Article pubs.acs.org/est
High-Magnesium Calcite Dissolution in Tropical Continental Shelf Sediments Controlled by Ocean Acidification R. R. Haese,*,† J. Smith,‡ R. Weber,‡ and J. Trafford‡ †
School of Earth Sciences, The University of Melbourne, Parkville, Victoria 3010, Australia Geoscience Australia, General Post Office Box 378, Canberra, Australian Capital Territory 2609, Australia
‡
S Supporting Information *
ABSTRACT: Increases in atmospheric CO2 cause the oceanic surface water to continuously acidify, which has multiple and profound impacts on coastal and continental shelf environments. Here we present the carbonate mineral composition in surface sediments from a range of continental shelf seabed environments and their current and predicted stability under ocean acidifying conditions. Samples come from the following four tropical Australian regions: (1) Capricorn Reef (southern end of the Great Barrier Reef), (2) the Great Barrier Reef Lagoon, (3) Torres Strait, and (4) the eastern Joseph Bonaparte Gulf. Beyond the nearshore zone, these regions typically have a carbonate content in surface sediments of 80 wt % or more. The abundance of high-magnesium calcites (HMC) dominates over aragonite (Arag) and low-magnesium calcite (LMC) and constitutes between 36% and 50% of all carbonate. HMC, with a magnesium content larger than 8−12 mol %, is more soluble than both Arag and LMC, and the solubility of HMC positively correlates with its magnesium concentration. From the solubility data of Plummer and Mackenzie (Am. J. Sci. 1974, 274, 61−83), 95% of HMC in the four regions is presently in metastable equilibrium relative to global mean tropical sea surface water. HMC is predicted to become destabilized in the four regions between 2040 and 2080 AD, with typical HMC decline rates between 2% and 5% per year. The range of respective estimated carbonate dissolution rates is expected to exceed current continental shelf carbonate accumulation rates, leading to net dissolution of carbonate during the period of HMC decline. In a geological context, the decline in HMC in tropical continental shelf environments is a global event triggered by reaching below-equilibrium conditions. The characteristic change in carbonate mineral composition in continental shelf sediments will serve as a geological marker for the proposed Anthropocene Epoch.
■
H 2O + CO2 + CO32 − → 2HCO3−
INTRODUCTION Atmospheric carbon dioxide (CO2) concentrations have increased by 40% from preindustrial levels, primarily from fossil fuel emissions.2 Rising atmospheric CO2 is tempered by oceanic uptake, which accounts for a third of anthropogenic carbon added to the atmosphere.3 Oceanic CO2 uptake, however, reduces pH and causes shift in seawater−carbonate chemistry in a process referred to as ocean acidification.4 Ocean acidification directly impacts a wide range of carbonate-producing marine organisms and has raised serious concerns about the potential effects on pelagic5,6 and benthic7,8 ecosystems. Studies on corals9 and pteropod molluscs10 from the tropical Australian continental shelf suggest ocean acidification is already detectable for recent decades. In shallow water areas of the inner continental shelf, the environmental impacts of ocean acidification may be exacerbated by other environmental and anthropogenic stressors such as episodic ocean warming leading to coral bleaching, overfishing, and land-based sources of pollution, leading to considerable economic and societal consequences.11 The reaction of CO2 with seawater reduces the availability of carbonate ions (eq 1), shifting the carbonate equilibrium and resulting in a lower saturation state with respect to calcium carbonate (CaCO3) minerals: © 2014 American Chemical Society
(1)
The saturation state (Ω) is commonly defined as the ratio of the concentration product to the stoichiometric solubility product (K′sp).12 A mineral is supersaturated when Ω > 1.0 and undersaturated when Ω < 1.0. For calculating the saturation state for calcite where calcium is partly replaced by magnesium, here referred to as high-magnesium calcite (HMC), the ratio of the ion activity product (IAP) to the solubility product based on observed apparent HMC solubilities (K′HMC) needs to be used and can be written as13 Ω HMC =
(a Mg 2+)x (aCa 2+)1 − x aCO32− K ′HMC
(2)
where x refers to the mole percent MgCO3. This conclusion has become a convention and follows a long controversy and discussion on the definition of a solubility product for a carbonate mineral with a given calcium-to-magnesium ratio Received: Revised: Accepted: Published: 8522
March 31, 2014 June 25, 2014 June 27, 2014 June 27, 2014 dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
Figure 1. Map of northeastern Australia including the continental shelf, showing sample locations of this study together with carbonate content of all samples registered in the Geoscience Australia MARS database. GBR, Great Barrier Reef; JBG, Joseph Bonaparte Gulf.
also deposit HMC of varying composition.12,18,19 In many coral reefs, HMC formed as the internal skeleton of coralline algae is part of the reef-stabilizing cement and thereby contributes to ecosystem sustainability beyond the life of its producing organism.20 Oceanic uptake of atmospheric CO2 primarily occurs in surface seawater extending over the continental shelf. Here, seawater CO2 concentration continuously increases and the saturation state with respect to carbonate minerals decreases (eq 2). HMC will be the first mineral phase subject to undersaturated conditions and is expected to dissolve because of its higher solubility than aragonite and calcite.13,14,16,20 Under the present rate of CO2 emissions, model calculations show that tropical surface seawater could become undersaturated with respect to HMC containing ≥12 mol % MgCO3 during this century.16 Dissolution of carbonate minerals consumes CO2 and produces alkalinity, and as a result, the pH and saturation state with respect to carbonate minerals increases. However, this potential buffer mechanism has been shown to be insignificant for two reasons. First, the relatively high rate of mixing with the deep open ocean rapidly dilutes alkalinityenriched continental shelf water, and second, the extent of carbonate dissolution is relatively small.16,21 As a result, a significant slowing or termination of ocean acidification through the oceanic buffer capacity is not expected. In this study we investigate the abundance and composition of carbonate minerals in surface sediments from four tropical regions of the Australian continental shelf. The observed carbonate mineralogy is compared to the current saturation state of carbonate minerals, and changes in mineral composition controlled by ocean acidification over the next 100 years are predicted.
exposed to water with variable ratios in the dissolved calcium to magnesium activities; see, for example, refs 12−14. It is further important to note that aragonite and HMC are metastable under current Earth surface conditions. However, as we report mineral stability in terms of the state of saturation (see above), we will use the terms under- and supersaturated for these minerals as well. Carbonate-rich sediments on the continental shelf represent a major CaCO3 reservoir that can rapidly react to the decreasing saturation state of seawater with respect to carbonate minerals. Three principal carbonate minerals are found in tropical continental shelf sediments: aragonite (Arag), low-magnesium calcite (LMC; 4 mol % MgCO3). Low- and highmagnesium calcites are often denoted simply as calcite and magnesium calcite, respectively. Currently, tropical surface seawaters are supersaturated with respect to LMC (−log K′ = 8.48) and Arag (−log K′ = 8.30).13,15 Differences in the solubility of HMC have been observed depending on sample preparation (see discussion in refs 12 and 13). Plummer and Mackenzie1 derived solubility data from experiments involving samples with minimal treatment, and on the basis of their results, HMC with 15 mol % MgCO3 is currently in equilibrium with tropical seawater (Ω = 1, −log K′ = 7.63).16 As HMC solubility correlates positively with magnesium concentration, HMC with a higher MgCO3 concentration is currently not thermodynamically favorable in tropical waters. The rate of carbonate dissolution can be described (e.g., ref 17) as
R = k(1 − Ω)n
(3)
where R is the dissolution rate, n is the reaction order, and k is the rate constant. The latter is strongly dependent on the reactive surface area, which decreases as dissolution progresses. Carbonate in marine sediments is predominantly of biogenic origin. The most common calcifiers forming HMC with significant mole percentages of MgCO3 are the red coralline algae, benthic foraminifera, bryozoans, and echinoderms, but other groups of organisms such as crustaceans, molluscs, annelid worms, calcareous sponges, barnacles, and brachiopods
■
MATERIALS AND METHODS This study utilizes previously collected sediment samples and data that have been archived at Geoscience Australia. A total of 216 surface sediment samples from the Australian continental shelf extending between 10° and 23° S were analyzed. Sample 8523
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
mmol Kg−1), Mg (8.892 mmol Kg−1), and CO32− (0.0099 mmol Kg−1) were calculated from average seawater concentrations,28 activity coefficients for Ca and CO32− at 25 °C,29 and an activity coefficient for Mg based on an estimated Mg/Ca activity ratio of 5.13 The solubility products for LMC and Arag are widely accepted with −log K′LMC = 8.48 and −log K′Arag = 8.3. In contrast, published K′sp values for HMC as a function of mole percent MgCO3 have been a matter of debate, and a good review of the discussion is provided by Morse et al.13 There are essentially two experimental solubility curves for biogenic magnesium calcites, and the curves are distinguished by the extent of preparation of the experimental materials. While it is not fully understood which solubility curve is most representative in natural environments, most studies12−14 suggest the use of HMC solubility data by Plummer and Mackenzie.1 For this reason, the Plummer and Mackenzie1 correlation (−log K′HMC = −0.063x + 8.586, where x = mol % MgCO3, for the range from 10−20 mol % MgCO3) was used in this study. Modeling the progression of ocean acidification requires a global biogeochemical model accounting for carbon fluxes across reservoir boundaries and changes in carbon, including carbonate ion reservoir sizes, over time. Here we use the model output derived by Gangstø et al.,30 which is based on the A2 scenario (IPCC SRES A2) for the period 2000−2100. The study provides a curve for the change in ΩArag in equatorial regions as defined by the region between 20°N and 20°S until 2100 AD. Present-day values of ΩHMCx, with x being the variable mole percent MgCO3, were calculated from the IAP and the HMC solubility data of Plummer and Mackenzie.1 The predicted ΩHMCx curves were then derived from the predicted global ΩArag curve for equatorial areas30 and the ratio of ΩHMCx to ΩArag.
