Alkalinity changes in the Sargasso Sea: geochemical evidence of calcification?

Strong seasonal patterns in upper ocean total carbon dioxide (TCO₂), alkalinity (TA) and calculated pCO₂ were observed in a time series of water column measurements collected at the US Joint Global Ocean Flux Study (JGOFS) BATS site (31 °50′N, 64 °10′W) in the Sargasso Sea. TA distribution was a conservative function of salinity. However, in February 1992, a non-conservative decrease in TA was observed, with maximum depletion of 25–30 μmoles kg⁻¹ occuring in the surface layer and at the depth of the chlorophyll maximum (˜ 80–100 m). Mixed-layer TCO₂ also decreased, while surface pCO₂ increased by 25–30 μatm. We suggest these changes in carbon dioxide species resulted from open-ocean calcification by carbonate-secreting organisms rather than physical processes. Coccolithophore calcification is the most likely cause of this event although calcification by foraminifera or pteropods cannot be ruled out. Due to the transient increase in surface pCO₂, the net annual transfer of CO₂ into the ocean at BATS was reduced. These observations demonstrate the potential importance of open-ocean calcification and biological community structure in the biogeochemical cycling of carbon.     

Short-term variability of fCO2 in seawater and air–sea CO2 fluxes in a coastal upwelling system (Ría de Vigo, NW Spain)

Coastal upwelling systems are regions with highly variable physical processes and very high rates of primary production and very little is known about the effect of these factors on the short-term variations of CO₂ fugacity in seawater (fCO₂^w). This paper presents the effect of short-term variability (<1 week) of upwelling–downwelling events on CO₂ fugacity in seawater (fCO₂^w), oxygen, temperature and salinity fields in the Ría de Vigo (a coastal upwelling ecosystem). The magnitude of fCO₂^w values is physically and biologically modulated and ranges from 285 μatm in July to 615 μatm in October. There is a sharp gradient in fCO₂^w between the inner and the outer zone of the Ría during almost all the sampling dates, with a landward increase in fCO₂^w.CO₂ fluxes calculated from local wind speed and air–sea fCO₂ differences indicate that the inner zone is a sink for atmospheric CO₂ in December only (−0.30 mmol m⁻² day⁻¹). The middle zone absorbs CO₂ in December and July (−0.05 and −0.27 mmol·m⁻² day⁻¹, respectively). The oceanic zone only emits CO₂ in October (0.36 mmol·m⁻² day⁻¹) and absorbs at the highest rate in December (−1.53 mmol·m⁻² day⁻¹).     

fCO2 variability at the CARIACO tropical coastal upwelling time series station

Monthly seawater pH and alkalinity measurements were collected between January 1996 and December 2000 at 10°30′N, 64°40′W as part of the CARIACO (CArbon Retention In A Colored Ocean) oceanographic time series. One key objective of CARIACO is to study temporal variability in Total CO₂ (TCO₂) concentrations and CO₂ fugacity (fCO₂) at this tropical coastal wind-driven upwelling site. Between 1996 and 2000, the difference between atmospheric and surface ocean CO₂ concentrations ranged from about − 64.3 to + 62.3 μatm. Physical and biochemical factors, specifically upwelling, temperature, primary production, and TCO₂ concentrations interacted to control temporal variations in fCO₂. Air–sea CO₂ fluxes were typically depressed (0 to + 10 mmol C m^{− 2} day^{− 1}) in the first few months of the year during upwelling. Fluxes were higher during June–November (+ 10 to 20 mmol C m^{− 2} day^{− 1}). Fluxes were generally independent of the slight changes in salinity normally seen at the station, but low positive flux values were seen in the second half of 1999 during a period of anomalously heavy rains and land-derived runoff. During the 5 years of monthly data examined, only two episodes of negative air–sea CO₂ flux were observed. These occurred during short but intense upwelling events in March 1997 (−10 mmol C m^{− 2} day^{− 1}) and March 1998 (− 50 mmol C m^{− 2} day^{− 1}). Therefore, the Cariaco Basin generally acted as a source of CO₂ to the atmosphere in spite of primary productivity in excess of between 300 and 600 g C m^{− 2} year^{− 1}.     

