Location: Subtropical Plant Pathology Research
2012 Annual Report
2. Developed a tool to calculate resources required and optimize the use of sampling resources.
3. Deliver user friendly Excel-spreadsheet based module that can be used by regulatory agencies.
Progress made on Sampling methods for early detection: 1. We have collated data from existing data resources from USDA as well as datasets collected by other research groups to characterize a number of contrasting pathogen case studies. This was used to assess the generality of the model and analyze its behavior for different epidemiological characteristics. 2. We have finalized the development and integration of a sampling module into the epidemic model to simulate a surveillance program, simulate the process of disease detection via the sampling module and use this to determine probability distributions of disease incidence at the point of detection. 3. Using a non-spatial variant of the model and a Baysian approach we have calculated the disease severity distribution at first detection from characteristics of the epidemic dynamics and sampling interval and sampling size. Using this approach we derived a simple approximation formula for the expected severity of the disease at first detection. This approximation now needs to be tested against the full spatially explicit stochastic model.
Progress made on validating the Multiple Pest Survey (MPS) methodology: We have developed a method to validate the MPS methodology on bases of historic data. This new development enables us to evaluate to which extend the presently much used MPS actually gives a suitable sampling plan. The tests will be done in the next year.
Progress made on development of a survey costs calculator: 1. In collaboration with USDA, APHIS, new data sets to parameterize the travel risk model were sought. One alternative is the AIQM (APHIS Software System) data currently collected at ports of entry from international travelers. Country of origin is captured by this data set which is highly useful. We explored the addition of new fields to the AIQM data form, specifically to collect information on destination zip code. 2. We started to obtain the data sets, zip code will be used to replace current census tracts in the preliminary model. U.S. zip codes are based on U.S. Census in a similar manner to census tracts. Even when new AIQM data forms are established and deployed, it will require time to collect and compile the data for model use. This may require another one or more years. 3. The preliminary model is currently restricted to Florida. It will be updated and enhanced by the use of zip codes and by extensive consideration of epidemiological characteristics used to parameterize risk. If successful the model will be extended to include predictability of introductions for additional US states. However, due to the complexity of the data sets nationally across multiple states, this will likely require additional years to finalize. Primary responsibility USDA, ARS, USHRL, Fort Pierce.