locations and year of sampling are given in Supporting Information. The samples were collected from Torres Strait in 2004 (n = 77), the eastern Joseph Bonaparte Gulf (JBG) in 2009 (n = 63), the Great Barrier Reef (GBR) Lagoon between 2003 and 2005 (n = 45), and the Capricorn Reef in 2004 and 2005 (n = 31) (Figure 1). All samples were taken from surface sediments in 2000 km. Narrow inter-reef channels separating reef edifices are important geomorphic features, and northeastern trade winds promote an overall northward-trending water circulation.22 The Capricorn Reef is located on the Tropic of Capricorn at the southern end of the GBR at significant distance and geomorphologically separated from the coast by deep channels. Torres Strait is located at the northern end of the Cape York Peninsula and forms a shallow-water (typically 15−25 m deep) low-relief plain with extensive seagrass areas.23 Samples from the eastern JBG come from a north−south transect west of Melville Island where a range of seabed environments including outer-shelf deep-water channels, bank environments interspersed by channels, and inner-shelf plains and shallow channels are found.24 Dried and archived samples were analyzed by X-ray diffraction (XRD) at Geoscience Australia. Sample metadata are stored in the MARS database. The carbonate content of all samples was measured at Geoscience Australia by the “carbonate bomb” method.25 The estimated accuracy and precision is equal to or better than 5%. The mineral composition of powdered samples was analyzed on a Siemens D-500 X-ray diffractometer, from 2° to 70° 2θ, using a Cu anode X-ray tube at 40 kV and 20 mA. Minerals were identified by Bruker Diffracplus Eva and quantified by use of Siroquant V3 software. The mole percent MgCO3 in the HMC was calculated by the method proposed by Chave,19 with fluorite as an internal standard for all samples except for samples from Torres Strait. For the latter samples, aragonite was initially used as an internal standard. The average peak difference between aragonite and fluorite was later used to recalibrate against fluorite. The relationship of Goldsmith et al.26 was employed to relate the peak shift of the d104 peak to the Mg content in calcite of the specific sample. The d space values were derived from the Siroquant and Eva programs, and comparison of the two showed agreement equivalent to ±0.5 mol % MgCO3, which is an acceptable margin of error. A total of 19 surface seawater samples were taken along a north−south transect in the eastern JBG during the SOL4934 expedition onboard the RV Solander. Temperature and salinity at sampling sites were recorded on a thermosalinograph. Total alkalinity (TA) was determined by Gran titration and by the Solver method,27 and pH was determined by the Tris buffer method and reported on the total hydrogen ion scale.28 Temperature, salinity, TA, and pH were used to calculate the state of saturation for aragonite by use of the CO2SYS program28 and the state of saturation of HMC (see below). The state of saturation, Ω, for Arag, LMC, and HMC was calculated according to eq 2. Ionic activities for Ca (1.778
■
RESULTS AND DISCUSSION Carbonate Content and Carbonate Mineral Composition. Carbonate content in continental shelf sediments of northern Australia varies significantly (Supporting Information) and is largely controlled by the discharge of siliciclastic detrital sediment by rivers and the geomorphology of the seafloor. The GBR Lagoon is part of the inner shelf and is characterized by high concentrations of terrestrial-derived muds and a carbonate content of 80%.22 Accordingly, the carbonate content range of the GBR Lagoon sample set is large, varying between 24 and 94 wt % with a mean of 80 wt % carbonate. In comparison, the carbonate content in Capricorn Reef samples ranges from 43 to 96 wt % with a mean of 87 wt %. Similarly, the carbonate content in Torres Strait samples varies between 72 and 95 wt % with a mean of 88 wt %. Samples from the eastern JBG show the highest proportion of detrital sediment, with a carbonate content in the range between 26 and 92 wt % and a mean concentration of 61 wt %. All samples contained typical siliciclastic minerals including kaolin, quartz, and muscovite (data not shown) as well as HMC, LMC, and Arag. The individual carbonate minerals were normalized to a percentage of total carbonate and are presented in Figure 2. Torres Strait and Capricorn Reef samples show higher mean proportions of HMC and a lower spread in the carbonate mineral distribution as compared to the GBR Lagoon and the eastern JBG samples. Torres Strait and Capricorn Reef have a mean HMC proportion of 45% (standard deviation σ = 6) and 50% (σ = 6), respectively, whereas the GBR Lagoon and 8524
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
Figure 3. Histograms showing the distribution of MgCO3 content in high-magnesium calcites in the four study areas.
Figure 2. Ternary diagrams showing proportions of high-magnesium calcite, low-magnesium calcite, and aragonite for the four study areas.
Lagoon, all study areas show a maximum magnesium content in HMC of 15.25 mol % MgCO3. This observation will be further discussed in the context of current and future HMC solubility. Solubility of High-Magnesium Calcite. On the basis of the solubility estimate for biogenic HMC with little treatment by Plummer and Mackenzie,1 surface seawater at 25 °C is currently in equilibrium with HMC containing 15.2 mol % MgCO3. This calculated equilibrium value agrees very well with an observed maximum MgCO3 content of 15.2 mol % in HMC in 95% of our samples. This suggests HMC with a higher MgCO3 content is currently undersaturated and therefore absent in most situations. However, these model calculations predict global changes in carbonate mineral stability and do not account for local or regional differences. Regional water chemistry data from areas of interest for this study show a high degree of agreement with the described global stability of HMC in tropical regions: Annual mean sea surface temperature varies between the Tropic of Capricorn and Torres Strait between 25 and 28 °C,35 and this temperature variability leads to a state of saturation for 15 mol % HMC, ΩHMC‑15, between 0.96 and 1.07 for 2200 μmol·kg−1 total alkalinity and pH 8.1 in seawater. The influence of upwelling events exposing the continental shelf temporarily to colder and more acidic water is considered to be minor or even negligible, given the shallow sample depths. For example, warm surface water outside the northern GBR is separated from deep water through a high-salinity layer typically located at a depth of 150 m. Only during the peak period of an upwelling event does the high-salinity layer extend over the shelf break.36 Surface water samples collected within the eastern JBG showed a north− south temperature gradient between 29.2 and 27.2 °C and median, minimum, and maximum ΩHMC‑15 values of 1.19, 0.94, and 1.27, respectively. Similarly, surface water with a temperature of 28 °C at Lizard Island (14°40′08″ S, 145°27′34″ E) in the northern part of the GBR has an average ΩArag of 4.2637 and a respective ΩHMC‑15 of 1.19. Reef flat waters, however, show large diurnal variability in carbonate water composition and respective carbonate mineral saturation in response to net
the eastern JBG have mean HMC values of 35% (σ = 10) and 36% (σ = 9), respectively. Multiple factors may contribute to the differences in carbonate mineral composition in surface sediments. Here, it is likely due to the larger size of the sample area and the greater diversity in seabed morphology found in the GBR Lagoon and the eastern JBG, which lead to a larger range of carbonate mineral composition as compared to Torres Strait and the Capricorn Reef. Importantly, the calculated HMC:LMC:Arag ratio for the four tropical continental shelf regions is 42:38:20, highlighting the dominance of HMC in these tropical continental shelf regions. The dominance of HMC was also observed in sediments from the Bahamas,31 but the derived ratio differs significantly from a global neritic ratio of 24:13:6313 based on data compiled earlier.32 Carbonate mineral solubility is temperature-dependent and solubility increases with decreasing temperatures. In tropical regions, higher water temperatures result in a higher degree of carbonate mineral stability and a larger proportion of HMC. Conversely, HMC is absent in polar regions due to undersaturation at cooler water temperatures. The higher degree of carbonate mineral stability in tropical waters also reflects a higher biogenic calcification rate since a linear correlation between the degree of (aragonite) saturation and the calcification rate has been found for benthic15,33,34 and planktonic5,6 calcifiers. As the solubility of HMC depends on its magnesium content, the mole percent MgCO3 in HMC was determined for each sample. Histograms showing the relative frequency of the magnesium content in increments of 0.25 mol % MgCO3 for each of the four study areas are shown in Figure 3. Samples from Torres Strait and eastern JBG show a unimodal distribution, whereas the sample distribution of Capricorn Reef and the GBR Lagoon is bimodal. HMC in Torres Strait has a mean magnesium content of 14.5 mol % MgCO3 (σ = 0.4), whereas the magnesium content in HMC in the other three areas ranges between 13.6 and 13.8 mol % MgCO3 (σ = 0.7−1.0). Except for a small subset of samples from the GBR 8525
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
1, before the year 2100 AD, which is in agreement with earlier estimates.16 Our results further show the solubility of HMC drops below equilibrium in increments of 1 mol % MgCO3 every 20−30 years. If it is assumed that the decrease in HMC is purely a function of mineral stability, the rate of HMC decrease as a proportion of the carbonate content and the rate of HMC dissolution can be estimated by combining the predicted changes in saturation state over time and the derived total carbonate content and carbonate mineral concentrations. Figure 5 shows the
community calcification. At Lady Elliot Island (24°6′50″ S, 152°42′40″ E) in the southern part of the GBR, seasonal median ΩArag varies between 3.0 and 3.5, but a total range in ΩArag between 1.13 and 6.46 was observed.38 The latter converts to a range in ΩHMC‑15 between 0.32 and 1.80, which gives evidence for short-term metastability of HMC. A significant fraction of carbonate deposited at the sediment surface is dissolved through the mineralization of organic matter. Organic matter degradation by aerobic respiration leads to proton production and has been shown to be quantitatively most important for carbonate dissolution in shallow continental shelf39 and deeper marine sediments.40 Apart from the organic matter deposition rate, it is primarily the oxygen flux to the zone of aerobic respiration controlling proton production and associated carbonate dissolution. These primary process controls on carbonate dissolution in sediments are linked to ocean acidification, as it is likely that the rate constant contributing to the HMC dissolution rate (eq 3) in sediments decreases as HMC becomes increasingly exposed to undersaturated bottom water conditions before reaching the zone of aerobic respiration. When we compare the solubility of biogenic HMC-15 with minimal treatment to aragonite, we find a K′HMC‑15:K′Arag ratio of 4.4. In comparison, a K′HMC‑15:K′Arag ratio of 1.2 is determined when biogenic HMC-15 solubility derived from cleaned and annealed samples is compared to aragonite as suggested earlier.13 Furthermore, the solubility of cleaned and annealed samples predicts HMC with a much higher MgCO3 content to be stable under current tropical surface seawater conditions, which is not observed. This difference gives further support for the Plummer and Mackenzie1 solubility data under natural conditions. Predicted Stability of High-Magnesium Calcite over the Next 100 Years. Saturation curves for HMC with variable mole percent MgCO3, Arag, and LMC in equatorial surface water until the end of this century are shown in Figure 4. While LMC and Arag remain highly supersaturated, HMC with 12− 15 mol % magnesium carbonate drops below equilibrium, Ω ≤
Figure 5. Predicted decrease in high-magnesium calcite concentration in surface sediments in the four study areas. GBR, Great Barrier Reef; JBG, Joseph Bonaparte Gulf.