Temperature dependence of CO2 fugacity in seawater

Ideally, the correction of the measured CO₂ fugacity (fCO₂) at temperature T_m to fCO₂ at the in-situ temperature T_in should be made by using at least 2 known parameters (pH-A_T, C_T-A_T,…) and the reliable constants for carbonic acid. In practice however, a measured CO₂ property pair is not always available. When fCO₂ is measured alone, one must make an estimate of the effect of temperature on seawater fCO₂ from the accurate knowledge of seawater salinity and temperature and the approximate knowledge of the carbonate parameters. In this paper we present an empirical relationship that can be used to estimate the effect of temperature on fCO₂. The equation is of the form:

ƒCO_2[t] − ƒCO_2[20]=A + Bt + Ct² + Dt³ + Et⁴

where fCO_2[t] and fCO_2[20] represent fCO₂ at temperatures t°C and 20°C, respectively; the parameters A, B, etc. are functions of the ratio X = C_T/A_T:

E = e₀ + e₁X + e₂X²ln(X) + e₃exp(X) + e₄/ln(X)

where the parameters a_i, b_i, etc. are functions of salinity.The 25-parameter equation is fitted by the values of fCO₂ calculated using the constants of Goyet and Poisson (1989), when X varies from 0.8 to 1.0, t varies from −1dgC to 40°C, and S varies from 30 to 40. For T_m - T_in within ± 10°C, direct measurements of fCO₂ as a function of the temperature (from −I to 30°C verify this equation within less than ±5 μatm.     

Seasonal and interannual variability of the air–sea CO2 flux in the Atlantic sector of the Barents Sea

The seasonal and interannual variability of the air–sea CO₂ flux (F) in the Atlantic sector of the Barents Sea have been investigated. Data for seawater fugacity of CO₂ (fCO₂^sw) acquired during five cruises in the region were used to identify and validate an empirical procedure to compute fCO₂^sw from phosphate (PO₄), seawater temperature (T), and salinity (S). This procedure was then applied to time series data of T, S, and PO₄ collected in the Barents Sea Opening during the period 1990–1999, and the resulting fCO₂^sw estimates were combined with data for the atmospheric mole fraction of CO₂, sea level pressure, and wind speed to evaluate F.The results show that the Atlantic sector of the Barents Sea is an annual sink of atmospheric CO₂. The monthly mean uptake increases nearly monotonically from 0.101 mol C m^− 2 in midwinter to 0.656 mol C m^− 2 in midfall before it gradually decreases to the winter value. Interannual variability in the monthly mean flux was evaluated for the winter, summer, and fall seasons and was found to be ± 0.071 mol C m^− 2 month^− 1. The variability is controlled mainly through combined variation of fCO₂^sw and wind speed. The annual mean uptake of atmospheric CO₂ in the region was estimated to 4.27 ± 0.68 mol C m^− 2.     

Surface water carbon dioxide in the southwest Indian sector of the Southern Ocean: a highly variable CO2 source/sink region in summer

Measurements of partial pressure of carbon dioxide (pCO₂), total dissolved inorganic carbon (TCO₂), total alkalinity (TA) and chlorophyll a (Chl a) have been made in surface water in the southwestern Indian sector of the Southern Ocean (20–85°E) in the austral summer (INDIVAT V cruise, January-February 1987). Between Antarctica and Africa, pCO₂ distribution was linked to the oceanic frontal zones and Chi a variations. The pCO₂ spatial structure was very close to that explored in summer 1967 in the same region but the pCO₂ differences between the ocean and the atmosphere were smaller in 1987 than 20 years ago. At all latitudes we found strongly contrasting surface pCO₂ characteristics between eastern (around 80°E) and western (around 25°E) regions; C02 sources were mainly in the west and CO₂ sinks in the east. South of 60°S, the contrast could be due to biological activity. Between 60°S and the Antarctic Polar Front, intensification of upwelling might be responsible for the higher _pCO₂ values in the west.     

