Decrease in CO2 efflux from northern hardwater lakes with increasing atmospheric warming



Boreal lakes are biogeochemical hotspots that alter carbon fluxes by sequestering particulate organic carbon in sediments1,2 and by oxidizing terrestrial dissolved organic matter to carbon dioxide (CO2) or methane through microbial processes3,4. At present, such dilute lakes release 1.4 petagrams of carbon annually to the atmosphere3,4, and this carbon efflux may increase in the future in response to elevated temperatures5 and increased hydrological delivery of mineralizable dissolved organic matter to lakes6,7. Much less is known about the potential effects of climate changes on carbon fluxes from carbonate-rich hardwater and saline lakes that account for about 20 per cent of inland water surface area4,8. Here we show that atmospheric warming may reduce CO2 emissions from hardwater lakes. We analyse decadal records of meteorological variability, CO2 fluxes and water chemistry to investigate the processes affecting variations in pH and carbon exchange9,10 in hydrologically diverse lakes of central North America. We find that the lakes have shifted progressively from being substantial CO2 sources in the mid-1990s to sequestering CO2 by 2010, with a steady increase in annual mean pH. We attribute the observed changes in pH and CO2 uptake to an atmospheric-warming-induced decline in ice cover in spring that decreases CO2 accumulation under ice, increases spring and summer pH, and enhances the chemical uptake of CO2 in hardwater lakes. Our study suggests that rising temperatures do not invariably increase CO2 emissions from aquatic ecosystems.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Temporal changes in summer pH and CO2 flux in six hardwater lakes of central Canada.
Figure 2: Effects of winter temperature on ice cover, lake chemistry and CO2 flux in lakes of central Canada.
Figure 3: Relationship between duration of ice cover, water-column pH and oxygen content before ice melt in Buffalo Pound Lake.


  1. 1

    Burgermeister, J. Missing carbon mystery: case solved? Nature Clim. Change 3, 36–37 (2007)

  2. 2

    Downing, J. A. et al. Sediment organic carbon burial in agriculturally eutrophic impoundments over the last century. Glob. Biogeochem. Cycles 22 GB1018 10.1029/2006GB002854 (2008)

  3. 3

    Cole, J. J. et al. Plumbing the global carbon cycle: integrating inland waters into the terrestrial carbon budget. Ecosystems 10, 172–185 (2007)

  4. 4

    Tranvik, L. J. et al. Lakes and impoundments as regulators of carbon cycling and climate. Limnol. Oceanogr. 54, 2298–2314 (2009)

  5. 5

    Marotta, H. et al. Greenhouse gas production in low-latitude lake sediments responds strongly to warming. Nature Clim. Change 4, 467–470 (2014)

  6. 6

    Rantakari, M. & Kortelainen, P. Interannual variation and climatic regulation of the CO2 emission from large boreal lakes. Glob. Change Biol. 11, 1368–1380 (2005)

  7. 7

    Sobek, S., Tranvik, L. J., Prairie, Y. T., Kortelainen, P. & Cole, J. J. Patterns and regulation of dissolved organic carbon: an analysis of 7,500 widely distributed lakes. Limnol. Oceanogr. 52, 1208–1219 (2007)

  8. 8

    Duarte, C. M. et al. CO2 emissions from saline lakes: a global estimate of a surprisingly large flux. J. Geophys. Res. 113 G04041 10.1029/2007JG000637 (2008)

  9. 9

    Finlay, K., Leavitt, P. R., Patoine, A. & Wissel, B. Magnitudes and controls of organic and inorganic carbon flux through a chain of hard-water lakes on the northern Great Plains. Limnol. Oceanogr. 55, 1551–1564 (2010)

  10. 10

    Finlay, K., Leavitt, P. R., Wissel, B. & Prairie, Y. T. Regulation of spatial and temporal variability of carbon flux in six hard-water lakes of the northern Great Plains. Limnol. Oceanogr. 54, 2553–2564 (2009)

  11. 11

    del Giorgio, P. A., Cole, J. J. & Cimbleris, A. Respiration rates in bacteria exceed phytoplankton production in unproductive aquatic ecosystems. Nature 385, 148–151 (1997)

  12. 12

    Ask, J. et al. Whole-lake estimates of carbon flux through algae and bacteria in benthic and pelagic habitats of clearwater lakes. Ecology 90, 1923–1932 (2009)

  13. 13

    Prairie, Y. T., Bird, D. F. & Cole, J. J. The summer metabolic balance in the epilimnion of southeastern Quebec lakes. Limnol. Oceanogr. 47, 316–321 (2002)