calculated relative decrease in HMC in increments of 0.25 mol % MgCO3. It shows almost all of the HMC decrease will occur between 2040 and 2080 AD in all four regions. Torres Strait is a precursor relative to the other regions due to the high mean MgCO3 content of 14.5 mol % in HMC. Based on the linear regression through the steepest concentration gradient over 20−30 years, the decrease in HMC for Torres Strait (2040−2060 AD) is 5% per year, for the GBR Lagoon (2040− 2070 AD) and the eastern JBG (2050−2080 AD), 3% per year, and for the Capricorn Reef (2035−2075 AD) 2% per year. Given the ubiquitous distribution of HMC in tropical and subtropical continental shelf sediments and the predicated measurable changes at time intervals on the order of 5 years post-2040 AD, the decline in maximum mole percent MgCO3 in HMC and the overall decline in HMC content in surface sediments may serve as indicators of ocean acidification. The above derived changes in the mineral composition of marine sediments are very rapid and include the assumption mineral dissolution kinetics do not limit the reaction rate. The latter depends on mineral-specific kinetic properties including the reactive surface area (see eq 3), which are not taken into account. The reactive surface area decreases during dissolution and may at some point limit the dissolution rate. With the above assumptions, carbonate dissolution rates controlled by ocean acidification can be calculated from the derived results if further assumptions are made about sediment porosity and the penetration depth of dissolution. Both parameters are highly variable in continental shelf sediments, depending on factors such as the energy regime of the depositional environment and carbon mineralization and bioturbation rates.40 For the purpose
Figure 4. Saturation state of calcite, aragonite, and high-magnesium calcites with variable concentrations of MgCO3. The horizontal black line at Ω = 1 represents fluid−mineral equilibrium. 8526
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
The presented results suggest the complete decline of HMC in surface sediments of the tropical and subtropical continental shelf between 2040 and 2080 AD based on mineral stability considerations only. Predicting dissolution rates over upcoming decades via carbonate dissolution kinetics is currently uncertain, mostly due to an unknown decline in the rate constant. However, reaction kinetics appear not to have limited HMC dissolution up to the present, since HMC in current surface sediments is either in equilibrium or supersaturated. In the broader context of impacts driven by the anthropogenic increase in atmospheric CO2, the rapid decline in HMC serves as an example of an event rather than a trend triggered by exceeding a threshold, in this case by reaching the state of mineral undersaturation. Because the predicted change is characteristic for the global tropical continental shelf and occurs as an event on the geological time scale, the decline in HMC may serve as one geological marker for the onset of the proposed Anthropocene Epoch.43
of estimating the order of magnitude of enhanced carbonate dissolution controlled by ocean acidification based on mineral stability only, a minimum and a maximum carbonate dissolution rate is calculated as follows: The minimum rate uses dissolution penetration depth of 2 cm, porosity of 70%, average mineral density of 2.65 g·cm−3, HMC decline of 2% per year, carbonate content of 80%, and HMC proportion of 40% of all carbonate. The maximum rate uses dissolution penetration depth of 5 cm, porosity of 40%, HMC decline rate of 5% per year, and the remaining parameters as above. The resulting range in enhanced carbonate dissolution rate is between 100 and 1300 g·m−2·year−1, which is in the range of carbonate dissolution rates observed so far.41 The derived HMC dissolution rates are also very similar to current carbonate accumulation rates, estimated as 1200, 250, and 30 g·m−2·year−1 in coral reef complexes, carbonate banks, and carbonate shelf environments, respectively.42 Given that the estimated carbonate dissolution range is equivalent to the given accumulation range, no net carbonate accumulation on the continental shelf during the period of HMC decline may be possible. Alternatively, the HMC dissolution rate can be calculated by use of eq 3; however, two uncertainties are introduced. First, estimates of the initial rate constant for bulk samples are uncertain, as the grain size and composition of biogenic carbonate fragments are highly variable between samples. Second, the decrease in rate constant over decades is critical and equally uncertain. This is illustrated by the following example: eq 3 is applied for HMC-13.5 with incremental time steps of 1 year. Sediment properties are chosen as above for the minimum carbonate dissolution rate, and ΩHMC‑13.5 is allowed to change over time as derived in this study. A reaction order of 3.5 is used.17 The rate constant is assumed to decline by an order of magnitude each 10 years, t10(k) = 10, and each 20 years, t10(k) = 20, starting with the first year of undersaturation. In the case of t10(k) = 10, low rate constants and low dissolution rates, respectively, are reached much earlier, and consequently HMC dissolution is much lower. In contrast, in the case of t10(k) = 20, the dissolution period is drawn out and much more extensive. Indeed, complete dissolution of HMC before 2110 AD is suggested (Figure 6).
■
ASSOCIATED CONTENT
S Supporting Information *
One table listing sample locations, depths, figure showing median, lower and upper ranges of carbonate content from four material is available free of charge via the pubs.acs.org.
■
and dates and one quartile, and total sample sets. This Internet at http://
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank crews for shipboard sampling and personnel at Geoscience Australia for archiving samples and maintaining the MARS sample database. Liz Webber and Murray Woods from Geoscience Australia provided XRD analysis and preparation of maps, respectively. We thank Scott Nichol and Lynda Radke from Geoscience Australia and three anonymous reviewers for their comments on an earlier manuscript version. This publication received permission for publication by the CEO of Geoscience Australia.
■
REFERENCES
(1) Plummer, L. N.; Mackenzie, F. T. Predicting mineral solubility from rate data: Application to the dissolution of magnesian calcites. Am. J. Sci. 1974, 274, 61−83. (2) IPCC. Climate Change 2013: The Physical Science Basis; Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change 2013; Cambridge University Press: Cambridge, U.K., and New York, 2013. (3) Sabine, C. L.; et al. The oceanic sink for anthropogenic CO2. Science 2004, 305 (5682), 367−371. (4) Doney, S. C.; Fabry, V. J.; Feely, R. A.; Kleypas, J. A. Ocean acidification: The other CO2 problem. Annu. Rev. Mar. Sci. 2009, 1 (1), 169−192. (5) Riebesell, U.; Zondervan, I.; Rost, B.; Tortell, P.; Zeebe, R.; Morell, F. M. M. Reduced calcification of marine plankton in response to increased atmospheric CO2. Nature 2000, 407, 364−367. (6) Moy, A. D.; Howard, W. R.; Bray, S. G.; Trull, T. W. Reduced calcification in modern Southern Ocean planktonic foraminifera. Nat. Geosci. 2009, 2, 276−280.
