Measuring mosaic diversity based on land use map in the region of Madrid, Spain

Abstract

Ecological sustainable landscapes require that their ecological processes can be sustained over time. Spatial heterogeneity is recognized as a very influent factor in biological diversity, but there is still a need to evaluate how this heterogeneity changes with scale and how pattern or processes change with the definition of the detail level, so this is the aim of this study.

A six level hierarchical land use map, comprising 7244 patches and 646 different land units was generated for the Region of Madrid. We calculated heterogeneity pattern indices, based on information theoretic measures (Shannon diversity index, Evenness) and fractal geometry index. The study area has been the Region of Madrid (8000 Km2) and three partitioning zones (sub-regions or macro-landscapes) of the territory, using vector-based techniques.

Territorial diversity is measured and different detail levels are compared from the hierarchical land use map. This analysis could establish the appropriate level of information necessary to reflect significant heterogeneity measurements.

The analysis shows that index values rise for each level and for the three studied zones. These values suggest that when the hierarchical classification has an increment in the detail level, the amount of information is broader and explains the spatial heterogeneity results. Zone analysis suggests that the woodland area contribution to Shannon index is higher than for less forested zones.

The conclusions drawn are that the regional level is not adequate for territorial diversity evaluation, whenever it contains macro-landscapes and sub-regions with wide physical or land-use differences. Another conclusion is the importance of increasing forest area in agricultural areas or high urban component, to increase landscape diversity.

This study allows for a better comprehension of territorial patterns and their meaning. It identifies the most influential land uses to heterogeneity from variations on the territorial pattern diversity, providing a simple and accessible methodology for the study of landscape processes.