  14. 14

    Bastviken, D., Tranvik, L. J., Downing, J. A., Crill, P. A. & Enrich-Prast, A. Freshwater methane emissions offset the continental carbon sink. Science 331, 50 (2011)

  15. 15

    Schindler, D. E., Carpenter, S. R., Cole, J. J., Kitchell, J. F. & Pace, M. Influence of food web structure on carbon exchange between lakes and the atmosphere. Science 277, 248–250 (1997)

  16. 16

    Hammer, U. T. Saline Lake Ecosystems of the World (Junk, 1986)

  17. 17

    Stumm, W. & Morgan, J. J. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters (Wiley, 1996)

  18. 18

    Pham, S. V., Leavitt, P. R., McGowan, S., Wissel, B. & Wassenaar, L. Spatial and temporal variability of prairie lake hydrology as revealed using stable isotopes of hydrogen and oxygen. Limnol. Oceanogr. 54, 101–118 (2009)

  19. 19

    Vogt, R. J., Rusak, J. A., Patoine, A. & Leavitt, P. R. Differential effects of energy and mass influx on the landscape synchrony of lake ecosystems. Ecology 92, 1104–1114 (2011)

  20. 20

    Falkowski, P. G. et al. The global carbon cycle: a test of our knowledge of earth as a system. Science 290, 291–296 (2000)

  21. 21

    Patoine, A., Graham, M. D. & Leavitt, P. R. Spatial variation of nitrogen fixation in lakes of the northern Great Plains. Limnol. Oceanogr. 51, 1665–1677 (2006)

  22. 22

    Leavitt, P. R. et al. Paleolimnological evidence of the effects on lakes of energy and mass transfer from climate and humans. Limnol. Oceanogr. 54, 2330–2348 (2009)

  23. 23

    Bonsal, B. R., Prowse, T. D., Duguay, C. R. & Lacroix, M. P. Impacts of large-scale teleconnections on freshwater-ice break/freeze-up dates over Canada. J. Hydrol. (Amst.) 330, 340–353 (2006)

  24. 24

    Striegl, R. G. & Michmerhuizen, C. M. Hydrologic influence on methane and CO2 dynamics in two north-central Minnesota lakes. Limnol. Oceanogr. 43, 1519–1529 (1998)

  25. 25

    Striegl, R. G. et al. Carbon dioxide partial pressure and 13C content of north temperate and boreal lakes at spring ice melt. Limnol. Oceanogr. 46, 941–945 (2001)

  26. 26

    Magnuson, J. J. et al. Historical trends in lake and river ice cover in the Northern Hemisphere. Science 289, 1743–1746 (2000)

  27. 27

    Baehr, M. M. & DeGrandepre, M. D. Under-ice CO2 and O2 variability in a freshwater lake. Biogeochemistry 61, 95–113 (2002)

  28. 28

    Kratz, T. K., Cook, R. B., Bowser, C. J. & Brezonik, P. L. Winter and spring pH depressions in northern Wisconsin lakes caused by increase in pCO2 . Can. J. Fish. Aquat. Sci. 44, 1082–1088 (2007)

  29. 29

    Wissel, B., Cooper, R. N., Leavitt, P. R. & Pham, S. V. Hierarchical regulation of pelagic invertebrates in lakes of the northern Great Plains: a novel model for inter-decadal effects of future climate change. Glob. Change Biol. 17, 172–185 (2011)

  30. 30

    Province of Saskatchewan. Saskatchewan’s State of the Environment Report (Government of Saskatchewan, 2008)

  31. 31

    Kouraev, A. V. et al. Sea ice cover in Caspian and Aral Seas from historical and satellite data. J. Mar. Syst. 47, 89–100 (2004)

  32. 32

    Lukashin, V. N. et al. An integrated study in the Northern Caspian Sea: the 30th voyage of the R/V Rift. Oceanology (Mosc.) 50, 439–443 (2010)

  33. 33

    Tuzhilkin, V. S., Katunin, D. N. & Nalbandov, Y. R. in Handbook of Environmental Chemistry Vol. 5P (eds Kostianoy, A. G. & Kosarev, A. N. ) 83–108 (Springer, 2005)

  34. 34

    Pham, S. V., Leavitt, P. R., McGowan, S. & Peres-Neto, P. Spatial variability of climate and land-use effects on lakes of the northern Great Plains. Limnol. Oceanogr. 53, 728–742 (2008)

