A data-mining approach to determine the spatio-temporal relationship between environmental factors and fish distribution

The interaction between environmental factors and the spatiotemporal dynamics of living organism is an important aspect in ecology. We describe here a data-mining approach—the spatiotemporal assignment mining model (STAMM)—to extract the spatiotemporal pattern, or assignment of environmental factors, which control the distribution of a living organism. In STAMM, the spatiotemporal assignment of environmental factors is expressed via neighbourhood rules which will reflect the fuzzy or uncertain prior knowledge about the relationship. The values of cells or points in the neighbourhood and the relationships are used to construct a decision table. Indices expressing the probabilities of the ecological association rules are recursively processed in order to determine the spatiotemporal assignment. These rules are objective assessments of our prior knowledge and they refine our knowledge and understanding of the ecosystem. As a case study, we used this model to study the temperature pattern which controls the assembling of fish in the Dasha area of the Yellow Sea in China.