Abstract
This paper introduces an Internet-PDA-based spatiotemporal survey and modeling service called, the quot;Ag Certification Networkquot;. This service was originally intended for the Caribbean Fruit Fly (Anastrepha suspensa Loew) Free Protocol certification program. Due to certain urgencies in tracking pest prevalence in Florida, the service was amplified to accommodate all important citrus insects and diseases, including citrus canker and greening diseases. One challenge was how this information can automatically be related to data from distributed global information networks, including other agricultural pest data acquisition systems. Our solution makes use of a universal algorithm for acquiring, aligning, and analyzing information from any agricultural crop-pest domain. Its foundation is a centralized data representation allowing relationships from input sources separated by structure or distance to be explored in real-time through mapping, mathematical modeling, and OLAP cube reporting. These features are valuable to growers because retrospective analyses against financial data become easier, and insects and diseases can be tracked in realtime to facilitate better management decisions. They are also useful to regulators who certify crops for sale to foreign or domestic markets. The features can also help researchers who wish to produce statistical analyses in real-time rather than wait until the end of an experiment to generate their results. We present our system architecture and show how observational information is acquired by centrally-located servers using handheld PDA computers. First-year mathematical and geostatistical modeling results suggest Caribbean Fruit Fly populations can be described by a logistic function and occur with an approximate 30-mile range of influence. This study utilizes a new computer language in development at MIT (the quot;Mquot; Language) that provides an interoperable data organization and modeling environment. The application of M supports automatic alignment and modeling of information from different agricultural, ecological, environmental, and financial domains. A commercial-level Ag Certification Network Pilot Trial scheduled to begin 1 September 2006 was recently ratified by the Indian River Citrus League. Information of the type collected in this trial could be useful in any surveillance program, including sentinel plot monitoring, grower self-surveys, pre-harvest inspections, laboratory-based passive tracking programs (e.g., the National Plant Diagnostic Network (NPDN), team-based research projects where data sharing is critical, and coordinated international insect and disease traceability programs.