Figure 6. Predicted changes in HMC concentration for two rate constants, k. The rate constants t10(k) = 10 and t10(k) = 20 decrease by an order of magnitude each 10 and 20 years, respectively. See text for other values used in calculation. 8527
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
(27) Dickson, A. G.; Sabine C. L.; Christian, J. R., Eds. Guide to best practices for ocean CO2 measurements; PICES Special Publication 3; North Pacific Marine Science Organization: Sidney, BC, Canada, 2007. (28) Pierrot, D.; Lewis, E.; Wallace, D. W. R. MS Excel program developed for CO2 systems calculations; ORNL/CDIAC-105a; Carbon Dioxide Information Analysis Centre, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN, 2006; DOI: 10.3334/ CDIAC/otg.CO2SYS_XLS_CDIAC105a. (29) Plummer, L. N.; Sundquist, E. T. The individual ion activity coefficients of calcium and carbonate in sea water at 25 °C and 35 ‰ salinity, and implications to the agreement between apparent and thermodynamic constants for calcite and aragonite. Geochim. Cosmochim. Acta 1982, 46, 247−258. (30) Gangstø, R.; Gehlen, M.; Schneider, B.; Bopp, L.; Aumont, O.; Joos, F. Modeling the marine aragonite cycle: changes under rising carbon dioxide and its role in shallow water CaCO3 dissolution. Biogeosciences 2008, 5, 1057−1072. (31) Morse, J. W.; Zullig, J. J.; Bernstein, L. D.; Millero, F. J.; Milne, F.; Mucci, A.; Choppin, G. R. Chemistry of calcium carbonate-rich shallow-water sediments in the Bahamas. Am. J. Sci. 1985, 285, 147− 183. (32) Land, L.S. Diagenesis of skeletal carbonates. J. Sed. Res. 1967, 37 (3), 914−930. (33) Leclercq, N.; Gattuso, J.-P.; Jaubert, J. CO2 partial pressure controls the calcification rate of a coral community. Global Change Biol. 2000, 6, 329−334. (34) Leclercq, N.; Gattuso, J.-P. Primary production, respiration, and calcification of a coral reef mesocosm under increased CO2 partial pressure. Limnol. Oceanogr. 2002, 47, 558−564. (35) Kuchinke, M.; Tilbrook, B.; Lenton, A. Seasonal variability of aragonite saturation state in the Western Pacific. Mar. Chem. 2014, 161, 1−13. (36) Andrews, J. C.; Gentien, P. Upwelling as a source of nutrients for the Great Barrier Reef ecosystem: A solution to Darwin’s question? Mar. Ecol.: Prog. Ser. 1982, 8, 257−269. (37) Diaz-Pulido, G.; Nash, M. C.; Anthony, K. R. N.; Bender, D.; Opdyke, B. N.; Reyes-Nivia, C.; Troitsch, U. Greenhouse conditions induce mineralogical changes and dolomite accumulation in coralline algae on tropical reefs. Nat. Commun. 2014, 5, No. 3310, DOI: 10.1038/ncomms4310. (38) Shaw, E. C.; McNeil, B. I.; Tilbrook, B. Impacts of ocean acidification in naturally variable coral reef flat ecosystems. J. Geophys. Res. 2012, 117, No. C03038, DOI: 10.1029/2011JC007655. (39) Burdige, D. J.; Zimmerman, R. C.; Hu, X. Rates of carbonate dissolution in permeable sediments estimated from pore-water profiles: the role of sea grass. Limnol. Oceanogr. 2008, 53, 549−565. (40) Pfeiffer, K.; Henson, C.; Adler, M.; Wenzhöfer, F.; Weber, B.; Schulz, H. D. Modeling of subsurface calcite dissolution, including the respiration and reoxidation process of marine sediments in the regional of the equatorial upwelling off Gabon. Geochim. Cosmochim. Acta 2002, 66, 4247−4259. (41) Andersson, A. J.; Gledhill, D. Ocean acidification and coral reefs: Effects of breakdown, dissolution, and net ecosystem calcification. Annu. Rev. Mar. Sci. 2013, 5, 321−348. (42) Milliman, J. D.; Droxler, A. W. Neritic and pelagic carbonate sedimentation in the marine environment: Ignorance is not bliss. Int. J. Earth Sci. 1996, 85, 496−504. (43) Zalasiewicz, I.; Williams, M.; Steffen, W.; Crutzen, P. The new world of the Anthropocene. Environ. Sci. Technol. 2010, 44 (7), 2228− 2231.
(7) Hoegh-Guldberg, O.; et al. Coral reefs under rapid climate change and ocean acidification. Science 2007, 318, 1737−1742. (8) Anthony, K. R. N.; Kline, D. I.; Diaz-Pulido, G.; Dove, K. S.; Hoegh-Guldberg, O. Ocean acidification causes bleaching and productivity loss in coral reef builiders. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17442−17446. (9) Wei, G.; McCulloch, M. T.; Mortimer, G.; Deng, W.; Xie, L. Evidence for ocean acidification in the Great Barrier Reef of Australia. Geochim. Cosmochim. Acta 2009, 73, 2332−2346. (10) Roger, L. M.; Richardson, A. J.; McKinnon, A. D.; Knott, B.; Matear, R.; Scadding, C. Comparison of the shell structure of two tropical Thecosomata (Creseis acicula and Diacavolinia longirostris) from 1963 to 2009: Potential implications of declining aragonite saturation. ICES J. Mar. Sci. 2012, 69, 465−474. (11) Hood, M.; Broadgate, W.; Urban, E.; Gaffey, O. Ocean Acidification: A Summary for Policymakers; Second Symposium on the Ocean in a High-CO2 World, Scientific Committee on Oceanic Research, Monaco, 2009. (12) Geochemistry of Sedimentary Carbonates. In Developments in Sedimentology; Morse, J., Mackenzie, F. T., Eds.; Elsevier: Amsterdam, 1990; Vol. 48. (13) Morse, J. W.; Andersson, A. J.; Mackenzie, F.T. Initial responses of carbonate-rich shelf sediments to rising atmospheric pCO2 and “ocean acidification”: Role of high Mg-calcites. Geochim. Cosmochim. Acta 2006, 70 (23), 5814−5830. (14) Mackenzie, F. T.; et al. Magnesian calcites: Low-temperature occurrence, solubility and solid-solution behaviour. In Carbonates: Mineralogy and Chemistry; Reviews in Mineralogy and Geochemistry, Vol. 11; Reeder, R. J., Ed.; Mineralogical Society of America: Chantilly, VA, 1983; pp 97−143. (15) Kleypas, J. A.; et al. Geochemical consequences of increased atmospheric carbon dioxide on coral reefs. Science 1999, 284 (5411), 118−120. (16) Andersson, A. J.; Mackenzie, F. T. Revisiting four scientific debates in ocean acidification research. Biogeosciences 2012, 9 (3), 893−905. (17) Walter, L. M.; Morse, J. W. The dissolution kinetics of shallow marine carbonates in sea water: A laboratory study. Geochim. Cosmochim. Acta 1985, 49, 1503−1513. (18) Agegian, C. R.; Mackenzie, F. T. Calcareous organisms and sediment mineralogy on a mid-depth bank in the Hawaiian archipelago. Pac. Sci. 1989, 43, 56−66. (19) Chave, K. E. Aspects of the biogeochemistry of magnesium 1. Calcareous marine organisms. J. Geol. 1954, 62 (3), 266−283. (20) Andersson, A. J.; Mackenzie, F. T.; Bates, N. R. Life on the margin: Implications of ocean acidification on Mg-calcite, high latitude and cold-water marine calcifiers. Mar. Ecol.: Prog. Ser. 2008, 373, 265− 273. (21) Andersson, A. J.; Mackenzie, F. T.; Ver, L. M. Solution of shallow-water carbonates: An insignificant buffer against rising atmospheric CO2. Geology 2003, 31 (6), 513−516. (22) Mathews, E. J.; Heap, A. D.; Woods, M. Inter-reefal seabed sediments and geomorphology of the Great Barrier Reef, a spatial analysis. Geoscience Australia Record 2007/09, 2007. (23) Daniell, J.; Hemer, M.; Heap, A.; Mathews, E.; Sbaffi, L.; Hughes, M.; Harris, P. Biophysical processes in the Torres Strait marine ecosystem II. Survey results and review of activities in response to CRC objectives. Geoscience Australia Report 2006/10, 2006. (24) Heap, A. D.; Przeslawski, R.; Radke, L.; Trafford, J.; Battershill, C. Shipboard party seabed environments of the Eastern Joseph Bonaparte Gulf, Northern Australia. SOL4934: Post-survey Report. Geoscience Australia Report 2010/09, 2010. (25) Mueller, G.; Gastner, M. The “Karbonat-Bombe”, a simple device for the determination of the carbonate content in sediments, soils, and other materials. Neues Jahrb. Mineral. 1971, 10, 466−469. (26) Goldsmith, J. R.; Graf, D. L.; Joensuu, O. I. The occurrence of magnesian calcites in nature. Geochim. Cosmochim. Acta 1955, 7 (5−6), 212−228. 8528
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
High-Magnesium Calcite Dissolution in Tropical Continental Shelf Sediments Controlled by Ocean Acidification R. R. Haese,*,† J. Smith,‡ R. Weber,‡ and J. Trafford‡ †
School of Earth Sciences, The University of Melbourne, Parkville, Victoria 3010, Australia Geoscience Australia, General Post Office Box 378, Canberra, Australian Capital Territory 2609, Australia
‡
S Supporting Information *
ABSTRACT: Increases in atmospheric CO2 cause the oceanic surface water to continuously acidify, which has multiple and profound impacts on coastal and continental shelf environments. Here we present the carbonate mineral composition in surface sediments from a range of continental shelf seabed environments and their current and predicted stability under ocean acidifying conditions. Samples come from the following four tropical Australian regions: (1) Capricorn Reef (southern end of the Great Barrier Reef), (2) the Great Barrier Reef Lagoon, (3) Torres Strait, and (4) the eastern Joseph Bonaparte Gulf. Beyond the nearshore zone, these regions typically have a carbonate content in surface sediments of 80 wt % or more. The abundance of high-magnesium calcites (HMC) dominates over aragonite (Arag) and low-magnesium calcite (LMC) and constitutes between 36% and 50% of all carbonate. HMC, with a magnesium content larger than 8−12 mol %, is more soluble than both Arag and LMC, and the solubility of HMC positively correlates with its magnesium concentration. From the solubility data of Plummer and Mackenzie (Am. J. Sci. 1974, 274, 61−83), 95% of HMC in the four regions is presently in metastable equilibrium relative to global mean tropical sea surface water. HMC is predicted to become destabilized in the four regions between 2040 and 2080 AD, with typical HMC decline rates between 2% and 5% per year. The range of respective estimated carbonate dissolution rates is expected to exceed current continental shelf carbonate accumulation rates, leading to net dissolution of carbonate during the period of HMC decline. In a geological context, the decline in HMC in tropical continental shelf environments is a global event triggered by reaching below-equilibrium conditions. The characteristic change in carbonate mineral composition in continental shelf sediments will serve as a geological marker for the proposed Anthropocene Epoch.
