Spatial patterns of aboveground phytogenic Si stocks in a grass-dominated \ncatchment – Results from UAS based high resolution remote sensing

Abstract

Abstract. Various studies have been performed to quantify silicon (Si) stocks in plant biomass and related Si fluxes in terrestrial biogeosystems. Most of these studies were performed at relatively small plots with an intended low heterogeneity in soils and plant canopy composition, and results were extrapolated to larger spatial units up to global scale implicitly assuming similar environmental conditions. However, the emergence of new technical features and increasing knowledge on details in Si cycling leads to a more complex picture at landscape or catchment scales. Dynamic and static soil properties change along the soil continuum and might influence not only the species composition of natural vegetation, but its biomass distribution and related Si stocks. Maximum Likelihood (ML) classification was applied to multispectral imagery captured by an Unmanned Aerial System (UAS) aiming the identification of land cover classes (LCC). Subsequently, the Normalized Difference Vegetation Index (NDVI) and ground-based measurements of biomass were used to quantify aboveground Si stocks in two Si accumulating plants (Calamagrostis epigejos and Phragmites australis) in a heterogeneous catchment and related corresponding spatial patterns of these stocks to soil properties. We found aboveground Si stocks of C. epigejos and P. australis to be surprisingly high (maxima of Si stocks reach values up to 98\u2009g\u2009Si\u2009m−2), i.e., comparable to or markedly exceeding reported values for the Si storage in aboveground vegetation of various terrestrial ecosystems. We further found spatial patterns of plant aboveground Si stocks to reflect spatial heterogeneities in soil properties. From our results we concluded that (i) aboveground biomass of plants seems to be the main factor of corresponding phytogenic Si stock quantities and (ii) a detection of biomass heterogeneities via UAS-based remote sensing represents a promising tool for the quantification of lifelike phytogenic Si pools at landscape scales.