Plant Disease Forecasting and Model Validation: Classic and Modern Approaches

apple scab;disease modeling;Alternaria mali;Alternaria blotch of apples;decision suport systems;disease forecasting;Venturia inequalis

Thayer, Charles Lucas ; Dr. Turner Bond Sutton, Committee Chair,Dr. Barbara B Shew, Committee Member,Dr. Roger D Magarey, Committee Member,Dr. James F Walgenbach, Committee Member,Thayer, Charles Lucas ; Dr. Turner Bond Sutton ; Committee Chair ; Dr. Barbara B Shew ; Committee Member ; Dr. Roger D Magarey ; Committee Member ; Dr. James F Walgenbach ; Committee Member

University:North Carolina State University

关键词: apple scab;    disease modeling;    Alternaria mali;    Alternaria blotch of apples;    decision suport systems;    disease forecasting;    Venturia inequalis;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/2002/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Alternaria blotch of apples emerged in the late 1980s in North Carolina threatening the nations seventh largest apple producing state.Alternaria blotch, causal agent Alternaria mali, has since spread throughout the entire Southeast apple growing region.Alternaria blotch proved to be difficult to control with the management programs in place at the time.Filajdic et al developed a weather-based disease-forecasting model to aid in the timing of fungicide applications to manage the disease.However, this model was based on the use of iprodione (Rovral), which never became fully registered for use on apples. The purpose of this project is to improve the prediction of the onset and progression of Alternaria blotch to aid in the timing of fungicide applications.The ability of four Alternaria models for predicting the occurrence of Alternaria blotch was examined over three growing seasons.Five sites across the apple production area of Henderson County, North Carolina were established to track disease progress and collect weather information.The data collected on-site were compared to the output of the disease models generated from on-site weather data as well as geospatial satellite based weather data.The initial results of the project show that none of the examined models exhibit a strong correlation with either the onset or the progression of the disease.However, examining the individual disease model attributes revealed key information about the disease progression. The detailed model analysis suggests that the initial inoculum level, in addition to the increasing susceptibility of the host, most readily influences the progression of the disease over the course of a growing season.The traditional approach of plant disease model validations is usually limited in scope to counties, states, or regions.With the possible threat of bio-terrorist attacks on our nation's agricultural system, the ability to create and validate models on a national scale is of paramount importance in protecting America's immense agro-economic infrastructure.The nearly non-existent national disease incidence and severity data sets for foliar fungal pathogens are a serious limitation for the accurate validation of risk prediction models such as The North Carolina State University / Animal and Plant Health Inspection Service Plant Pathogen Forecasting System (NAPPFAST).The future of rapid disease model validations is dependent on the existence of research tools to give researchers the ability to accurately predict the potential establishment of exotic diseases.NAPPFAST is a template based modeling tool that uses daily 10-km2 geospatial weather input to create empirical infection models.An adequate validation of NAPPFAST models requires a data set of disease observations with national coverage for multiple years, pathogens and crops.The object of this project was to investigate the potential suitability of meta-analysis techniques to validate the NAPPFAST infection model using disease observation data from fungicide trials available in Fungicide and Nematicide Tests reports.Data on the incidence of apple scab, caused by Venturia inaequalis, were obtained for 10 years for six locations and 25-years for one location.There was a poor correlation between the output from the NAPPFAST model and the observed incidence of apple scab for the locations and years examined.The model appears to be missing variables such as inoculum density or phenological data that may play a key role in apple scab forecasting.