■
H 2O + CO2 + CO32 − → 2HCO3−
INTRODUCTION Atmospheric carbon dioxide (CO2) concentrations have increased by 40% from preindustrial levels, primarily from fossil fuel emissions.2 Rising atmospheric CO2 is tempered by oceanic uptake, which accounts for a third of anthropogenic carbon added to the atmosphere.3 Oceanic CO2 uptake, however, reduces pH and causes shift in seawater−carbonate chemistry in a process referred to as ocean acidification.4 Ocean acidification directly impacts a wide range of carbonate-producing marine organisms and has raised serious concerns about the potential effects on pelagic5,6 and benthic7,8 ecosystems. Studies on corals9 and pteropod molluscs10 from the tropical Australian continental shelf suggest ocean acidification is already detectable for recent decades. In shallow water areas of the inner continental shelf, the environmental impacts of ocean acidification may be exacerbated by other environmental and anthropogenic stressors such as episodic ocean warming leading to coral bleaching, overfishing, and land-based sources of pollution, leading to considerable economic and societal consequences.11 The reaction of CO2 with seawater reduces the availability of carbonate ions (eq 1), shifting the carbonate equilibrium and resulting in a lower saturation state with respect to calcium carbonate (CaCO3) minerals: © 2014 American Chemical Society
(1)
The saturation state (Ω) is commonly defined as the ratio of the concentration product to the stoichiometric solubility product (K′sp).12 A mineral is supersaturated when Ω > 1.0 and undersaturated when Ω < 1.0. For calculating the saturation state for calcite where calcium is partly replaced by magnesium, here referred to as high-magnesium calcite (HMC), the ratio of the ion activity product (IAP) to the solubility product based on observed apparent HMC solubilities (K′HMC) needs to be used and can be written as13 Ω HMC =
(a Mg 2+)x (aCa 2+)1 − x aCO32− K ′HMC
(2)
where x refers to the mole percent MgCO3. This conclusion has become a convention and follows a long controversy and discussion on the definition of a solubility product for a carbonate mineral with a given calcium-to-magnesium ratio Received: Revised: Accepted: Published: 8522
March 31, 2014 June 25, 2014 June 27, 2014 June 27, 2014 dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
Figure 1. Map of northeastern Australia including the continental shelf, showing sample locations of this study together with carbonate content of all samples registered in the Geoscience Australia MARS database. GBR, Great Barrier Reef; JBG, Joseph Bonaparte Gulf.
also deposit HMC of varying composition.12,18,19 In many coral reefs, HMC formed as the internal skeleton of coralline algae is part of the reef-stabilizing cement and thereby contributes to ecosystem sustainability beyond the life of its producing organism.20 Oceanic uptake of atmospheric CO2 primarily occurs in surface seawater extending over the continental shelf. Here, seawater CO2 concentration continuously increases and the saturation state with respect to carbonate minerals decreases (eq 2). HMC will be the first mineral phase subject to undersaturated conditions and is expected to dissolve because of its higher solubility than aragonite and calcite.13,14,16,20 Under the present rate of CO2 emissions, model calculations show that tropical surface seawater could become undersaturated with respect to HMC containing ≥12 mol % MgCO3 during this century.16 Dissolution of carbonate minerals consumes CO2 and produces alkalinity, and as a result, the pH and saturation state with respect to carbonate minerals increases. However, this potential buffer mechanism has been shown to be insignificant for two reasons. First, the relatively high rate of mixing with the deep open ocean rapidly dilutes alkalinityenriched continental shelf water, and second, the extent of carbonate dissolution is relatively small.16,21 As a result, a significant slowing or termination of ocean acidification through the oceanic buffer capacity is not expected. In this study we investigate the abundance and composition of carbonate minerals in surface sediments from four tropical regions of the Australian continental shelf. The observed carbonate mineralogy is compared to the current saturation state of carbonate minerals, and changes in mineral composition controlled by ocean acidification over the next 100 years are predicted.
exposed to water with variable ratios in the dissolved calcium to magnesium activities; see, for example, refs 12−14. It is further important to note that aragonite and HMC are metastable under current Earth surface conditions. However, as we report mineral stability in terms of the state of saturation (see above), we will use the terms under- and supersaturated for these minerals as well. Carbonate-rich sediments on the continental shelf represent a major CaCO3 reservoir that can rapidly react to the decreasing saturation state of seawater with respect to carbonate minerals. Three principal carbonate minerals are found in tropical continental shelf sediments: aragonite (Arag), low-magnesium calcite (LMC; 4 mol % MgCO3). Low- and highmagnesium calcites are often denoted simply as calcite and magnesium calcite, respectively. Currently, tropical surface seawaters are supersaturated with respect to LMC (−log K′ = 8.48) and Arag (−log K′ = 8.30).13,15 Differences in the solubility of HMC have been observed depending on sample preparation (see discussion in refs 12 and 13). Plummer and Mackenzie1 derived solubility data from experiments involving samples with minimal treatment, and on the basis of their results, HMC with 15 mol % MgCO3 is currently in equilibrium with tropical seawater (Ω = 1, −log K′ = 7.63).16 As HMC solubility correlates positively with magnesium concentration, HMC with a higher MgCO3 concentration is currently not thermodynamically favorable in tropical waters. The rate of carbonate dissolution can be described (e.g., ref 17) as
R = k(1 − Ω)n
(3)
where R is the dissolution rate, n is the reaction order, and k is the rate constant. The latter is strongly dependent on the reactive surface area, which decreases as dissolution progresses. Carbonate in marine sediments is predominantly of biogenic origin. The most common calcifiers forming HMC with significant mole percentages of MgCO3 are the red coralline algae, benthic foraminifera, bryozoans, and echinoderms, but other groups of organisms such as crustaceans, molluscs, annelid worms, calcareous sponges, barnacles, and brachiopods
■
MATERIALS AND METHODS This study utilizes previously collected sediment samples and data that have been archived at Geoscience Australia. A total of 216 surface sediment samples from the Australian continental shelf extending between 10° and 23° S were analyzed. Sample 8523
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
mmol Kg−1), Mg (8.892 mmol Kg−1), and CO32− (0.0099 mmol Kg−1) were calculated from average seawater concentrations,28 activity coefficients for Ca and CO32− at 25 °C,29 and an activity coefficient for Mg based on an estimated Mg/Ca activity ratio of 5.13 The solubility products for LMC and Arag are widely accepted with −log K′LMC = 8.48 and −log K′Arag = 8.3. In contrast, published K′sp values for HMC as a function of mole percent MgCO3 have been a matter of debate, and a good review of the discussion is provided by Morse et al.13 There are essentially two experimental solubility curves for biogenic magnesium calcites, and the curves are distinguished by the extent of preparation of the experimental materials. While it is not fully understood which solubility curve is most representative in natural environments, most studies12−14 suggest the use of HMC solubility data by Plummer and Mackenzie.1 For this reason, the Plummer and Mackenzie1 correlation (−log K′HMC = −0.063x + 8.586, where x = mol % MgCO3, for the range from 10−20 mol % MgCO3) was used in this study. Modeling the progression of ocean acidification requires a global biogeochemical model accounting for carbon fluxes across reservoir boundaries and changes in carbon, including carbonate ion reservoir sizes, over time. Here we use the model output derived by Gangstø et al.,30 which is based on the A2 scenario (IPCC SRES A2) for the period 2000−2100. The study provides a curve for the change in ΩArag in equatorial regions as defined by the region between 20°N and 20°S until 2100 AD. Present-day values of ΩHMCx, with x being the variable mole percent MgCO3, were calculated from the IAP and the HMC solubility data of Plummer and Mackenzie.1 The predicted ΩHMCx curves were then derived from the predicted global ΩArag curve for equatorial areas30 and the ratio of ΩHMCx to ΩArag.
