Effects of tree species diversity and identity on remotely sensed canopy variables in two forest biodiversity experiments

Remote sensing can link biodiversity and its effects on ecosystem functioning in forests, but in observational studies environmental covariates may confound results. We tested effects of tree species richness and functional diversity on remotely-sensed canopy variables at community and species levels in two forest biodiversity experiments established under homogeneous environmental conditions. At the BIOTREE sites in Kaltenborn and Bechstedt, we combined airborne imaging spectroscopy-based vegetation indices, sun-induced chlorophyll fluorescence, and land surface temperature at 1-m resolution, with ground-based tree stem growth as a productivity measure. At Kaltenborn, increasing species richness (one to four species) did not increase community productivity because of contrasting species-level responses. The normalized difference vegetation index (NDVI) decreased with richness in both needle-leaved species but increased in both broadleaved species. At Bechstedt, increasing functional diversity of mixtures planted with four tree species lowered community-level normalized difference water and photochemical reflectance indices; responses differed between four functional groups of tree species. NDVI was negatively correlated with basal area; correlations among other spectral, thermal, and fluorescence traits varied among tree functional groups. These results show that forest biodiversity experiments can test remotely sensed indicators of tree diversity and ecosystem functioning and inform interpretation of observational studies.