The dissociation constants of carbonic acid in seawater at salinities 5 to 45 and temperatures 0 to 45°C

The pK₁^* and pK₂^* for the dissociation of carbonic acid in seawater have been determined from 0 to 45°C and S = 5 to 45. The values of pK₁^* have been determined from emf measurements for the cell:

Pt](1 − X)H2 + XCO2|NaHCO3, CO2 in synthetic seawater|AgC1; Ag

where X is the mole fraction of CO₂ in the gas. The values of pK₂^* have been determined from emf measurements on the cell:

Pt, H2(g, 1 atm)|Na2CO3, NaHCO3 in synthethic seawater|AgC1; Ag

The results have been fitted to the equations:

lnK^*₁ = 2.83655 − 2307.1266/T − 1.5529413 lnT + (−0.20760841 − 4.0484/T)S^0.5 + 0.08468345S − 0.00654208S¹

InK^*₂ = −9.226508 − 3351.6106/T− 0.2005743 lnT + (−0.106901773 − 23.9722/T)S^0.5 + 0.1130822S − 0.00846934S^1.5

where T is the temperature in K, S is the salinity, and the standard deviations of the fits are σ = 0.0048 in lnK₁^* and σ = 0.0070 in lnK₂^*.Our new results are in good agreement at S = 35 (±0.002 in pK₁^*and ±0.005 in pK₂^*) from 0 to 45°C with the earlier results of Goyet and Poisson (1989). Since our measurements are more precise than the earlier measurements due to the use of the Pt, H₂|AgCl, Ag electrode system, we feel that our equations should be used to calculate the components of the carbonate system in seawater.     

o,p′-DDT induction of vitellogenesis and its inhibition by tamoxifen in Nile tilapia (Oreochromis niloticus)

In order to investigate the mechanism by which o,p′-DDT disrupts endocrine functioning of Nile tilapia in vivo, the estrogenicity of o,p′-DDT was investigated in conjunction with 17β-estradiol (E₂) and tamoxifen. Mature, male tilapia were treated intraperitoneally with o,p′-DDT (60 mg/kg, one dose) or E₂ (5 mg/kg, four doses) in the presence or absence of tamoxifen (5 mg/kg, six doses) for 12 days and then plasma vitellogenin (Vtg) (measured as alkaline-labile phosphorous), E₂, and testosterone (T) were measured. Vtg levels were increased dramatically by E₂ (1744±171 μg/ml) and moderately by o,p′-DDT (82±15 μg/ml) compared with controls (23±3.5 μg/ml). Tamoxifen alone had no effect on Vtg production, but inhibited both E₂ and o,p′-DDT stimulated vitellogenesis. T levels were reduced with E₂ administration (1688±383 pg/ml) and declined further with the combined treatment of E₂ and tamoxifen (281±70 pg/ml), compared with controls (6558±1438 pg/ml). Tamoxifen or o,p′-DDT alone did not affect T levels, but their combined treatment did (2069±647 pg/ml). The results of this study suggest that o,p′-DDT is weakly estrogenic in male tilapia, and that this activity may be mediated through the estrogen receptor.     

Increase of carbon dioxide in the bottom water of the Weddell Sea, Antarctica

High precision total CO₂ (TCO₂) data are presented from the NW Weddell Sea obtained during two cruises which were 3 years apart. A TCO₂ increase from 1993 to 1996 was observed in the newly formed bottom water, whereas no TCO₂ increase was found in the surrounding water masses. Accompanying this TCO₂ increase in the bottom water was an oxygen decrease. Obviously, bottom water with variable characteristics is produced along the margins of the Weddell Sea. Examination of possible causes leads to the conclusion that the bottom water variability is largely due to varying amounts of Warm Deep Water intruding onto the shelves of the Weddell Sea, thus changing the shelf water end-member of bottom water formation. Analysis of the data, using the observed differences of oxygen to perform a correction, suggested that some part of the TCO₂ increase of the bottom water is due to the increased level of anthropogenic CO₂. The TCO₂ increase of the bottom water is commensurate to a tentative annual increase of about 1 μmol kg⁻¹ in the surface water source of this bottom water. This would agree fairly well with the increase of the partial pressure of CO₂ in the atmosphere.     