locations and year of sampling are given in Supporting Information. The samples were collected from Torres Strait in 2004 (n = 77), the eastern Joseph Bonaparte Gulf (JBG) in 2009 (n = 63), the Great Barrier Reef (GBR) Lagoon between 2003 and 2005 (n = 45), and the Capricorn Reef in 2004 and 2005 (n = 31) (Figure 1). All samples were taken from surface sediments in 2000 km. Narrow inter-reef channels separating reef edifices are important geomorphic features, and northeastern trade winds promote an overall northward-trending water circulation.22 The Capricorn Reef is located on the Tropic of Capricorn at the southern end of the GBR at significant distance and geomorphologically separated from the coast by deep channels. Torres Strait is located at the northern end of the Cape York Peninsula and forms a shallow-water (typically 15−25 m deep) low-relief plain with extensive seagrass areas.23 Samples from the eastern JBG come from a north−south transect west of Melville Island where a range of seabed environments including outer-shelf deep-water channels, bank environments interspersed by channels, and inner-shelf plains and shallow channels are found.24 Dried and archived samples were analyzed by X-ray diffraction (XRD) at Geoscience Australia. Sample metadata are stored in the MARS database. The carbonate content of all samples was measured at Geoscience Australia by the “carbonate bomb” method.25 The estimated accuracy and precision is equal to or better than 5%. The mineral composition of powdered samples was analyzed on a Siemens D-500 X-ray diffractometer, from 2° to 70° 2θ, using a Cu anode X-ray tube at 40 kV and 20 mA. Minerals were identified by Bruker Diffracplus Eva and quantified by use of Siroquant V3 software. The mole percent MgCO3 in the HMC was calculated by the method proposed by Chave,19 with fluorite as an internal standard for all samples except for samples from Torres Strait. For the latter samples, aragonite was initially used as an internal standard. The average peak difference between aragonite and fluorite was later used to recalibrate against fluorite. The relationship of Goldsmith et al.26 was employed to relate the peak shift of the d104 peak to the Mg content in calcite of the specific sample. The d space values were derived from the Siroquant and Eva programs, and comparison of the two showed agreement equivalent to ±0.5 mol % MgCO3, which is an acceptable margin of error. A total of 19 surface seawater samples were taken along a north−south transect in the eastern JBG during the SOL4934 expedition onboard the RV Solander. Temperature and salinity at sampling sites were recorded on a thermosalinograph. Total alkalinity (TA) was determined by Gran titration and by the Solver method,27 and pH was determined by the Tris buffer method and reported on the total hydrogen ion scale.28 Temperature, salinity, TA, and pH were used to calculate the state of saturation for aragonite by use of the CO2SYS program28 and the state of saturation of HMC (see below). The state of saturation, Ω, for Arag, LMC, and HMC was calculated according to eq 2. Ionic activities for Ca (1.778
■
RESULTS AND DISCUSSION Carbonate Content and Carbonate Mineral Composition. Carbonate content in continental shelf sediments of northern Australia varies significantly (Supporting Information) and is largely controlled by the discharge of siliciclastic detrital sediment by rivers and the geomorphology of the seafloor. The GBR Lagoon is part of the inner shelf and is characterized by high concentrations of terrestrial-derived muds and a carbonate content of 80%.22 Accordingly, the carbonate content range of the GBR Lagoon sample set is large, varying between 24 and 94 wt % with a mean of 80 wt % carbonate. In comparison, the carbonate content in Capricorn Reef samples ranges from 43 to 96 wt % with a mean of 87 wt %. Similarly, the carbonate content in Torres Strait samples varies between 72 and 95 wt % with a mean of 88 wt %. Samples from the eastern JBG show the highest proportion of detrital sediment, with a carbonate content in the range between 26 and 92 wt % and a mean concentration of 61 wt %. All samples contained typical siliciclastic minerals including kaolin, quartz, and muscovite (data not shown) as well as HMC, LMC, and Arag. The individual carbonate minerals were normalized to a percentage of total carbonate and are presented in Figure 2. Torres Strait and Capricorn Reef samples show higher mean proportions of HMC and a lower spread in the carbonate mineral distribution as compared to the GBR Lagoon and the eastern JBG samples. Torres Strait and Capricorn Reef have a mean HMC proportion of 45% (standard deviation σ = 6) and 50% (σ = 6), respectively, whereas the GBR Lagoon and 8524
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
Figure 3. Histograms showing the distribution of MgCO3 content in high-magnesium calcites in the four study areas.
Figure 2. Ternary diagrams showing proportions of high-magnesium calcite, low-magnesium calcite, and aragonite for the four study areas.
Lagoon, all study areas show a maximum magnesium content in HMC of 15.25 mol % MgCO3. This observation will be further discussed in the context of current and future HMC solubility. Solubility of High-Magnesium Calcite. On the basis of the solubility estimate for biogenic HMC with little treatment by Plummer and Mackenzie,1 surface seawater at 25 °C is currently in equilibrium with HMC containing 15.2 mol % MgCO3. This calculated equilibrium value agrees very well with an observed maximum MgCO3 content of 15.2 mol % in HMC in 95% of our samples. This suggests HMC with a higher MgCO3 content is currently undersaturated and therefore absent in most situations. However, these model calculations predict global changes in carbonate mineral stability and do not account for local or regional differences. Regional water chemistry data from areas of interest for this study show a high degree of agreement with the described global stability of HMC in tropical regions: Annual mean sea surface temperature varies between the Tropic of Capricorn and Torres Strait between 25 and 28 °C,35 and this temperature variability leads to a state of saturation for 15 mol % HMC, ΩHMC‑15, between 0.96 and 1.07 for 2200 μmol·kg−1 total alkalinity and pH 8.1 in seawater. The influence of upwelling events exposing the continental shelf temporarily to colder and more acidic water is considered to be minor or even negligible, given the shallow sample depths. For example, warm surface water outside the northern GBR is separated from deep water through a high-salinity layer typically located at a depth of 150 m. Only during the peak period of an upwelling event does the high-salinity layer extend over the shelf break.36 Surface water samples collected within the eastern JBG showed a north− south temperature gradient between 29.2 and 27.2 °C and median, minimum, and maximum ΩHMC‑15 values of 1.19, 0.94, and 1.27, respectively. Similarly, surface water with a temperature of 28 °C at Lizard Island (14°40′08″ S, 145°27′34″ E) in the northern part of the GBR has an average ΩArag of 4.2637 and a respective ΩHMC‑15 of 1.19. Reef flat waters, however, show large diurnal variability in carbonate water composition and respective carbonate mineral saturation in response to net
the eastern JBG have mean HMC values of 35% (σ = 10) and 36% (σ = 9), respectively. Multiple factors may contribute to the differences in carbonate mineral composition in surface sediments. Here, it is likely due to the larger size of the sample area and the greater diversity in seabed morphology found in the GBR Lagoon and the eastern JBG, which lead to a larger range of carbonate mineral composition as compared to Torres Strait and the Capricorn Reef. Importantly, the calculated HMC:LMC:Arag ratio for the four tropical continental shelf regions is 42:38:20, highlighting the dominance of HMC in these tropical continental shelf regions. The dominance of HMC was also observed in sediments from the Bahamas,31 but the derived ratio differs significantly from a global neritic ratio of 24:13:6313 based on data compiled earlier.32 Carbonate mineral solubility is temperature-dependent and solubility increases with decreasing temperatures. In tropical regions, higher water temperatures result in a higher degree of carbonate mineral stability and a larger proportion of HMC. Conversely, HMC is absent in polar regions due to undersaturation at cooler water temperatures. The higher degree of carbonate mineral stability in tropical waters also reflects a higher biogenic calcification rate since a linear correlation between the degree of (aragonite) saturation and the calcification rate has been found for benthic15,33,34 and planktonic5,6 calcifiers. As the solubility of HMC depends on its magnesium content, the mole percent MgCO3 in HMC was determined for each sample. Histograms showing the relative frequency of the magnesium content in increments of 0.25 mol % MgCO3 for each of the four study areas are shown in Figure 3. Samples from Torres Strait and eastern JBG show a unimodal distribution, whereas the sample distribution of Capricorn Reef and the GBR Lagoon is bimodal. HMC in Torres Strait has a mean magnesium content of 14.5 mol % MgCO3 (σ = 0.4), whereas the magnesium content in HMC in the other three areas ranges between 13.6 and 13.8 mol % MgCO3 (σ = 0.7−1.0). Except for a small subset of samples from the GBR 8525
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
1, before the year 2100 AD, which is in agreement with earlier estimates.16 Our results further show the solubility of HMC drops below equilibrium in increments of 1 mol % MgCO3 every 20−30 years. If it is assumed that the decrease in HMC is purely a function of mineral stability, the rate of HMC decrease as a proportion of the carbonate content and the rate of HMC dissolution can be estimated by combining the predicted changes in saturation state over time and the derived total carbonate content and carbonate mineral concentrations. Figure 5 shows the
community calcification. At Lady Elliot Island (24°6′50″ S, 152°42′40″ E) in the southern part of the GBR, seasonal median ΩArag varies between 3.0 and 3.5, but a total range in ΩArag between 1.13 and 6.46 was observed.38 The latter converts to a range in ΩHMC‑15 between 0.32 and 1.80, which gives evidence for short-term metastability of HMC. A significant fraction of carbonate deposited at the sediment surface is dissolved through the mineralization of organic matter. Organic matter degradation by aerobic respiration leads to proton production and has been shown to be quantitatively most important for carbonate dissolution in shallow continental shelf39 and deeper marine sediments.40 Apart from the organic matter deposition rate, it is primarily the oxygen flux to the zone of aerobic respiration controlling proton production and associated carbonate dissolution. These primary process controls on carbonate dissolution in sediments are linked to ocean acidification, as it is likely that the rate constant contributing to the HMC dissolution rate (eq 3) in sediments decreases as HMC becomes increasingly exposed to undersaturated bottom water conditions before reaching the zone of aerobic respiration. When we compare the solubility of biogenic HMC-15 with minimal treatment to aragonite, we find a K′HMC‑15:K′Arag ratio of 4.4. In comparison, a K′HMC‑15:K′Arag ratio of 1.2 is determined when biogenic HMC-15 solubility derived from cleaned and annealed samples is compared to aragonite as suggested earlier.13 Furthermore, the solubility of cleaned and annealed samples predicts HMC with a much higher MgCO3 content to be stable under current tropical surface seawater conditions, which is not observed. This difference gives further support for the Plummer and Mackenzie1 solubility data under natural conditions. Predicted Stability of High-Magnesium Calcite over the Next 100 Years. Saturation curves for HMC with variable mole percent MgCO3, Arag, and LMC in equatorial surface water until the end of this century are shown in Figure 4. While LMC and Arag remain highly supersaturated, HMC with 12− 15 mol % magnesium carbonate drops below equilibrium, Ω ≤
Figure 5. Predicted decrease in high-magnesium calcite concentration in surface sediments in the four study areas. GBR, Great Barrier Reef; JBG, Joseph Bonaparte Gulf.
