Microbial carbon processing in present-day lacustrine food webs: A multidimensional approach using stable isotopes, membrane lipid chemistry and modelling

Abstract

Freshwater ecosystems, although relatively small in volume, have a disproportionately large carbon storage capacity compared to oceans, as a result of relatively high production rates and fast sedimentation rates. Although of clear importance in the global carbon cycle, the role of lakes in terms of processing and storage of organic carbon is not very well characterized. Also, complete lacustrine carbon budgets are scarce and need to be better constrained in order to provide a more consistent view on the function of freshwater systems in the global carbon cycle. At the base of the food web, the microbial segment plays a crucial role in lacustrine food-web functioning and carbon transfer, thereby influencing carbon storage and CO2 emission and uptake in freshwater environments. Microbial communities, and thereby relative rates of production and consumption, are particularly sensitive to variability in nutrient supply, oxygenation, temperature, light and atmospheric CO2 levels. Most of these parameters vary between ecosystems, but also with water depth and season. This thesis aims to increase understanding of the processes within food webs that affect lacustrine carbon budgets, focusing specifically on the effects of depth, trophic states, and season on microbial carbon processing. To this end, field studies were carried out in freshwater systems with varying nutrient conditions: 1) Rotsee, a shallow, eutrophic Swiss lake, 2) Lake Lucerne, a large, oligotrophic Swiss lake, 3) two basins of Lake Taihu (China), one highly eutrophic and dominated by toxic algal blooms, and the other was subject to biomanipulation and restoration efforts and is now dominated by macrophytes, and 4) Lake Naarden (Naardermeer), a shallow Dutch peat lake. Stable carbon (δ13C) and nitrogen (δ15N) isotopes at either natural abundance or after tracer addition were analysed for total carbon and nitrogen pools and carbon isotopes were analysed for specific membrane lipids (PLFAs). We observed 1) substantial variability in absolute and relative amounts of production and consumption of organic matter, concluding that nutrient levels are of clear importance to microbial metabolic activities. 2) Nonetheless, despite lower nutrient levels in the macrophyte-dominated part of Lake Taihu we found total ecosystem productivity to be much higher compared to the still eutrophic basin. 3) We observed that carbon cycling pathways had shifted from phytoplankton-dominated to macrophyte-dominated, confirming the increased importance of the microbial food web as a consequence of eutrophication. 4) We found that severe DIC-limitation forced phytoplankton to derive significant amounts of their carbon from DOC, in addition to DIC. 5) Also large variability with depth was found in microbial communities and dominant metabolic processes, showing that multiple sampling depths are of crucial importance, especially in vertically stratified lakes. 6) We found that coupling between phytoplankton and bacterial production seems related to trophic state and 7) varied with water depth and during the year. We conclude that the potential for burial of organic matter in lake sediments is not only highly variable among different systems but also throughout the year and hence mass flux calculations and carbon budget calculations should be based on multiple sampling depths and campaigns.