Spectral heterogeneity of QuickBird satellite data is related to fine-scale plant species spatial turnover in semi-natural grasslands

Abstract Question: Can satellite data be related to fine-scale species diversity and does the integrated use of field and satellite data provide information that can be used in the estimation of fine-scale species diversity in semi-natural grassland sites? Location: The Baltic Island of Oland (Sweden). Methods: Field work including the on-site description of 62 semi-natural grassland sites (represented by three 0.5m0.5m plots per site) was performed to record response variables (total species richness, mean species richness and species spatial turnover) and field-measured explanatory variables (field-layer height and distance between plots). Within each site, QuickBird satellite data were extracted from a standardized sample area by associating each field plot with a 33 pixel window (1 pixel = 2.4m2.4 m). Explanatory variables (the normalized difference vegetation index and spectral heterogeneity) were generated from the satellite data. Correlation tests, univariate regressions, variance partitioning and multivariate linear regressions were used to analyse the associations between response and explanatory variables. Results: There was a significant association between the spectral heterogeneity of the near-infrared band and the field-measured spatial turnover of species. The most parsimonious explanatory model for each response variable included both field-measured and satellite-generated explanatory variables. The models explained 30–35% of the variation in species diversity (total richness 36%, mean richness 31%, species turnover 33%). Conclusions: High spatial resolution satellite data are capable of supplying fine-scale habitat information that is relevant for the monitoring and conservation management of fine-scale plant diversity in semi-natural grasslands