calculated relative decrease in HMC in increments of 0.25 mol % MgCO3. It shows almost all of the HMC decrease will occur between 2040 and 2080 AD in all four regions. Torres Strait is a precursor relative to the other regions due to the high mean MgCO3 content of 14.5 mol % in HMC. Based on the linear regression through the steepest concentration gradient over 20−30 years, the decrease in HMC for Torres Strait (2040−2060 AD) is 5% per year, for the GBR Lagoon (2040− 2070 AD) and the eastern JBG (2050−2080 AD), 3% per year, and for the Capricorn Reef (2035−2075 AD) 2% per year. Given the ubiquitous distribution of HMC in tropical and subtropical continental shelf sediments and the predicated measurable changes at time intervals on the order of 5 years post-2040 AD, the decline in maximum mole percent MgCO3 in HMC and the overall decline in HMC content in surface sediments may serve as indicators of ocean acidification. The above derived changes in the mineral composition of marine sediments are very rapid and include the assumption mineral dissolution kinetics do not limit the reaction rate. The latter depends on mineral-specific kinetic properties including the reactive surface area (see eq 3), which are not taken into account. The reactive surface area decreases during dissolution and may at some point limit the dissolution rate. With the above assumptions, carbonate dissolution rates controlled by ocean acidification can be calculated from the derived results if further assumptions are made about sediment porosity and the penetration depth of dissolution. Both parameters are highly variable in continental shelf sediments, depending on factors such as the energy regime of the depositional environment and carbon mineralization and bioturbation rates.40 For the purpose
Figure 4. Saturation state of calcite, aragonite, and high-magnesium calcites with variable concentrations of MgCO3. The horizontal black line at Ω = 1 represents fluid−mineral equilibrium. 8526
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
The presented results suggest the complete decline of HMC in surface sediments of the tropical and subtropical continental shelf between 2040 and 2080 AD based on mineral stability considerations only. Predicting dissolution rates over upcoming decades via carbonate dissolution kinetics is currently uncertain, mostly due to an unknown decline in the rate constant. However, reaction kinetics appear not to have limited HMC dissolution up to the present, since HMC in current surface sediments is either in equilibrium or supersaturated. In the broader context of impacts driven by the anthropogenic increase in atmospheric CO2, the rapid decline in HMC serves as an example of an event rather than a trend triggered by exceeding a threshold, in this case by reaching the state of mineral undersaturation. Because the predicted change is characteristic for the global tropical continental shelf and occurs as an event on the geological time scale, the decline in HMC may serve as one geological marker for the onset of the proposed Anthropocene Epoch.43
of estimating the order of magnitude of enhanced carbonate dissolution controlled by ocean acidification based on mineral stability only, a minimum and a maximum carbonate dissolution rate is calculated as follows: The minimum rate uses dissolution penetration depth of 2 cm, porosity of 70%, average mineral density of 2.65 g·cm−3, HMC decline of 2% per year, carbonate content of 80%, and HMC proportion of 40% of all carbonate. The maximum rate uses dissolution penetration depth of 5 cm, porosity of 40%, HMC decline rate of 5% per year, and the remaining parameters as above. The resulting range in enhanced carbonate dissolution rate is between 100 and 1300 g·m−2·year−1, which is in the range of carbonate dissolution rates observed so far.41 The derived HMC dissolution rates are also very similar to current carbonate accumulation rates, estimated as 1200, 250, and 30 g·m−2·year−1 in coral reef complexes, carbonate banks, and carbonate shelf environments, respectively.42 Given that the estimated carbonate dissolution range is equivalent to the given accumulation range, no net carbonate accumulation on the continental shelf during the period of HMC decline may be possible. Alternatively, the HMC dissolution rate can be calculated by use of eq 3; however, two uncertainties are introduced. First, estimates of the initial rate constant for bulk samples are uncertain, as the grain size and composition of biogenic carbonate fragments are highly variable between samples. Second, the decrease in rate constant over decades is critical and equally uncertain. This is illustrated by the following example: eq 3 is applied for HMC-13.5 with incremental time steps of 1 year. Sediment properties are chosen as above for the minimum carbonate dissolution rate, and ΩHMC‑13.5 is allowed to change over time as derived in this study. A reaction order of 3.5 is used.17 The rate constant is assumed to decline by an order of magnitude each 10 years, t10(k) = 10, and each 20 years, t10(k) = 20, starting with the first year of undersaturation. In the case of t10(k) = 10, low rate constants and low dissolution rates, respectively, are reached much earlier, and consequently HMC dissolution is much lower. In contrast, in the case of t10(k) = 20, the dissolution period is drawn out and much more extensive. Indeed, complete dissolution of HMC before 2110 AD is suggested (Figure 6).
■
ASSOCIATED CONTENT
S Supporting Information *
One table listing sample locations, depths, figure showing median, lower and upper ranges of carbonate content from four material is available free of charge via the pubs.acs.org.
■
and dates and one quartile, and total sample sets. This Internet at http://
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank crews for shipboard sampling and personnel at Geoscience Australia for archiving samples and maintaining the MARS sample database. Liz Webber and Murray Woods from Geoscience Australia provided XRD analysis and preparation of maps, respectively. We thank Scott Nichol and Lynda Radke from Geoscience Australia and three anonymous reviewers for their comments on an earlier manuscript version. This publication received permission for publication by the CEO of Geoscience Australia.
■
REFERENCES
(1) Plummer, L. N.; Mackenzie, F. T. Predicting mineral solubility from rate data: Application to the dissolution of magnesian calcites. Am. J. Sci. 1974, 274, 61−83. (2) IPCC. Climate Change 2013: The Physical Science Basis; Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change 2013; Cambridge University Press: Cambridge, U.K., and New York, 2013. (3) Sabine, C. L.; et al. The oceanic sink for anthropogenic CO2. Science 2004, 305 (5682), 367−371. (4) Doney, S. C.; Fabry, V. J.; Feely, R. A.; Kleypas, J. A. Ocean acidification: The other CO2 problem. Annu. Rev. Mar. Sci. 2009, 1 (1), 169−192. (5) Riebesell, U.; Zondervan, I.; Rost, B.; Tortell, P.; Zeebe, R.; Morell, F. M. M. Reduced calcification of marine plankton in response to increased atmospheric CO2. Nature 2000, 407, 364−367. (6) Moy, A. D.; Howard, W. R.; Bray, S. G.; Trull, T. W. Reduced calcification in modern Southern Ocean planktonic foraminifera. Nat. Geosci. 2009, 2, 276−280.
