A fuzzy logic method to assess the relationship between landscape patterns and bird richness of the Rolling Pampas

The loss of biodiversity in productive ecosystems is a global concern of the last decades. The Rolling Pampas of Argentina is an intensively cropped region that underwent important land use and landscape change, with different impacts on biodiversity of both plants and animals. Land use type and habitat complexity are hypothesized to be the most important factors determining species richness in agro-ecosystems. But it is not easy to define these attributes in an unambiguous fashion, or determine their interactions at different spatial scales. A fuzzy logic approach allows overcoming some of these problems by using linguistic variables and logic rules to relate them and formulate hypothesis. We constructed fuzzy logic models to study how bird species richness in the Rolling Pampas is related to land use and habitat complexity, and how these variables interact at two spatial scales. Results showed that at the local scale, landscape complexity is the most important factor determining species numbers; trees and bodies of water are the most influential complexities. The effect of local scale landscape attributes was modified depending on the context at broader scales, so that agricultural sites were enriched when surrounded by more favorable landscapes. There was a high dispersion in the predicted/observed value relationship, indicating that landscape factors interact in more complex ways than those captured by the models we used. We suggest that the fuzzy logic approach is suitable for working with biological systems, and we discuss the advantages and disadvantages of its use.