  35. 35

    Buffalo Pound Water Administration Board. 2012 Annual Report (City of Regina, 2012)

  36. 36

    Millero, F. J. The marine inorganic carbon cycle. Chem. Rev. 107, 308–341 (2007)

  37. 37

    Kling, G. W., Kipphut, G. W. & Miller, M. C. The flux of CO2 and CH4 from lakes and rivers in arctic Alaska. Hydrobiologia 240, 23–36 (1992)

  38. 38

    Cole, J. J. & Caraco, N. F. Atmospheric exchange of carbon dioxide in a low-wind oligotrophic lake measured by the addition of SF6 . Limnol. Oceanogr. 43, 647–656 (1998)

  39. 39

    Hoover, T. E. & Berkshire, D. C. Effects of hydration on carbon dioxide exchange across an air–water interface. J. Geophys. Res. 74, 456–464 (1969)

  40. 40

    Wanninkhof, R. & Knox, M. Chemical enhancement of CO2 exchange in natural waters. Limnol. Oceanogr. 41, 689–697 (1996)

  41. 41

    Wanninkhof, R. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res. 97, 7373–7382 (1992)

  42. 42

    Zou, H. & Hastie, T. Regularization and variable selection via the elastic net. J. R. Stat. Soc. B 67, 301–320 (2005)

  43. 43

    Hastie, T., Tibshirani, R. & Friedman, J. The Elements of Statistical Learning; Data Mining, Inference, and Prediction 2nd edn (Springer, 2009)

  44. 44

    Friedman, J., Hastie, T. & Tibshirani, R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 33, 1–22 (2010)

  45. 45

    R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2014)

  46. 46

    Bethke, C. M. & Yeakel, S. The Geochemist’s Workbench Release 9.0 Reaction Modeling Guide (Aqueous Solutions, LLC, 2013)

  47. 47

    Shapley, M. D., Ito, E. & Donovan, J. J. Authigenic calcium carbonate flux in groundwater-controlled lakes: implications for lacustrine paleoclimate records. Geochim. Cosmochim. Acta 69, 2517–2533 (2005)

  48. 48

    Luhmann, A. J. et al. Experimental dissolution of dolomite by CO2-charged brine at 100°C and 150bar: evolution of porosity, permeability, and reactive surface area. Chem. Geol. 380, 145–160 (2014)

  49. 49

    Tutolo, B. M., Kong, X. Z., Seyfried, W. E., Jr & Saar, M. O. Internal consistency in aqueous geochemical data revisited: applications to the aluminum system. Geochim. Cosmochim. Acta 133, 216–234 (2014)

  50. 50

    Berggren, M., Lapierre, J.-F. & del Giorgio, P. A. Magnitude and regulation of bacterioplankton respiratory quotient across freshwater environmental gradients. Int. Soc. Microb. Ecol. J. 6, 984–993 (2012)

Download references


We thank I. Phillips for estimates of lake area in Saskatchewan; D. Conrad for water chemistry data from Buffalo Pound Lake, Katherine Miller for compilation and analysis of Buffalo Pound data; J. Piwowar for assistance with GIS data; S. Pham and the University of Regina Limnology Laboratory for field work in 1996–2010; and Y. T. Prairie, E. G. Stets, J. A. Downing and D. E. Schindler for reviewing the manuscript. NSERC Canada, the Canada Foundation for Innovation, the Province of Saskatchewan, and the Canada Research Chair programme provided funding for this work.

Author information

P.L. and K.F. designed the study. P.L. provided data from Qu’Appelle lakes. P.L. and B.W. provided data from other hardwater lakes. B.T. conducted geochemical modelling, G.S. conducted elastic net analysis, and all authors contributed additional numerical analysis. P.L., K.F. and R.V. wrote the manuscript. All authors edited the manuscript.

Correspondence to Peter R. Leavitt.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Map of study region in Saskatchewan, Canada.