Figure 6. Predicted changes in HMC concentration for two rate constants, k. The rate constants t10(k) = 10 and t10(k) = 20 decrease by an order of magnitude each 10 and 20 years, respectively. See text for other values used in calculation. 8527
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
Environmental Science & Technology
Article
(27) Dickson, A. G.; Sabine C. L.; Christian, J. R., Eds. Guide to best practices for ocean CO2 measurements; PICES Special Publication 3; North Pacific Marine Science Organization: Sidney, BC, Canada, 2007. (28) Pierrot, D.; Lewis, E.; Wallace, D. W. R. MS Excel program developed for CO2 systems calculations; ORNL/CDIAC-105a; Carbon Dioxide Information Analysis Centre, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN, 2006; DOI: 10.3334/ CDIAC/otg.CO2SYS_XLS_CDIAC105a. (29) Plummer, L. N.; Sundquist, E. T. The individual ion activity coefficients of calcium and carbonate in sea water at 25 °C and 35 ‰ salinity, and implications to the agreement between apparent and thermodynamic constants for calcite and aragonite. Geochim. Cosmochim. Acta 1982, 46, 247−258. (30) Gangstø, R.; Gehlen, M.; Schneider, B.; Bopp, L.; Aumont, O.; Joos, F. Modeling the marine aragonite cycle: changes under rising carbon dioxide and its role in shallow water CaCO3 dissolution. Biogeosciences 2008, 5, 1057−1072. (31) Morse, J. W.; Zullig, J. J.; Bernstein, L. D.; Millero, F. J.; Milne, F.; Mucci, A.; Choppin, G. R. Chemistry of calcium carbonate-rich shallow-water sediments in the Bahamas. Am. J. Sci. 1985, 285, 147− 183. (32) Land, L.S. Diagenesis of skeletal carbonates. J. Sed. Res. 1967, 37 (3), 914−930. (33) Leclercq, N.; Gattuso, J.-P.; Jaubert, J. CO2 partial pressure controls the calcification rate of a coral community. Global Change Biol. 2000, 6, 329−334. (34) Leclercq, N.; Gattuso, J.-P. Primary production, respiration, and calcification of a coral reef mesocosm under increased CO2 partial pressure. Limnol. Oceanogr. 2002, 47, 558−564. (35) Kuchinke, M.; Tilbrook, B.; Lenton, A. Seasonal variability of aragonite saturation state in the Western Pacific. Mar. Chem. 2014, 161, 1−13. (36) Andrews, J. C.; Gentien, P. Upwelling as a source of nutrients for the Great Barrier Reef ecosystem: A solution to Darwin’s question? Mar. Ecol.: Prog. Ser. 1982, 8, 257−269. (37) Diaz-Pulido, G.; Nash, M. C.; Anthony, K. R. N.; Bender, D.; Opdyke, B. N.; Reyes-Nivia, C.; Troitsch, U. Greenhouse conditions induce mineralogical changes and dolomite accumulation in coralline algae on tropical reefs. Nat. Commun. 2014, 5, No. 3310, DOI: 10.1038/ncomms4310. (38) Shaw, E. C.; McNeil, B. I.; Tilbrook, B. Impacts of ocean acidification in naturally variable coral reef flat ecosystems. J. Geophys. Res. 2012, 117, No. C03038, DOI: 10.1029/2011JC007655. (39) Burdige, D. J.; Zimmerman, R. C.; Hu, X. Rates of carbonate dissolution in permeable sediments estimated from pore-water profiles: the role of sea grass. Limnol. Oceanogr. 2008, 53, 549−565. (40) Pfeiffer, K.; Henson, C.; Adler, M.; Wenzhöfer, F.; Weber, B.; Schulz, H. D. Modeling of subsurface calcite dissolution, including the respiration and reoxidation process of marine sediments in the regional of the equatorial upwelling off Gabon. Geochim. Cosmochim. Acta 2002, 66, 4247−4259. (41) Andersson, A. J.; Gledhill, D. Ocean acidification and coral reefs: Effects of breakdown, dissolution, and net ecosystem calcification. Annu. Rev. Mar. Sci. 2013, 5, 321−348. (42) Milliman, J. D.; Droxler, A. W. Neritic and pelagic carbonate sedimentation in the marine environment: Ignorance is not bliss. Int. J. Earth Sci. 1996, 85, 496−504. (43) Zalasiewicz, I.; Williams, M.; Steffen, W.; Crutzen, P. The new world of the Anthropocene. Environ. Sci. Technol. 2010, 44 (7), 2228− 2231.
(7) Hoegh-Guldberg, O.; et al. Coral reefs under rapid climate change and ocean acidification. Science 2007, 318, 1737−1742. (8) Anthony, K. R. N.; Kline, D. I.; Diaz-Pulido, G.; Dove, K. S.; Hoegh-Guldberg, O. Ocean acidification causes bleaching and productivity loss in coral reef builiders. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17442−17446. (9) Wei, G.; McCulloch, M. T.; Mortimer, G.; Deng, W.; Xie, L. Evidence for ocean acidification in the Great Barrier Reef of Australia. Geochim. Cosmochim. Acta 2009, 73, 2332−2346. (10) Roger, L. M.; Richardson, A. J.; McKinnon, A. D.; Knott, B.; Matear, R.; Scadding, C. Comparison of the shell structure of two tropical Thecosomata (Creseis acicula and Diacavolinia longirostris) from 1963 to 2009: Potential implications of declining aragonite saturation. ICES J. Mar. Sci. 2012, 69, 465−474. (11) Hood, M.; Broadgate, W.; Urban, E.; Gaffey, O. Ocean Acidification: A Summary for Policymakers; Second Symposium on the Ocean in a High-CO2 World, Scientific Committee on Oceanic Research, Monaco, 2009. (12) Geochemistry of Sedimentary Carbonates. In Developments in Sedimentology; Morse, J., Mackenzie, F. T., Eds.; Elsevier: Amsterdam, 1990; Vol. 48. (13) Morse, J. W.; Andersson, A. J.; Mackenzie, F.T. Initial responses of carbonate-rich shelf sediments to rising atmospheric pCO2 and “ocean acidification”: Role of high Mg-calcites. Geochim. Cosmochim. Acta 2006, 70 (23), 5814−5830. (14) Mackenzie, F. T.; et al. Magnesian calcites: Low-temperature occurrence, solubility and solid-solution behaviour. In Carbonates: Mineralogy and Chemistry; Reviews in Mineralogy and Geochemistry, Vol. 11; Reeder, R. J., Ed.; Mineralogical Society of America: Chantilly, VA, 1983; pp 97−143. (15) Kleypas, J. A.; et al. Geochemical consequences of increased atmospheric carbon dioxide on coral reefs. Science 1999, 284 (5411), 118−120. (16) Andersson, A. J.; Mackenzie, F. T. Revisiting four scientific debates in ocean acidification research. Biogeosciences 2012, 9 (3), 893−905. (17) Walter, L. M.; Morse, J. W. The dissolution kinetics of shallow marine carbonates in sea water: A laboratory study. Geochim. Cosmochim. Acta 1985, 49, 1503−1513. (18) Agegian, C. R.; Mackenzie, F. T. Calcareous organisms and sediment mineralogy on a mid-depth bank in the Hawaiian archipelago. Pac. Sci. 1989, 43, 56−66. (19) Chave, K. E. Aspects of the biogeochemistry of magnesium 1. Calcareous marine organisms. J. Geol. 1954, 62 (3), 266−283. (20) Andersson, A. J.; Mackenzie, F. T.; Bates, N. R. Life on the margin: Implications of ocean acidification on Mg-calcite, high latitude and cold-water marine calcifiers. Mar. Ecol.: Prog. Ser. 2008, 373, 265− 273. (21) Andersson, A. J.; Mackenzie, F. T.; Ver, L. M. Solution of shallow-water carbonates: An insignificant buffer against rising atmospheric CO2. Geology 2003, 31 (6), 513−516. (22) Mathews, E. J.; Heap, A. D.; Woods, M. Inter-reefal seabed sediments and geomorphology of the Great Barrier Reef, a spatial analysis. Geoscience Australia Record 2007/09, 2007. (23) Daniell, J.; Hemer, M.; Heap, A.; Mathews, E.; Sbaffi, L.; Hughes, M.; Harris, P. Biophysical processes in the Torres Strait marine ecosystem II. Survey results and review of activities in response to CRC objectives. Geoscience Australia Report 2006/10, 2006. (24) Heap, A. D.; Przeslawski, R.; Radke, L.; Trafford, J.; Battershill, C. Shipboard party seabed environments of the Eastern Joseph Bonaparte Gulf, Northern Australia. SOL4934: Post-survey Report. Geoscience Australia Report 2010/09, 2010. (25) Mueller, G.; Gastner, M. The “Karbonat-Bombe”, a simple device for the determination of the carbonate content in sediments, soils, and other materials. Neues Jahrb. Mineral. 1971, 10, 466−469. (26) Goldsmith, J. R.; Graf, D. L.; Joensuu, O. I. The occurrence of magnesian calcites in nature. Geochim. Cosmochim. Acta 1955, 7 (5−6), 212−228. 8528
dx.doi.org/10.1021/es501564q | Environ. Sci. Technol. 2014, 48, 8522−8528
