Umeå University's logo

Global warming and land use scenarios suggest increased 21^st century nitrogen (N) inputs to aquatic systems. Nitrogen affects in-lake processing and, potentially, atmospheric exchange of greenhouse gases, probably being most relevant in unproductive systems. Here, we test for the first time the effect of a whole-lake experimental increase (threefold) in external nitrate loads on the atmospheric exchange of carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) from N-limited unproductive boreal lakes. Nitrate enrichment effects were assessed within a paired Before/After-Control/Impact framework based on 2-hourly to biweekly surface-water sampling of dissolved gas concentrations, and monthly whole-lake inventory surveys, carried out over 4 yrs in six lakes. Nitrate enrichment did not affect gas exchange during summer stratification and whole-lake gas inventories during summer and winter stratification. This finding specifically emphasizes the modest role of internal carbon fixation for the CO₂ dynamics of unproductive boreal lakes. A global synthesis of 52 published studies revealed a wide range of nutrient fertilization effects, both in systems similar to our experimental lakes, and other more productive systems. Effects depended mainly on the spatiotemporal scale of the study and became more pronounced when N enrichment was combined with phosphorous. Conclusively, although short-term and habitat-specific effects can occur, changes in N supply have only weak whole-ecosystem effects on greenhouse gas emissions from unproductive boreal lakes.

Forest clearcutting generally increase exports of carbon and nitrogen to downstream aquatic systems. Although these losses affect the greenhouse gas budget of managed forests, it is unknown if they modify greenhouse gas emissions of recipient inland waters. To assess this question, we quantified atmospheric fluxes of carbon dioxide (CO₂) methane (CH₄) and nitrous oxide (N₂O) of humic lakes and their inlet streams in four boreal catchments of which two were treated with forest clearcuts (18% and 44% of the catchment area) using a Before-After/Control-Impact-experiment. We measured gas concentrations and hydrological and physicochemical water characteristics in hillslope groundwater, along stream transects and at multiple locations in lakes at 2-hourly to biweekly intervals throughout the snow-free season over a four year period. These measurements were combined with atmospheric gas transfer measurements and models to calculate aquatic greenhouse gas emissions. Forest clearcutting did not change greenhouse gas emissions from streams or lakes, despite significant increases of CO₂ and CH₄ concentrations in hillslope groundwater. Clearcut effects on groundwater were likely buffered in the riparian zone. Hence, the greenhouse gas budget of forests initially after clearcutting is unlikely to be confounded by aquatic greenhouse gas emissions. However, our findings should be extrapolated with caution to other environments. Here, site-specific conditions makes our study system representative for systems where clearcutting causes only a limited initial impact on catchment hydrology and biogeochemistry.

Lakes emit carbon dioxide (CO₂) at globally significant rates. These emissions may be controlled by land use and climate. Disentangling these interactions is a prerequisite to accurately predicting anthropogenic impacts on carbon cycling in lake rich regions. We used environmental monitoring data of lakes collected in autumn (n = 439) and throughout the whole open-water season (n = 22) from a wet and a dry year in Swedish forest catchments to evaluate direct and indirect effects of temperature, precipitation, wind speed, forest productivity and forest clearcutting on lake CO₂ concentrations. We found that trends in CO₂ concentrations along climate and forestry gradients do not always correspond between autumn and open-water season averages, implying the need of validation of patterns derived from snap-shot data alone. According to spatially resolved data, autumn CO₂ concentrations increased with mean annual air temperature (dry year) or catchment forest productivity (wet year) but were not coupled to colored dissolved organic matter (CDOM) concentrations. In contrast, open-water season averaged CO₂ concentrations were constant across temperature and productivity gradients but increased with CDOM. These antagonistic results suggest that our fundamental understanding of the controls of CO₂ in lakes depends on whether spatially or temporally resolved data is used. Hence, trade-offs in sampling efforts clearly limit the questions that can be addressed regarding climate and land use effects on lake carbon cycling.

Whole-lake metabolism estimates based on free-water oxygen measurements may be compromised in unproductive lakes if the transport of oxygen by physical processes is neglected and large relative to biologically induced changes. Here, we quantify how oxygen fluxes associated with atmospheric exchange, vertical mixing and internal waves modify metabolism estimates in five unproductive boreal, brown water lakes. Water temperature and dissolved oxygen concentrations were measured at five minute intervals at 5-8 depths for 8-70 days per lake from May to September 2015. We estimated daily metabolism using three inverse Bayesian models accounting for atmospheric gas exchange using a (1) conventional wind-speed model and a (2) surface renewal model based on near-surface turbulence, and (3) accounting for within lake mixing by deepening of the actively mixing layer and the coefficient of eddy diffusivity. Gross primary production (GPP), ecosystem respiration (ER) and net ecosystem production (NEP) ranged from 0 to 0.9, -0.3 to -2.5 and -0.2 to -1.6 g C m^-3 d^-1, respectively. Metabolism estimates were reduced by up to 400% if oxygen time series were filtered to remove effects of internal waves and thermocline up- and downwelling, which have likely caused large changes in concentrations. Metabolism estimates differed by 0-25% depending on the atmospheric gas transfer model used and by 0-120% depending on whether fluxes from vertical mixing were considered. Almost half of the metabolism estimates were unreasonable (GPP<0 or ER>0), and these largely coincided with enhanced (up to 300%) diel cycles in physical oxygen fluxes. We conclude that physical processes should be explicitly assessed in three dimensions when using the free-water oxygen method to model metabolism in unproductive lakes.