Integration of data is critical to achieving new levels of understanding of ecological systems and processes. Typically, data integration is achieved only through a painstaking manual process that rules out large-scale integration. We believe that many of the techniques related to uncertain reasoning (fuzzy logic, Baysian networks, and evolutionary algorithms) and data mining might be usefully applied to ecological data integration. Here we present two case studies. One characterizes a traditional approach to integration. The second focuses on using software system integration to integrate geospatial and research data, along with providing data discovery services. We discuss those case studies where advanced techniques might prove useful and where modifications are needed to support scientific research.

Orlando, FL