Abstract :
[en] Urban ponds are increasingly widespread as cities expand worldwide, yet their contribution to greenhouse gas (GHG) emissions remains poorly studied. However, they may also act as significant sources of GHG to the atmosphere. This thesis aims to unravel the processes controlling the production and emission of carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) in the urban ponds of Brussels (Belgium).
We first surveyed 22 ponds (0.1-4.6 ha) across Brussels to characterize spatial and seasonal patterns of GHG concentrations and identify their main drivers. Biological activity was the dominant seasonal control, while landscape context and external inputs shaped spatial variability. Smaller ponds emitted more CO2 due to higher allochthonous inputs, macrophyte-dominated ponds emitted more CH4, and N2O emissions peaked in the city center under higher nitrogen deposition. To move from citywide perspective toward a more detailed understanding, we monitored four representative ponds (two clear-water, two turbid-water) 46 times over 2.5 years. Clear-water ponds emitted more ebullitive CH4, likely from macrophyte-derived organic matter. Both CH4 and CO2 fluxes responded to meteorological conditions: CH4 increased with temperature, and CO2 with rainfall. We also showed that hydrogenotrophic methanogenesis predominated in clear-water ponds and that acetoclastic methanogenesis predominated in turbid-water ponds, with a shift towards hydrogenotrophic methanogenesis in fall. Methane oxidation was higher in turbid-water ponds, linked to suspended matter that enhanced bacterial activity and reduced light inhibition of methanotrophs, and accounted for most dissolved CH4 removal. Finally, by narrowing the focus to the finest temporal scale, high-frequency measurements conducted hourly from dawn to dusk in two ponds revealed that sub-daily variations were mainly driven by fluctuations in wind speed, with photosynthetic activity modulating daytime CO2 dynamics during warmer seasons; however, these short-term variations were low compared to seasonal and inter-pond differences, emphasizing that spatial heterogeneity among ponds constitutes the main source of uncertainty in emission estimates. Altogether, this work demonstrates that GHG dynamics in urban ponds result from a complex interplay between biological activity, hydromorphological characteristics, landscape context, and climatic forcing. It emphasizes the need to account for the ecological state of ponds, particularly macrophyte presence, when scaling up emissions, and to prioritize spatial and seasonal samplings for estimating GHG budgets from urban ponds.