Modelling the monthly sea surface fCO2 fields in the Indian Ocean

In order to construct monthly fields of sea surface fugacity of carbon dioxide (f_CO2) on a large scale in the Indian Ocean, we use a one-dimensional model which takes into account the main physical and biogeochemical processes controlling f_CO2 variations in the ocean. Physical and biogeochemical processes are constrained by the monthly variations of sea surface temperature, salinity, chlorophyll concentration, wind speed and mixed-layer depth. The model is applied to four locations in the Indian Ocean and it well predicts observed temporal variations in f_CO2 at these locations. Regarding to monthly f_CO2 observations, the model also well simulates the f_CO2 distribution and its temporal variations along a track located between 20 ° and 50 °S with a maximal error of + 10 μatm. The model is also used to predict f_CO2 for 2 ° × 2 ° grids over the entire Indian Ocean and simulates seasonal cycles that are consistent with observations. The monthly f_CO2 fields derived from the model are used to estimate a global air-sea CO₂ flux over the Indian Ocean basin. We estimate a net sink of 0.5 Gt/yr C for the Indian Ocean (20 °N-50 °S), with the main sink located between 20 ° and 50 °S.     

The international at-sea intercomparison of fCO2 systems during the R/V Meteor Cruise 36/1 in the North Atlantic Ocean

The ‘International Intercomparison Exercise of fCO₂ Systems’ was carried out in 1996 during the R/V Meteor Cruise 36/1 from Bermuda/UK to Gran Canaria/Spain. Nine groups from six countries (Australia, Denmark, France, Germany, Japan, USA) participated in this exercise, bringing together 15 participants with seven underway fugacity of carbon dioxide (fCO₂) systems, one discrete fCO₂ system, and two underway pH systems, as well as systems for discrete measurement of total alkalinity and total dissolved inorganic carbon. Here, we compare surface seawater fCO₂ measured synchronously by all participating instruments. A common infrastructure (seawater and calibration gas supply), different quality checks (performance of calibration procedures for CO₂, temperature measurements) and a common procedure for calculation of final fCO₂ were provided to reduce the largest possible amount of controllable sources of error. The results show that under such conditions underway measurements of the fCO₂ in surface seawater and overlying air can be made to a high degree of agreement (±1 μatm) with a variety of possible equilibrator and system designs. Also, discrete fCO₂ measurements can be made in good agreement (±3 μatm) with underway fCO₂ data sets. However, even well-designed systems, which are operated without any obvious sign of malfunction, can show significant differences of the order of 10 μatm. Based on our results, no “best choice” for the type of the equilibrator nor specifics on its dimensions and flow rates of seawater and air can be made in regard to the achievable accuracy of the fCO₂ system. Measurements of equilibrator temperature do not seem to be made with the required accuracy resulting in significant errors in fCO₂ results. Calculation of fCO₂ from high-quality total dissolved inorganic carbon (C_T) and total alkalinity (A_T) measurements does not yield results comparable in accuracy and precision to fCO₂ measurements.     

