Functional Context Similarity Units for Supporting Out-scaling of Sustainable Land Management in European Mediterranean Bioregion

It is widely recognized that Sustainable Land Management (SLM) are much needed for achievingLand Degradation Neutrality (LDN). Adoption and effectiveness of SLM depend on specificcontexts. Contextual diversity makes uniform blanket policies promoting SLM less effectiveand out-scaling of site-specific SLM difficult. We propose the functional context similar unit(fCSU) approach to overcome these challenges by grouping common biophysical, economicand social contextual factors of land use and land management into distinct context similarityunits that shape SLM adoption and resulting LDN.The presented study focuses on the European Mediterranean Bioregion within Mediterraneanbasin. The socio-ecological contextual factors selected for analysis includes 25 spatial variablesof biophysical condition (climate, topography, soil, land cover and livestock density), physicalaccessibility, demography and economic development status at the spatial resolution of 500m. We identified and mapped context types using spatially explicit (grid-based) multivariatestatistics. Core factors differentiating socio-ecological context were identified using principalcomponent analysis (PCA). Six distinct types of socio-ecological context were defined andmapped by cluster analysis. The empirical nesting of these clusters with aggregated land coverclasses resulted in 47 Context Similarity Units (CSU) in the study region.The functionality of the derived SCUs were evaluated by unbalanced, grid-based ANOVA thatmeasured and tested the differences in inter-annual trend of Net Primary Productivity (2001-2014 period) among the SCU within each aggregated land cover type. The result demonstratesthe potential of the function CSU approach to further our understanding LDN in socio-ecologicalcontexts. The result of fCSU mapping can be directly used by LDN-oriented projects/programs and citizen scientists in the region. For example given limited resource and aims, wecan know approximately where efforts should be focused by managing or coping with whatdrivers. The resulted fCSU is also being implemented as SLM extrapolation domains in theGeoinformatics Options by Context (GeOC) tool tailored to LANDSUPORT's S-DSS: given SLMoutcomes in a number of project sites, we can identify where similar intervention options havea potential of success based on contextual similarity.The presented study is under the framework of and supported by LANDSUPPORT project.