Hardwater lakes of the Qu’Appelle catchment (triangles) were monitored every two weeks from May to September during 1995–2010 (ref. 19), and closed-basin lakes were monitored monthly (open circles) or annually (filled circles) during 2002–2010 (except 2006)29. Weekly monitoring of pH occurred at Buffalo Pound Lake (black triangle) during 1979–2007. All lakes are situated in prairie grassland ecozones with pronounced precipitation deficits (annual precipitation − potential evaporation) of 40–60 cm yr−1 (dashed lines)18,29. Source data

Extended Data Figure 2 Principal components analysis of the relationship between mean annual surface water pH, annual meteorological conditions and mean summer lake parameters during 1995–2010.

a, Ordination of mean summer pH in Qu’Appelle lakes (n = 6) during 1 May to 31 August in relation to mean annual meteorological conditions revealed that pH was correlated positively with mean annual and spring (not shown) temperatures, correlated negatively with the date of ice melt, and was unrelated to mean annual levels of precipitation or irradiance. Variables include log10-transformed mean annual temperature (temperature), total annual rainfall (rain), total annual precipitation (precipitation), total snowfall (snow), untransformed daily hours of bright sunlight (irradiance) and the calendar day of the year when ice was completely melted from the lake surface (ice melt date). b, Ordination of mean summer pH in relation to coeval chemical, hydrological and physical conditions in Qu’Appelle lakes, as well as indices of relevant global climate systems. Abbreviations include water temperature (T°CH20), total inorganic carbon (TIC), dissolved organic carbon (DOC), total dissolved nitrogen (TDN), log10-transformed chlorophyll a (Chl), conductivity (Cond), log10-transformed soluble reactive phosphorus (SRP), log10-transformed total dissolved phosphorus (TDP), log10-transformed dissolved ammonia/ammonium (NH4), turbidity (Secchi depth), log10-transformed volume of river inflow (inflow), dissolved oxygen (O2), log10-transformed dissolved nitrite + nitrate (NO3) and climate indices representing the Pacific Decadal Oscillation (PDO), the North Atlantic Oscillation (NAO) and the winter (SOIwinter) or annual (SOImean) Southern Oscillation Index. This analysis reveals that mean summer pH is correlated positively with the PDO and negatively with the SOI, consistent with the interpretation that warm winters and reduced ice cover result in higher summer pH in Qu’Appelle lakes. Source data

Extended Data Figure 3 Elastic net analysis to identify and rank predictors of changes in under-ice pH in Buffalo Pound Lake during winter.

Water quality parameters at 1.5 m above the lake bottom were analysed weekly using uniform methods during 1985–2003 from the date of ice-cover formation to the date of ice melt. Analysis was performed using 125 weekly observations with complete water chemistry. Parameters include concentrations of dissolved oxygen (O2), sodium (Na+), carbonate (CO32−), loge-transformed dissolved aluminium (logeAl), fluoride (F), potassium (K+), loge-transformed orthophosphate (logePO43−), calcium (Ca2+), dissolved magnesium (Mg+), loge-transformed nitite + nitrate (logeNO3), loge-transformed dissolved manganese (logeMn), bromide (Br), total phosphorus (TP), dissolved iron (Fe), chloride (Cl), loge-transformed ammonium (logeNH4+), bicarbonate (HCO3), sulphate (SO42−) and temperature (temp). Coloured lines indicate how standardized regression coefficients (y axis, left) develop (right to left) as the initial pool of predictors (y axis, right) is refined by removing collinear and non-significant variables. Evaluation of the standardized coefficients of the most parsimonious model (vertical dashed line; equation under graph) demonstrates that changes in microbial metabolism (O2 decline × respiratory quotient of 1.2 = CO2 production)4,11 was the main factor regulating variation in water-column pH under ice, showing a nearly fourfold greater coefficient (0.14) than did either HCO3 or Ca2+ (0.04). Although dissolved logeAl, logeNH4+ and CO32− were also significant predictors of changes in winter pH (standardized coefficients 0.03–0.07), concentrations of these solutes (means ± s.e.m.; n = 17) were too low (<0.01 M) to regulate lake-water pH relative to the effects of changes in O2 (0.62 M), HCO3 (3.69 M) or Ca2+ (1.22 M). This analysis suggests that metabolically produced CO2 mainly regulates variation in winter pH by the production of carbonic acid (reduces pH), but that pH decline is slightly tempered by CO2-induced dissolution of sedimentary CaCO3, the main form of sedimentary carbon in Buffalo Pound9. Source data

Extended Data Figure 4 Effects of ice melt on water chemistry and spring carbon efflux from Buffalo Pound Lake, 1985–2003.

