7.4. The Carbonate Pump

Key Points:

• Many plankton species make shells or skeletons from calcium carbonate (CaCO₃).

• The process of calcification takes up CO₃^2- from the surface ocean and exports it to the deep.

• Export of CO₃^2- reduces TA and DIC in a 2:1 ratio in the surface, and increases them in the deep, in a process known as the carbonate pump.

• Calcification lowers ocean pH, shifting carbon speciation and causing the release of CO₂ to the atmosphere.

Calcification

Many groups of marine organisms make shells or skeletons from calcium carbonate (CaCO₃). You are probably familiar with bivavles and gastropods that produce the shells you find on the sea shore, and corals that produce massive reef structures found in tropical coastal environments. Some examples you may not have encountered are the planktonic organisms found throughout the surface ocean that produce microscopic CaCO₃ structures. Three groups of marine plankton are responsible the majority of calcification in the open ocean (Fig. 7.25):

• Coccolithophores are small (3-50 μm) phytoplankton that produce intricate calcium carbonate plates called coccoliths. Coccolithophores are a major calcifyer in the open ocean, demonstrated by the fact that coccoliths are the main constituent of massive chalk deposits found in uplifted marine sediments (e.g. the White Cliffs of Dover). The exact purpose of these coccoliths is unclear, although hypotheses include buoyancy control, protection from grazing and light scattering.

• Foraminifera are slightly larger (50-500 μm) unicellular zooplankton which produce calcium carbonate shells called tests. These tests have a distinctive multi-chambered structure, which grows gradually throughout the organism’s life cycle. Some species, like the one pictured below, also create long mineralised spines that are home to symbiotic algae.

• Pteropods are larger-still (0.5-10 mm) multicellular zooplankton who’s name literally translates to ‘wing foot’. Pteropods are gastropods, which (mostly) produce a recognisable calcium carbonate ‘snail shell’ structure, with the soft parts of the body adapted into ‘wing’ that allow them some limited movement through the water.

Fig. 7.25 The three main open-ocean calcifiers. An SEM micrograph of the coccolithophore Emiliania huleyi, and light micrographs of the planktic foraminifera Orbulina universa and a Pteropod. Images from wikimedia commons (coccolithophore, pteropod) and Oscar Branson.

The shells of these organisms are formed from one of two ‘polymorphs’ of CaCO₃ that are stable at Earth surface conditions: calcite and aragonite. Coccolithophores and foraminifera produce calcite, while pteropods produce aragonite.

The formation of these shells is actively controlled by the organism, and involves taking up dissolved Ca²⁺ and CO₃^2- ions from the surrounding water, and using them to build the a mineral structure. This process is known as calcification, and can be written as:

(7.24)\[\mathrm{Ca^{2+} + CO_{3}^{2-} \xrightarrow{calcification} CaCO_{3}}\]

Calcification and CO₂

Removing CO₃^2- from the water and trapping it in mineral form both directly removes DIC from surface water, and alters the speciation of the remaining DIC in the surface ocean, causing CO₂ to be released to the atmosphere, further reducing surface DIC.

Just like organic matter in The Biological Pump, the CO₃^2- trapped in shells is exported from the surface ocean as shell-bearing organisms die and sink. The uptake and export of CO₃^2- directly removes DIC from the surface ocean.

We also saw when considering The Solubility Pump that the DIC species exist in an equilibrium with each other and pH. Removing CO₃^2- causes this speciation adjust, reducing pH because a HCO₃^- dissociates to replace it, releasing H⁺ into the solution. We can re-draw DIC speciation diagram (Fig. 7.10) to see how the abundance of CO₃^2- relates to the concentration of the other species (Fig. 7.26):

Fig. 7.26 Change in DIC speciation as a function of CO₃^2- concentration. This equilbrium is calculated at constant DIC concentration - there is no gas exchange with the atmosphere. If gas exchange were allowed to happen CO₂ would be released to the atmosphere as CO₃^2- is removed from the surface ocean. This is shown in Fig. 7.27.

