Habitat landscape pattern and connectivity indices : used at varying spatial scales for harmonized reporting in the EBONE project
暂无分享,去创建一个
This study is motivated by biodiversity related policy information needs on ecosystem fragmentation and connectivity. The aim is to propose standardized and repeatable methods to characterize ecosystem landscape structure in a harmonized way at varying spatial scales and thematic resolutions (habitat in situ versus land cover satellite based observations). Habitat landscape pattern was assessed in terms of configuration, interface mosaic context and structural/functional connectivity on the basis of three available conceptual models (morphological analysis, landscape composition moving window, network graph theory) that were customized, automated and partly combined. Input data were from the EBONE General Habitat Categories maps available over sixty 1 km2 in-situ samples at fine scale (400 m2 Minimum Mapping Unit). Demonstration focused on the focal forest phanerophyte habitat. Forest spatial pattern, edge interfaces and connectivity related maps and indices were obtained for all samples, and then reported per European Environmental Zones. A prototype web-based mapping client (http://forest.jrc.ec.europa.eu/ebone) was also developed to view and query the map layers and indices. Finally, the same models and indices were applied to the satellite based European and regional land cover maps available at broad (25 ha MMU) and medium (1ha MMU) scales. Differences in patterns across the three scales were highlighted over the only common 1 km2 analysis unit. Further, the satellite based patterns were reported at the more suitable fixed area grid of 25 km x 25 km. The overlay with the 1 km2 in situ habitat pattern enabled to inform the macro-scale landscape structure context of the squares and compare with their micro-scale pattern. Such study should be repeated to study spatio-temporal patterns relationships across scales once multi-temporal and larger in situ dataset will be available.