Field margin quantification and characterization using very high resolution multisource geospatial data in Wallonia, Belgium

Details

Supervisors

Defourny, Pierre ; Boland, Basile

Faculty

Faculté des bioingénieurs

Degree label

Master [120] en sciences agronomiques et industries du vivant, à finalité spécialisée: Geo-Information Science and Earth Observation for Environmental Modelling and Management

Abstract

Field margins are important structural and ecological components of the agroecosystem. The primary objective of this study is to develop a spatial approach that integrates LiDAR, PICC vector data, and very highresolution land cover information to accurately delineate field margins across 30 sites in Wallonia. This approach specifically aims to detect and characterize the natural elements present within these field margins, including vegetation, ditches, and grass margins, while evaluating the compatibility and complementarity of these three datasets for representing very small-scale landscape features. Furthermore, the study identifies and applies appropriate spatial metrics to quantify the extent, configuration, connectivity, and fragmentation patterns of these margin elements. This study shows that integrating SIGEC data, LiDAR- derived vegetation structure, PICC vector layer, and land cover maps provides an effective methodology for delineating and characterising the field margins across 30 sites in Wallonia. The margin detection algorithm achieved strong validation performance, while width estimation yielded a moderate predictive fit with some overestimation attributable to transitional vegetation zones. Landscape metric analysis revealed high spatial variability across sites, with most exhibiting Margin to total ratios exceeding one, indicating that field margins remain dominant linear ecological features, though a subset of intensively managed landscapes showed reduced connectivity and greater fragmentation. The weak correlations between individual landscape metrics reinforcing the necessity of a multi-indicator approach. Foliage Height Diversity analysis confirmed substantial vertical structural complexity across the study sites. DTM-based extraction identified a more extensive ditch network than PICC, including 455 predominantly shallow and narrow ditches, thereby providing improved hydrological characterization. Despite some limitations in field margin extraction, this study offers a foundational framework for more detailed characterization of field margin ecosystems in future research. Overall, the findings underscore the ecological significance of field margins within Wallonia’s agricultural landscapes, and the derived spatial metrics provide a useful basis for assessing margin connectivity and vegetation structure at the site level.