Sentinel-2 images reveal functional biophysical heterogeneities in crop mosaics

Résumé En

Context Identifying landscape structure and understanding its functions are crucial for biological control. However, the relationship between the crop mosaic phenological heterogeneity and crop phenology at the field scale remains a blind spot. This hinders the understanding of crop dynamics and associated biodiversity. Remote sensing images are commonly used in landscape ecology because they allow for regular fine-scale monitoring of large areas. Objective The objective of this study was to understand the influence of biophysical heterogeneity of the crop mosaic on crop phenology and biodiversity using optical satellite images. Methods Indicators of wheat phenology and biophysical heterogeneity were derived from Sentinel-2 images using the Weighted Difference Vegetation Index (WDVI). A landscape gradient was studied in 2017 and 2018 using six study sites in Brittany, Picardy (France) and Wallonia (Belgium). First, we analyzed relationships among the crop mosaic, landscape grain and biophysical heterogeneity. Second, we studied effects of biophysical heterogeneity on wheat phenology. Last, we used WDVI to estimate the distribution of two carabid beetle species. Results The biophysical heterogeneity derived from WDVI correlated strongly with the crop mosaic gradient and landscape grain. Biophysical heterogeneity appeared to benefit wheat growth in fine-grain landscapes but disadvantage it in coarse-grain landscapes during the stem-extension and ripening periods. Biophysical heterogeneity estimated the distribution of carabid beetle species accurately. Conclusion The biophysical heterogeneity metric is continuous, consistent across locations and crop types and enables to address ecological issues using freely available satellite images covering the Earth. Future studies could use this metric to study the dynamics of species.