Figure 1.
Seasonal dynamics of phytoplankton blooms in Lake Volkerak.
(A) Changes in phytoplankton population density (strongly dominated by the cyanobacterium Microcystis) and measured dissolved CO2 concentration ([CO2]) during two consecutive years. The dashed line is the expected dissolved CO2 concentration ([CO2*]) when assuming equilibrium with atmospheric pCO2. Dark shading indicates that the lake is supersaturated with CO2, while light shading indicates undersaturation. (B) Changes in pH, bicarbonate and total DIC concentration. Sampling details are described in Text S1.
Figure 2.
Changes in inorganic carbon chemistry during phytoplankton growth in two chemostat experiments.
Left panels: Chemostat experiment with low pCO2 of 200 ppm in the gas flow and 500 µmol L−1 bicarbonate in the mineral medium. Right panels: Chemostat experiment with high pCO2 of 1,200 ppm in the gas flow and 2,000 µmol L−1 bicarbonate in the mineral medium. Both chemostats were inoculated with Microcystis CYA140. (A, B) Population density (expressed as biovolume) and light intensity penetrating through the chemostat (IOUT), (C, D) dissolved CO2, bicarbonate and carbonate concentrations, (E, F) total DIC concentration and pH, and (G, H) alkalinity (ALK) and concentrations of dissolved inorganic nitrogen (DIN) and phosphorus (DIP). Symbols represent measurements, lines show the model fits. The model and its parameter values are detailed in Text S2.
Figure 3.
Trajectories of dissolved CO2 and population density.
Trajectories predicted by the model for chemostats with (A) low pCO2 of 200 ppm in the gas flow and 500 µmol L−1 bicarbonate in the mineral medium, and (B) high pCO2 of 1,200 ppm in the gas flow and 2,000 µmol L−1 bicarbonate in the mineral medium. The trajectories start from a series of different initial conditions, and all converge to the same equilibrium point. Arrows indicate the direction of the trajectories. The model assumes species parameters specific for Microcystis CYA140, and is detailed in Text S2.
Figure 4.
Steady-state patterns of phytoplankton population density and inorganic carbon chemistry in chemostat experiments.
Steady-state results are shown for 6 chemostats with Microcystis HUB5-2-4 exposed to different pCO2 levels in the gas flow and two different bicarbonate concentrations in the mineral medium (0.5 or 2.0 mmol L−1). (A) Phytoplankton population density (expressed as biovolume), (B) light intensity penetrating through the chemostat (IOUT), (C) dissolved CO2 concentration, (D) bicarbonate concentration, (E) pH, (F) alkalinity, (G) DIC concentration, and (H) carbon sequestration rate. Symbols show the mean (± s.d.) of 5 measurements in each steady-state chemostat, lines show the model fits. For comparison, dashed lines show steady-state patterns predicted for chemostats without phytoplankton. Shading indicates the level of carbon limitation (LC) predicted by the model. The model and its parameter values are detailed in Text S2.
Figure 5.
Steady-state patterns predicted for phytoplankton blooms in low-alkaline lakes.
Steady-state predictions of the model evaluated across a wide range of atmospheric pCO2 levels. (A) Phytoplankton population density (expressed as biovolume), (B) light intensity reaching the lake sediment (IOUT), (C) dissolved CO2 concentration, (D) bicarbonate concentration, (E) pH, (F) alkalinity, (G) DIC concentration, and (H) carbon sequestration rate. Shading indicates the level of carbon limitation (LC). For comparison, dashed lines show steady-state patterns predicted for low-alkaline waters without phytoplankton. The model parameters are representative for eutrophic low-alkaline lakes (ALKIN = 0.5 mEq L−1) dominated by the cyanobacterium Microcystis HUB5-2-4. The model and its parameter values are detailed in Text S2 and Text S3.
Figure 6.
Contour plots of phytoplankton blooms predicted for different pCO2 levels and alkalinities.
Model predictions of (A) the level of carbon limitation, and (B) phytoplankton population density (expressed as biovolume, in mm3 L−1). The vertical solid line represents the present-day atmospheric CO2 level of ∼400 ppm, while the vertical dashed line shows the atmospheric CO2 level of 750 ppm predicted for the year 2150 by the RCP6 scenario of the Fifth Assessment Report of the IPCC. The model predictions are based on steady-state solutions across a grid of 40×50 = 2,000 simulations, using the model and parameter values detailed in Text S2 and Text S3.
Figure 7.
Sensitivity of the model predictions to variation in phytoplankton traits.
Contour plots of the level of carbon limitation (left panels) and steady-state phytoplankton population density (right panels, expressed as biovolume, in mm3 L−1) predicted for different atmospheric pCO2 levels and phytoplankton traits. The phytoplankton traits are (A, B) the half-saturation constant for CO2 uptake (HCO2), (C, D) the half-saturation constant for bicarbonate uptake (HHCO3), (E, F) the maximum CO2 uptake rate (uMAX, CO2), and (G, H) the cellular N:C ratio (cN). The model considers a low-alkaline lake (ALKIN = 0.5 mEq L−1). Vertical lines represent atmospheric CO2 levels of 400 ppm (present-day) and 750 ppm (predicted for the year 2150 by the RCP6 scenario of the IPCC). Horizontal dotted lines represent our default parameter values. The contour plots are based on steady-state solutions across a grid of 40×50 = 2,000 simulations.
Table 1.
Normalized sensitivity coefficients of selected model parameters at atmospheric CO2 levels of 400 ppm (SC400) and 750 ppm (SC750).
Figure 8.
Sensitivity of the model predictions to variation in lake properties.
Contour plots of the level of carbon limitation (left panels) and steady-state phytoplankton population density (right panels, expressed as biovolume, in mm3 L−1) predicted for different atmospheric pCO2 levels and lake properties. The lake properties are (A, B) lake depth (zMAX), (C, D) CO2 gas transfer velocity (v), (E, F) DIC concentration of the influx ([DIC]IN), and (G, H) salinity (Sal). The model considers a low-alkaline lake (ALKIN = 0.5 mEq L−1). Vertical lines represent atmospheric CO2 levels of 400 ppm (present-day) and 750 ppm (predicted for the year 2150 by the RCP6 scenario of the IPCC). Horizontal dotted lines represent our default parameter values. In (E, F), the dotted line indicates equilibrium with the atmospheric CO2 pressure. The contour plots are based on steady-state solutions across a grid of 40×50 = 2,000 simulations.