Spatiotemporal variation in climate and weather, allochthonous carbon loads, and autochthonous factors such as lake metabolism (photosynthesis and respiration) interacts to regulate atmospheric CO2 exchange of lakes. Understanding this interplay in diverse basin types at different timescales is required to adequately place lakes into the global carbon cycle and predict CO2 flux through space and time. We analyzed 18 years of data from seven moderately hard lakes in an agricultural prairie landscape in central Canada. We applied generalized additive models and sensitivity analyses to evaluate the roles of metabolic and climatic drivers in regulating CO2 flux at the intra-annual scale. At mean conditions with respect to other predictors, metabolic controls resulted in uptake of atmospheric CO2 when surface waters exhibited moderate primary production but released CO2 only when primary production was very low (<8 μg/L or when dissolved nitrogen was elevated (>2,000 μg/L), implying that respiratory controls offset photosynthetic CO2 uptake under these conditions. Climatically, dry conditions increased the likelihood of in-gassing, likely due to evaporative concentration of base cations and/or reduced allochthonous carbon loads. While more research is required to establish the relative importance of climate and metabolism at other timescales (diel, autumn/winter), we conclude that these hard fresh waters characteristic of continental interiors are mainly affected by metabolic drivers of pCO2 at daily-monthly timescales, are climatically controlled at interannual intervals, and are more likely to in-gas CO2 for a given level of algal abundance than are soft water, boreal ecosystems.
Bibliographical noteFunding Information:
This work was funded by the NSERC Canada Discovery grants program (P. R. L. and G. L. S.) and Canada Research Chairs, Canada Foundation for Innovation, the Province of Saskatchewan, and the University of Regina (P. R. L.). This paper is a contribution to the Qu’Appelle Long-term Ecological Research Program. All R code and open data for analysis and figures has been archived with the doi 10.5281/zenodo.1284524. The version containing private data has been archived at https://github.com/ simpson-lab/jgr-co2-flux-private-data. Queries regarding private data belonging to the Qu’Appelle Long-term Ecological Research Program may be directed to Peter.Leavitt@uregina.ca. We want to thank past and present members of the Limnology Laboratory for assistance with data collection since 1994 and A Chalom for his quick response to changing code on the R PSE package. B. Tutolo provided very useful answers to some water chemistry questions. K. Hodder provided the map for the study sites. Finally, we thank anonymous reviewers for their valuable contributions to improving the manuscript.
- carbon dioxide