Spatial explicit assessment of rural land abandonment in the Mediterranean area

This study adopts the “syndrome approach”, originally defined by the Potsdam Institute for Climate Impact Research (PIK), (Downing et al., 2002) to assess and map rural land abandonment (RLA), that occurred during the period 1990–2005 within the wider Mediterranean area. The basic idea behind the syndrome approach is to describe change processes by archetypical, dynamic, and co-evolutionary patterns of civilization–nature interactions. In the frame of the Rural Exodus Syndrome the RLA can be interpreted as the occurrence of environmental degradation through the abandonment of traditional agricultural practices. Multi-source spatial data, including biophysical-related variables mainly derived from Earth Observation as well as socio-economical GIS-based data, were used to define proxies for expected underlying processes and drivers of the mentioned syndrome. The analysis of data is rooted in the fuzzy set theory and approximate reasoning techniques which allows for the handling of uncertain and imprecise knowledge of environmental systems. Generalized Conjunction/Disjunction operators (GCD) were applied to compute intermediate indicator score maps representing the conditions that may affect the RLA, and a bipolar operator was used to combine mandatory and favouring conditions with the aim of generating a RLA indicator. The indicator expresses the detailed location and severity, or degree, of the syndrome. The Northern Mediterranean was generally found to suffer from RLA to a distinctly higher degree than the Southern Mediterranean. Reported abandonment studies from the existing literature, the European CORINE land cover map, and the Less Favoured Areas (LFA) map all supported the findings by confirming plausibility through convergence of evidence from comparisons with different types of independent information. This spatially highly-detailed results obtained may be of particular interest to policy and decision makers involved in rural development planning in the conceived areas.

► We map rural land abandonment and occurrence in the Mediterranean area (1990–2005).
► Mountainous areas in North Mediterranean countries are most seriously affected.
► We integrate satellite derived biophysical data and socio-economic data.
► We account for data uncertainty and incompleteness by soft integration.