Figure 1. Methodology developed for characterization and mapping of Socio-Ecosystem Functional Types

Preliminary list of ESEFVs - Searching spatial databases

Once we analyze the results of the online survey and have a preliminary list of Essential Social-Ecological Functional Variables (ESEFVs), we will conduct a data sources search and collect spatial data of as many variables as possible from national and regional databases, in order to develop a first proof of concept of Socio-Ecosystem Functional Types (SEFTs) mapping in the south of Spain.



Methodology for characterizing SEFTs

To characterize SEFTs, we will applied a multidimensional approach using a composite index structure (Sullivan & Meigh, 2007). First, we will standardize each variable at the municipality level and transform into indicators applying the min-max normalization method (Sullivan, 2002) to allow comparisons between them. We consider appropriate for our approach the municipality level because its relevance to social processes and decision making (Raudsepp-Hearne et al., 2010). Then, we will explore the correlation between variables and apply Principal Component Analysis. We will select from each functional dimension as many variables as necessary to gather an 80% of the variance of each dimension. From these selected variables, we will develop weighted additive indexes for each dimension, only in case that variables are coherent for being aggregated (Sullivan & Meigh, 2007). A priori, we consider that variables from the following dimensions: well-being and development, provisioning ES supply, regulating ES supply, cultural ES supply, ecosystem disservices supply, ES demand, human pressure on environment, and social-ecological coupling; could be aggregated in weighted additive indexes. This indexes will constitute an integrated expression of the cited functional dimensions, in order to make easier and more intuitive the integration of these dimensions and ESEFVs into a characterization SEFTs. Weighting factors will be derived from online survey results. For the dimension “human population dynamics” and the component “ecosystem”, selected ESEFVs will be used independently, without being aggregated.

A similar approach was develop by Sullivan (2002) to construct a Water Poverty Index by integrating of biophysical and social data. In this study, a set of variables, relevant to water management issues, was also structured into several components. Based on perceptions of scientists, managers and stakeholder, the author derived the weighting factors for each variable to calculate the composite indexes for each component. Jemmali & Sullivan (2014) refined this method by using a PCA analysis to obtain the weighting factors. In our approach, as it’s been explained, we will combine both techniques (expert opinions and PCA) in: 1) an online survey to agree on a first list of ESEFVs; 2) PCA analysis to select the variables that we will integrate into a functional classification and mapping of social-ecological systems (SESs); and 3) a weighting of the variables, also from the survey’s result, to calculate composite indexes for functional dimensions cited.



Methodology for mapping SEFTs

At this point, a social-ecological functional characterization of each unit of analysis (municipality) may be visualized by using spider diagrams. Each vertex of the diagram should be interpreted as a functional attribute for SESs characterization. Finally, to map SEFTs from selected ESEFVs and calculated composite indexes, we will conduct a cluster analysis (Raudsepp-Hearne et al., 2010; Rodríguez-Loinaz et al., 2015; Hamann et al., 2015) to identify groups of municipalities with similar social-ecological characteristics. We will use rose-wind diagrams to visualize social-ecological functional attributes of each cluster, corresponding to a Socio-Ecosystem Functional Type (SEFT).

As a result, we expect to obtain a classification and mapping of SEFTs. The spatial patterns of SEFTs will provided an integrative characterization of the functional heterogeneity of both the biophysical and human components of SESs. We aim to show SEFTs potential for mapping land patches with similar dynamics of ecosystem services supply and demand that could be used in Earth System and ecosystem services modelling.