Abstract

Stream ecosystems form an active component of the carbon (C) cycle, and are identified as “hotspots” for carbon dioxide (CO2) emissions. However, the mechanisms driving CO2 emissions from streams are not completely understood. Beside the input of C in the form of CO2 from groundwater, streams receive organic matter from aquatic and terrestrial origins which is partly mineralized to inorganic nutrients and CO2. Future predictions suggest enhanced input of terrestrial organic matter into streams. As such, surrounding land use may highly influence dissolved organic matter (DOM) composition and turnover in streams. The quality, i.e. bioavailability or lability, of aquatic and terrestrial organic matter, as well as which quality feature provides which bioavailability, is controversially discussed and the research is still in its infancy. Thus, the main goal of my thesis is to enhance the understanding of the role of organic matter quality as a potential driver for organic matter turnover in stream ecosystems. A further goal is to shed light on C dynamics with main focus on CO2 of streams surrounded by different land use. The presented work is based on an experimental approach in the laboratory, supported by seasonal field studies and a developed model in order to explore C dynamics and the corresponding drivers in stream ecosystems. The underlying mechanisms and the importance of DOM quality as a main driver was assessed on the small scale in laboratory experiments. The C emissions from streams were quantified and the influence of DOM quality was examined on a stream reach scale by investigating two stream types with different organic matter quality inputs. By developing a process-based model, the understanding of the daily and seasonal scale of C turnover in stream ecosystems was amplified. The results from the experiment under controlled conditions demonstrate that DOM quality governs microbial metabolism (i.e. respiration and bacterial protein production). Moreover, I revealed significant quality differences between two terrestrial DOM sources, while respiration and bacterial protein production increased with the available proportion of the labile DOM source. The molecular weight of DOM was the strongest predictor of bacterial protein production and respiration, while among others, the concentration of low molecular weight substances was another highly influential predictor. The importance of molecular size/weight and DOM quality for microbial metabolism was further confirmed on the stream reach scale where we demonstrated among others a significant linkage between molecular size of DOM and pCO2 across agricultural and forest streams. Moreover, agricultural streams contained significantly higher pCO2 compared to forest streams during all seasons. However, CO2 emissions measured with the powerful drifting chamber method were not significantly different between the stream types. Modeled dissolved oxygen (O2) and CO2 dynamics calibrated with field data resulted in respiratory quotients (RQ = mole of CO2 produced per mole of O2 consumed), which are intimately linked to the elemental composition of the respired compounds across four seasons and two stream types. RQ values were not related to adjacent land use or season. Nevertheless, I found significant relationships between RQ values and DOM quality indicators, such as fluorescing component characteristic for higher plant material and molecule size of DOM in agricultural streams. In conclusion, this thesis demonstrates that DOM quality is an important driver for organic matter turnover in streams. Consequently, my results indicate that ongoing and future land use change and enhanced terrestrial DOM input into streams may influence CO2 emissions, and underline the status of streams as C turnover “hotspots”. Thus, my thesis contributes to the mechanistic understanding of organic matter cycling in stream ecosystems and their role in the regional and global C cycle. Therefore, organic matter quality should be considered in future models and studies with respect to C cycling.