PA, WV, and New England, list scab as a disease of major importance. Since resistant cultivars are not available and cultural controls alone are inadequate, scab is primarily managed by application of consecutive fungicide sprays on a calendar basis. This approach results in unnecessary fungicide applications when environmental conditions are unfavorable and loss of disease control when conditions are highly favorable. The New Jersey Pest Management Strategic Plan states "a better understanding of the epidemiology should allow more effective and efficient use of newer fungicides". The major goal of the proposed project is to develop a forecasting model for predicting inoculum availability for infection. Specifically, sporulation of overwintering twig lesions, the major source of inoculum, will be quantitatively described as a function of temperature. A forecasting model algorithm will be created from this temperature relationship and previously published results. Predictions of this model will be field validated over two seasons. Implementation of the model will allow optimized fungicide application timing, thereby simultaneously reducing the potential for fungicide overuse and the likelihood of yield loss. Improvements in fungicide application efficiency will decrease amount of fungicide in the environment, enhance grower profitability, reduce applicator and field crew pesticide exposure, and decrease risk of pesticide residues on harvested fruit.
Objectives: 1. Model Development. The proper creation, validation, and deployment of a commercial disease forecasting model requires successful completion of many steps conducted over several years. Since we are at the beginning of this process, the major goal of this project proposal is to create the knowledge base necessary for the initial development of a peach scab forecasting model. Specifically, there are three sequential objectives for the proposed research project: (1) Quantitatively describe the temperature-sporulation relationship (2) Create model algorithm based on new and currently available information (3) Field validate the biological criteria for the model, i.e., the sporulation predictions Proposal Progress Report USDA CRIS data |
|
Final Report: |
|
Outcomes Treatment Effects. Spore production was slow during the first 12 h of incubation, but then increased rapidly between 12 and 48 hours. The greatest numbers of conidia were produced between 15-25C, reaching a maximum after 72 h incubation. Moderate levels of sporulation were observed at 10C and 30C and slight amounts at 5C; little or no sporulation occurred at the 1C and 35C temperature extremes. The optimum temperature for sporulation across all incubation durations ranged between 15C and 25C. Both dependent variables responded similarly to both environmental factors. Estimation of K parameter. The carrying capacity K for dependent variables S1 and S2 was estimated as the maximum value, which occurred at 72 h incubation. The relationship between maximum #conidia/lesion and temperature was similar in all three years. However, maximum levels of sporulation between 10C and 25C in 2009 were consistently lower than observed in 2007-08. When the asymptotes were fit as a function of temperature via the Gaussian model, graphical examination of the observed and predicted values showed a reasonably good fit of the model to the data for each year of the study. Overall, the fit of the model appeared somewhat better for the S2 dependent variable, especially at higher temperatures ( e 30°C) for the 2008 data. Results of the nonlinear regression analyses, which agree with this visual observation, showed that the total coefficient of variation for the parameter estimates, CVT, was 13.5% higher for S1 than S2 in 2008. Estimation of r parameter. The quadratic model provided a reasonably good fit to the Richards model rate parameters derived at each temperature level. Across all three years of the study, temperature described 79 to 83% of the variation in the rate parameter for the S1 dependent variable and 81 to 87% of the variation for the S2 dependent variable. Standard errors of the quadratic model parameters ranged from 14.2 to 19.1% and 10.7 to 18.4% of their estimated values for S1 and S2, respectively. Although these overall statistics were good, the model tended to under-predict the observed rate at 30°C and over-predict the rate at 35°C. These deviations were particularly evident for the 2008 and 2009 data. Sporulation Model. Each of the regression functions derived by fitting the Richards model (m = 1.01) to the dependent variables had highly significant F-values. Overall, the set of models provided very good data fits, explaining 80-85% of the variation in sporulation. When expressing sporulation as #condia per cm of twig (S2), Ra2 values were only slightly reduced. The nonlinear relationship between S, H, and T is expressed as: S = K [(1+exp(-rH)]**[1/(1-m)] (1) where K = maximum S, rH = b1H + b2HT + b3HT**2, and m = 1.01. If solved for various values of H and T, eq. 1 describes a three-dimensional response surface depicting sporulation. Model validation. Full and reduced models fitted to the 2007+2008 and 2009 data sets were statistically compared at each step of the development process. Comparison of the Gaussian models used to estimate the asymptote K as a function of temperature resulted in F-values of 5.30 (P = 0.01) and 1.29 (P = 0.31) for S1 and S2, respectively. Thus, the model parameters estimated from the 2009 data were not statistically different from those generated by the 2007+2008 data for the S2 dependent variable. However, the models describing sporulation as number of spores per lesion (S1) had one or more significantly different parameters. Although statistical comparisons of individual parameter estimates were not performed, comparison of the parameter values and graphs indicated that the model amplitudes (a) were most different. Full and reduced quadratic models employed in step three, which described the Richards rate parameter as a function of temperature, were also compared. The analyses of variance produced highly insignificant F-values of 0.00 (P = 1.00) and 0.11 (P = 0.95) for S1 and S2, respectively. Thus, the estimated quadratic model parameters for the 2009 model were not significantly different from their corresponding parameters generated from fitting the 2007+2008 data. Although models describing the K parameter were significantly different for S1, the multidimensional sporulation models (eq. 1) derived from the 2009 and 2007+2008 data sets were found to be statistically equivalent for both dependent variables. When full and reduced models were compared, insignificant F-values of 0.0022 (P = 0.99) and 0.03 (P = 0.99) were calculated for S1 and S2, respectively. Estimates for the b1, b2, and b3 parameters of the 2009 and 2007+2008 sporulation models varied by 2.2, 1.6, and 1.4% for S1, while those estimates for the S2 models varied by only 0.3, 0.0, and 1.3%, respectively. Unlike the parameter estimation models created during the intermediate steps, the sporulation models were derived from the much larger data sets including both independent variables. The larger sample sizes increased the degrees of freedom and, consequently, the power of the F-tests. Nevertheless, field validations will be necessary to insure that predictions are accurate under variable temperature and relative humidity conditions. Outreach activities to commercial growers, agents, and industry personnel consisted of presentations during experiment station field days in 2009 and 2010. Results were disseminated to professional plant pathologists by giving a presentation at the 2009 Annual Meeting of The American Phytopathological Society in Portland, OR. A published abstract can be found at Phytopathology 99:S68. A research manuscript has recently been written and submitted for publication. |
|
Impacts As stated in the original proposal, the major goal of this research-only project was to create the knowledge base necessary for the initial development of a peach scab forecasting model. Although this goal was achieved, it required an additional third year of data collection (in lieu of field validation). Thus, it is too early in the model development process to evaluate impacts. Most of the potential benefits of implementing the model, such as reduction in fungicide usage, better disease control, improved profitability, and reduced applicator/worker exposure to fungicide (see Problem, Background, and Justification section for details) are still possibilities. However, implementation of the model and subsequent determination of these benefits will depend on successful field validation under variable temperature and relative humidity conditions. |
| Report Appendices |
| Close Window |
|
Northeastern IPM Center 340 Tower Road Cornell University Ithaca, NY 14853 NortheastIPM.org |
 |
Developed by the Center for IPM © Copyright CIPM 2004-2026 |
|