Socio-ecological determinants of multiple ecosystem services on the Mediterranean landscapes of the Ionian Islands (Greece)

•The local ecosystems of the Ionian Islands may be affected by intense human pressure.•Relationships among ecosystem services (ES), and their supply and demand, were explored.•Random Forest was used to reveal the factors that contribute to ES bundles.•Landscape structure, topography, and population...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Ecological modelling 2020-04, Vol.422, p.108994, Article 108994
Hauptverfasser: Lorilla, Roxanne Suzette, Poirazidis, Konstantinos, Detsis, Vassilis, Kalogirou, Stamatis, Chalkias, Christos
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•The local ecosystems of the Ionian Islands may be affected by intense human pressure.•Relationships among ecosystem services (ES), and their supply and demand, were explored.•Random Forest was used to reveal the factors that contribute to ES bundles.•Landscape structure, topography, and population contributed to the supply bundles.•Topography and population contributed to the demand bundles. Mediterranean islands are widely recognized as biodiversity hotspots, with a long history of human activities shaping multi-functional landscapes. Socioeconomic and environmental factors are among the most important factors driving the creation of diverse landscapes, with a high supply of ecosystem services (ES). However, these factors, along with climate change, could also have irreversible consequences on local ecosystems. Thus, this study aimed to reveal the importance of socio-ecological factors in shaping ES bundles to manage natural resources efficiently and enhance human well-being. Using the Ionian Islands as a case study, we explored the relationships among multiple ES, including their supply and demand indicators. We identified bundles of ES to distinguish regions in which supply and demand exhibit different characteristics. An ensemble machine learning method (Random Forest - RF) was used to identify the most important socio-ecological variables out of 17 tested that contribute to ES bundles. Our results produced five bundles of ES supply and six bundles of ES demand. The most important variables for the distribution of ES supply bundles were landscape heterogeneity, elevation, slope, landscape connectivity, and population. In comparison, variables representing elevation, slope, and population were among the most important variables contributing to ES demand bundles. RF exhibited both good classification and predictability, which was supported by the accuracy measures. Our findings demonstrated that research on ES should account for underlying socio-ecological drivers that influence the supply and demand of ES to improve our understanding of the possible impacts of future management decisions regarding the diverse Mediterranean landscapes of the Ionian Islands.
ISSN:0304-3800
1872-7026
DOI:10.1016/j.ecolmodel.2020.108994