Interpretation of connectivity metrics at different spatial resolutions: The case of Candarli Bay-Izmir, Turkey

Landscape metrics play an important role in the quantification of landscape structure and associated processes. But the outputs of landscape structural analysis are affected by the spatial resolution of source data. Whilst the influence of spatial resolution on landscape metrics has received considerable critical attention [e.g. 1, 2], recent studies suggest an association between connectivity metrics and the scale of the source data [3]. In this study, we analysed the structure and distribution of natural and semi-natural vegetation patches in Çandarli Bay-Izmir using a land cover map with 6 broad land cover types (Forests, Shrub and Herbaceous Vegetation Associations, Agricultural Areas, Artificial Surfaces, Wetlands and Water Bodies) and 9 landscape metrics (FRAGSTATS) across different spatial resolutions (2m, 5m, 10m, 25m, 50m and 100m). Then, we investigated the influence of those landscape metrics on the identification and prioritisation of land cover types in terms of their contribution to physical connectivity. Our results suggested that the results of individual landscape metrics are affected by spatial resolution, but when they interpreted and evaluated altogether, the prioritised land cover types did not change. © by PSP.