Removing CO₃^2- from the water causes the fraction of DIC that exists as CO₂ to increase (the equilibrium shifts to the left in Fig. 7.26), causing CO₂ to be released to the atmosphere. Calcification therefore removes more DIC that we would predict based solely on the uptake of CO₃^2- from the surface ocean, because its effect on DIC speciation make it a source of CO₂ to the atmosphere (Fig. 7.27).

Fig. 7.27 The change in pH and DIC caused by the removal of CO₃^2- from the surface ocean by calcification when it is equilibrating with an atmosphere at 400 ppm pCO₂. We can see that DIC is reduced directly by the removal of CO₃^2-, and indirectly by the resulting shift in DIC speciation (Fig. 7.26), causing CO₂ to be released to the atmosphere.

The mechanism of exporting of CO₃^2- from the surface ocean via the formation and sinking of CaCO₃ structures is intuitive, but how is CO₃^2- released back into the deep ocean? The bacterial breakdown that applies to organic matter doesn’t apply here, because there is no energetic or nutritional benefit for the microbes to actively break down CaCO₃. To consider the remineralisation of CaCO₃, we must briefly consider mineral solubility and saturation states.

Saturation State

Just like the DIC speciation equations that drive The Solubility Pump (7.9), the calcification reaction (7.24) has an equilibrium point that is defined by the thermodynamic properties of the reactants and product, and is described by an equilibrium solubility product (\(K_{sp}\)) for the mineral phase:

\[K_{sp} = (\mathrm{[Ca^{2+}][CO_{3}^{2-}]})_{eq}\]

Which is the concentration of Ca²⁺ and CO₃^2- at which the mineral is in equilibrium with the surrounding water, and will neither grow nor dissolve. Note that, just like \(K_1\) and \(K_2\), \(K_{sp}\) varies with temperature, salinity and pressure. From this, we can define the saturation state of seawater with respect to CaCO₃:

\[\Omega = \frac{[Ca^{2+}][CO_{3}^{2-}]}{K_{sp}}\]

Values of \(\Omega > 1\) will favour the formation of CaCO₃, while values of \(\Omega < 1\) will favour dissolution.

Ca is relatively abundant and constant in seawater (~10.2 mmol kg^-1) compared to CO₃^2- (0.1-0.3 mmol kg^-1 at normal ocean surface conditions). This means that the formation and dissolution of CaCO₃ is predominantly controlled by the concentration of CO₃^2- in the water.

The majority of the modern surface ocean is saturated (i.e. \(\Omega > 1\)) with respect to CaCO₃, and so calcification is favoured. For CaCO₃ to dissolve in the deep ocean, the deep ocean must be undersaturated (i.e. \(\Omega < 1\)) with respect to CaCO₃.

CaCO₃ Dissolution

The dissolution of CaCO₃ is driven by a combination of changes in pressure and the impact of the biological pump.

Pressure

The CaCO₃ minerals become more soluble at higher pressure; \(K_{sp}\) increases with depth. The two polymorphs of CaCO₃, calcite and aragonite, also have different \(K_{sp}\) values - calcite is more stable than aragonite, so has a lower \(K_{sp}\). The speciation of DIC is also sensitive to pressure, which causes [CO₃^2-] to decrease with depth. In combination, this means that as a CaCO₃ particle sinks through the ocean it will become less stable as pressure increases, while [CO₃^2-] in the surrounding water simultaneously decreases. This causes the mineral to dissolve when it sinks below the depth at which [CO₃^2-] is lower than \(\frac{K_{sp}}{[Ca^{2+}]}\) (Fig. 7.28). Thus, changes in pressure alone are sufficient to dissolve a large fraction of the CaCO₃ in the deep ocean.