a, Mean pH recorded at 1.5 m above the lake bottom from four weeks before to four weeks after ice melt (week = 0). The rate of pH increase was linear with time (r = 0.96, P < 0.0001) with a slightly higher magnitude of increase (0.28 units) occurring in the two weeks after ice melt. b, Changes in mean concentration (mg C l−1) of total inorganic carbon (TIC) during spring. The maximum extent of TIC decline (6.57 mg C l−1; 14.6% of TIC stock at week 0) occurred during the two weeks after ice melt as a result of a combination of CO2 efflux (0.86 mg C l−1) and dilution by snowmelt (5.71 mg C l−1), as documented in d. c, Changes in concentrations of HCO3 and CO32−, and log10 of calculated CaCO3 saturation index. Patterns reveal that the decline in TIC was due to a decrease in HCO3 concentrations, rather than to the precipitation of CaCO3. d, Changes in mean concentrations (mg l−1) of chloride (Cl) and fluoride (F) during spring. Because concentrations of conservative tracers declined by an average of 12.7% in the two weeks after ice melt, yet TIC declined by 14.6%, we estimated that 1.9% of the decline in TIC stock at week 0 (0.86 mg C l−1) was caused by loss of inorganic C, particularly atmospheric evasion. Geochemical modelling of water chemistry observed at week 0 demonstrated that this magnitude of CO2 loss should increase the pH by 0.23 ± 0.08 units by week 2, a value equivalent to the observed increase in pH (see a). e, Observed changes in water temperature were too small (1.7 °C) to influence water chemistry strongly during the two weeks after ice melt. f, Calculated changes in chemically enhanced CO2 efflux modelled with observed water chemistry. Ice cover was assumed to prevent potential atmospheric CO2 exchange (grey shading), whereas most CO2 efflux seemed to occur one to two weeks after ice melt. All water samples were collected weekly at 1.5 m above the bottom of 3.0-m-deep Buffalo Pound Lake. Error bars represent s.e.m. (n = 17). During each year, sampling intervals were standardized to the documented week of ice melt (week = 0) before the calculation of long-term means. Changes in CaCO3 saturation and water-column were modelled with observed water-column parameters and Geochemists Workbench version 9.0.9 (see Methods). Source data

Extended Data Figure 5 Relationship between surface water pH in hardwater lakes of the Qu’Appelle River catchment and hydrologically closed lakes of southern Saskatchewan.

a, Qu’Appelle lakes (n = 6) were monitored every two weeks during summer 1995–2010 (ref. 19); closed-basin lakes (n = 20 in most years) were sampled monthly or seasonally (spring, summer and autumn) during 2002–2009 (except 2006, when there were no samples)29. b, Seasonal change in mean surface water pH of 15 closed-basin lakes monitored monthly during 2002–2009. Error bars represent one s.e.m. These patterns demonstrate that the pH of closed-basin lakes varied synchronously with that of Qu’Appelle lakes on both annual and seasonal scales, despite large differences in hydrological properties10,19. Source data

Extended Data Figure 6 Global map of regions where climatic conditions and soil types resemble those of southern Saskatchewan, Canada.

Hardwater lake distribution is not well quantified; however, this map depicts the region in which subsoil composition favours hardwater lakes and where climatic conditions produce substantial winter ice cover. Soil data originate from the FAO–UNESCO Soil Map of the World; regions highlighted in black have subsoil concentrations of CaCO3 in excess of 10% (for example Cambisol, Xerosol, Yarsol, Kastanozem and Chernozem soils). These data were overlain with temperature data (10 arcminute resolution, averaged monthly during 1950–2000) obtained from the WorldClim Global Climate Data that were restricted to regions where the monthly average temperature was below 0 °C for December–February (June–August for the Southern Hemisphere) but where the temperature was above 0 °C in October (April in the Southern Hemisphere), to exclude high-latitude lakes with permanent ice cover. The highlighted area (15,200,000 km2) has pronounced winter and calcareous soils, spanning the prairie and steppe regions of North America, South America, Europe and Asia. If we assume this region to have a similar surface water distribution to that of southern Saskatchewan, the area occupied by permanent lakes should be between 740,678 km2 (at 1:50,000 scale) and 538,892 km2 (at 1:250,000). If these basins also experienced a decline in CO2 efflux of 100 g C m−2 per summer during the past 15 years (Fig. 1d), global hardwater lakes may have sequestered 74.1 Mt (53.9 Mt at the coarser resolution) more C per summer than they did during the mid-1990s, a change greater than 5% of global efflux from dilute boreal lakes3,4. This value should increase in the future as ice cover declines. Source data

Extended Data Table 1 Physical and chemical characteristics of hardwater lakes of the Qu’Appelle River catchment, Saskatchewan, Canada

PowerPoint slides

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Finlay, K., Vogt, R., Bogard, M. et al. Decrease in CO2 efflux from northern hardwater lakes with increasing atmospheric warming. Nature 519, 215–218 (2015) doi:10.1038/nature14172

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.