|
Abstract text
|
Dissolved organic matter (DOM) is important for the function of aquatic ecosystems and can be used as a representation of the lake’s metabolome. DOM can enter an aquatic system via the runoff of the rainfall (or melting tundra) over the ecosystem’s watershed or from in water algal or microbial production. DOM optically detectable fraction – the colored dissolved organic matter (CDOM) – is often used as a proxy for dissolved organic matter. CDOM absorbs radiation in the ultraviolet and visible region of the spectrum and can be identified from satellite imagery. It originates from the degradation of plant materials and other organisms or from terrestrially imported substances. The source of CDOM is important information to understand environmental-driven dynamics in aquatic systems. Fluorescence spectroscopic techniques, such as the excitation–emission matrix (EEM) and the parallel factor analysis (PARAFAC), have been used to distinguish between allochthonous (humic-like) and autochthonous (protein-like) sources. Here, we assess the relationship between these fluorescent components and optical properties such as remote sensing reflectance and inherent optical properties (IOPs) in 19 lakes located within the Mecklenburg–Brandenburg Lake district in the North German Lowland. These lakes differ in size, shape, depth, trophic state and biogeochemical characteristics. Most lakes are connected in-series by rivers and natural or manmade channels. Water samples from these lakes were analyzed for absorbance and fluorescence measurements by using spectrophotometers. These samples were also used for the computation of the absorption coefficients of phytoplankton, CDOM, and non-algal particles, which were used as the IOP dataset. Remote sensing reflectance was calculated from radiometric measurement at the water surface using two handheld spectroradiometers (ASD, JETI). We started calculating 2 PARAFAC components yielding in a high correlation between both (Spearman’s rs of 0.88) indicating that it is difficult to differentiate these two components. Calculating 4 PARAFAC components, we observed a high correlation between component 1 and 2 (Spearman’s rs of 0.98) and between component 1 and 3 (Spearman’s rs of 0.87). Component 4 was the least correlated with Spearman’s rs of 0.23, 0.25 and 0.09 with components 1, 2 and 3 respectively. This indicates that it would be possible to differentiate components 1, 2 and 3 from component 4. The 2-dimensional correlation plot with the remote sensing reflectance and each component showed that for components 1, 2 and 3 the reflectance ratio at wavelengths 620 nm/590 nm was the most appropriate, while for component 4 the reflectance ratio was most appropriate at 825 nm/665 nm. In relation to the IOPs, the correspondence analysis showed that components 1, 2 and 3 are related to the absorption coefficient of CDOM and the absorption coefficient of non-algal particles while component 4 was related to the absorption coefficient of CDOM and the absorption coefficient of phytoplankton. These results indicate that components 1, 2 and 3 are related to the allochthonous CDOM while component 4 seems to be related to the autochthonous CDOM. Additionally, it also shows the potential of remote sensing for the identification CDOM sources which can help to understand aquatic ecosystem dynamics under environmental change.
|