Fig. 7.28 The concentration of CO₃^2- in seawater at which the CaCO₃ minerals calcite and aragonite are in equilibrium with seawater across a range of water depths. The shaded regions are supersaturated with respect to each mineral. The horizontal dashed line shows the [CO₃^2-] for typical deep ocean water. The equilibrium [CO₃^2-] for each mineral both increase with depth, while [CO₃^2-] decreases with depth. The mineral will begin to dissolve below where the dashed line intersects with the solid lines. Calculated at 4 °C and salinity of 35.

Biological Pump

The effects of pressure on solubility and speciation then combine with the biological pump to make CaCO₃ dissolution occur at shallower depths than would be expected from pressure alone. The uptake of CO₂ by photosynthesis in surface waters raises pH and [CO₃^2-], and the release of CO₂ by remineralisation in the deep ocean causes pH and [CO₃^2-] to decrease. In combination with the pressure effects on CaCO₃ solubility, this causes CaCO₃ dissolution to occur at shallower depths (Fig. 7.29)}).

Fig. 7.29 Similar to Fig. 7.28, but showing a real [CO₃^2-] profile for the Southern Atlantic. The action of the biological pump causes [CO₃^2-] to be elevated in surface waters, where CO₂ is being removed by photosynthesis, and elevated at depth, where CO₂ is being released by remineralistion.

Aggregates and Local Environments

Another layer of complexity when considering CaCO₃ is the effect of particle aggregates. Most material leaving the surface ocean is in the form of aggregates - collections of smaller particles that clump together to form larger particles. The respiration of bacteria in these aggregates can drive very high local CO₂ concentrations, dramatically reducing pH and CO₃^2- and creating a highly locally undersaturated environment. This effectively ‘concentrates’ the impact of the biological pump in a way that you would not pick up from bulk water chemistry measurements. A lot of CaCO₃ dissolution occurs within these local environments, which can cause a substantial fraction of CaCO₃ to dissolve much shallower than predicted by pressure and bulk water chemistry alone.

The Carbonate Pump

The export of CaCO₃ from the surface ocean to the deep ocean is known as the carbonate pump. The impact of the carbonate pump on carbon concentration and speciation considered above can also be described in terms of DIC and TA, which describe the concentration and speciation of carbon in the water. The carbonate pump acts to move TA and DIC from the surface to the deep ocean in a 2:1 ratio. This follows from our definitions of DIC and TA:

(7.25)\[\begin{split}\mathrm{DIC} &= {\color{red}\mathrm{[CO_3^{2-}]}} + \mathrm{[HCO_3^-]} + \mathrm{[CO_2^*]} \\ \mathrm{TA} &= {\color{red}\mathrm{2[CO_3^{2-}]}} + \mathrm{[HCO_3^-]} + [\dots] - \mathrm{[H^+]} \\\end{split}\]

Because calcification and dissolution affect TA and DIC in different amounts, it will alter both the concentration and speciation of DIC in the water (Fig. 7.13).

The carbonate pump is inextricably connected to the biological pump, because CaCO₃ production is driven by biological productivity. This means that the carbonate pump also depends on the supply of nutrients to the surface ocean, which fuels biological productivity. However, this connection is not one-way: the carbonate pump can also affect the efficiency of the biological pump by altering the density of sinking particles.

Ballasting

We saw in The Biological Pump that the proportion of primary production that is exported to the deep ocean depends to a large extent on particle sinking speed, which is a function of particle density. Seawater has a density of ~1025 kg m^-3, organic matter has a typical density of ~1030-1500 kg m^-3. Calcite has a density of 2710 kg m^-3. The inclusion of CaCO₃ in particle aggregates will therefore act to increase their density, allowing them to sink faster and increasing the proportion of primary production that is exported to the deep ocean. This means that, even thought calcification is an immediate source of CO₂ to the atmosphere, increasing the efficiency of the biological may offset this effect and make calcification a net (indirect) sink for CO₂. This feedback is known as ballasting, and the amount of dense CaCO₃ relative to organic matter in sinking particles is a critical parameter in determining the efficiency of the biological pump.

Ocean Acidification

In The Solubility Pump we saw that pCO₂ and pH are inversely related, and that elevating atmospheric pCO₂ will cause the surface ocean to become more acidic. This would reduce [CO₃^2-], and therefore the saturation state (Ω) of the surface ocean (Fig. 7.30). Reducing the saturation state has the potential to reduce CaCO₃ production, by marine organisms, and enhance CaCO₃ dissolution at shallow depths, thus reducing the strength of the carbonate pump.

Fig. 7.30 The impact of ocean acidification on surface ocean \(\Omega\) at 0 and 30 °C. Because of the influence of temperature on solubility and DIC speciation, the high latitudes are much more affected by ocean acidification than the tropics. The polar waters may become undersaturated with respect to aragonite as soon as ~2050.

In the immediate term, this could act as a negative feedback on atmospheric pCO₂, because reductions in calcification will reduce the rate of associated CO₂ release from the surface ocean. However, when the effect of CaCO₃ production on ballasting the biological pump is considered, it is less clear whether impact of ocean acidification on the carbonate pump will have a net postitive or negative effect on atmospheric pCO₂. The interaction between ballasting and ocean acidification is a significant source of uncertainty in our predictions of the uptake of CO₂ by the ocean in future.

Modelling the Carbonate Pump

Fig. 7.31 To model the carbonate pump, we need to include the production of CaCO_{3 by biological life, which takes up DIC and TA in the surface ocean and exports it to the deep ocean.}

Calcification is biologically-mediated, and takes up CO₃^-2 from the surface ocean and exports it to the deep. We will therefore model calcification as a function of biological productivity in the surface ocean, assuming that some fraction of biological carbon export (\(f_{calc}\)) is accompanied by calcification:

\[f_{calc} = \frac{CO_{3,calc}^{2-}}{DIC_{org}}\]

The amount of CO₃^2- exported from a surface box is therefore given by:

\[\dv{CO_{3,i}^{2-}}{t} = f_{calc,i} \underbrace{106 \frac{V_i}{\tau^P_i} P_i}_{biological~C~export}\]

From our definitions of TA and DIC (7.25), this modifies TA and DIC in our model in a 2:1 ratio. We therefore model the impact of calcification on DIC as:

(7.26)\[\begin{split}\dv{DIC_L}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{CO_2~exch.} \\ \mathrm{bio.~pump.} \end{bmatrix} - f_{calc,L} 106 \frac{1}{\tau^P_L} P_L \\ \dv{DIC_H}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{CO_2~exch.} \\ \mathrm{bio.~pump.} \end{bmatrix} - f_{calc,H} 106 \frac{1}{\tau^P_H} P_H \\ \dv{DIC_D}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{bio.~pump.} \end{bmatrix} + \left(f_{calc,L} 106 \frac{V_L}{\tau^P_L} P_L + f_{calc,H} 106 \frac{V_H}{\tau^P_H} P_H \right) / V_D\end{split}\]

where the terms in square brackets represents the transport (7.2), gas exchange (7.16) and biological pump (7.22) terms. For TA, the expression is included as:

(7.27)\[\begin{split}\dv{TA_L}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{bio.~pump.} \end{bmatrix} - 2 f_{calc,L} 106 \frac{1}{\tau^P_L} P_L \\ \dv{TA_H}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{bio.~pump.} \end{bmatrix} - 2 f_{calc,H} 106 \frac{1}{\tau^P_H} \\ \dv{TA_D}{t} &= \begin{bmatrix} \mathrm{transport} \\ \mathrm{bio.~pump.} \end{bmatrix} + \left( 2 f_{calc,L} 106 \frac{V_L}{\tau^P_L} P_L + 2 f_{calc,H} 106 \frac{V_H}{\tau^P_H} P_H \right) / V_D\end{split}\]