Interannual variations of the Antarctic Ocean CO2 uptake from 1986 to 1994

The interannual variations of CO₂ sources and sinks in the surface waters of the Antarctic Ocean (south of 50°S) were studied between 1986 and 1994. An existing, slightly modified one-dimensional model describing the mixed-layer carbon cycle was used for this study and forced by available satellite-derived and climatological data. Between 1986 and 1994, the mean Antarctic Ocean CO₂ uptake was 0.53 Pg C year⁻¹ with an interannual variability of 0.15 Pg C year⁻¹.Interannual variation of the Antarctic Ocean CO₂ uptake is related to the Antarctic Circumpolar Wave (ACW), which affects sea surface temperature (SST), wind-speed and sea-ice extent. The CO₂ uptake in the Antarctic Ocean has increased from 1986 to 1994 by 0.32 Pg C. It was found that over the 9 years, the surface ocean carbon dioxide fugacity (fCO₂) increase was half that of the atmospheric CO₂ increase inducing an increase of the air–sea fCO₂ gradient. This effect is responsible for 60% of the Antarctic Ocean CO₂ uptake increase between 1986 and 1994, as the ACW effect cancels out over the 9 years investigated.     

Time series of carbonate system variables off Otaru coast in Hokkaido, Japan

We report several biogeochemical parameters (dissolved inorganic carbon (DIC), total alkalinity (TA), dissolved oxygen (DO), phosphate (PO₄), nitrate + nitrite (NO₃ + NO₂), silicate (Si(OH)₄)) in a region off Otaru coast in Hokkaido, Japan on a “weekly” basis during the period of April 2002–May 2003. To better understand the long-term temporal variations of the main factors affecting CO₂ flux in this coastal region and its role as a sink/source of atmospheric CO₂, we constructed an algorithm of DIC and TA using other hydrographic properties. We estimated the CO₂ flux across the air–sea interface by using the classical bulk method. During 1998–2003 in our study region, the estimated fCO_2sea ranged about 185–335 μatm. The maximum of fCO_2sea in the summer was primarily due to the change of water temperature. The minimum of fCO_2sea in the early spring can be explained not only by the change of water temperature but also the change of nutrients and chlorophyll-a. To clarify the factors affecting fCO_2sea (water temperature, salinity, and biological activity), we carried out a sensitivity analysis of these effects on the variation of fCO_2sea. In spring, the biological effect had the largest effect for the minimum of fCO_2sea (40%). In summer, the water temperature effect had the largest effect for the maximum of fCO_2sea (25%). In fall, the water temperature effect had the largest effect for the minimum of fCO_2sea (53%). In winter, the biological effect had the largest effect for the minimum of fCO_2sea (35%).We found that our study region was a sink region of CO₂ throughout a year (−0.78 mol/m²/yr). Furthermore, we estimated that the increase of fCO_2sea was about 0.56 μatm/yr under equilibrium with the atmospheric CO₂ content for the period 1998–2003, with the temporal changes in the variables (T, S, PO₄) on fCO_2sea, thus as the maximum trend of each variable on fCO_2sea was 0.22 μatm/yr, and the trend of residual fCO₂ including gas exchange was 0.34 μatm/yr. This result suggests that interaction among variables would affect gas exchange between air and sea effects on fCO_2sea. We conclude that this study region as a representative coastal region of marginal seas of the North Pacific is special because it was measured, but there is no particular significance in comparison to any other area.     

Spatial variability in the partial pressures of CO2 in the northern Bering and Chukchi seas

In the summers of 1999 and 2003, the 1st and 2nd Chinese National Arctic Research Expeditions measured the partial pressure of CO₂ in the air and surface waters (pCO₂) of the Bering Sea and the western Arctic Ocean. The lowest pCO₂ values were found in continental shelf waters, increased values over the Bering Sea shelf slope, and the highest values in the waters of the Bering Abyssal Plain (BAP) and the Canadian Basin. These differences arise from a combination of various source waters, biological uptake, and seasonal warming. The Chukchi Sea was found to be a carbon dioxide sink, a result of the increased open water due to rapid sea-ice melting, high primary production over the shelf and in marginal ice zones (MIZ), and transport of low pCO₂ waters from the Bering Sea. As a consequence of differences in inflow water masses, relatively low pCO₂ concentrations occurred in the Anadyr waters that dominate the western Bering Strait, and relatively high values in the waters of the Alaskan Coastal Current (ACC) in the eastern strait. The generally lower pCO₂ values found in mid-August compared to at the end of July in the Bering Strait region (66–69°N) are attributed to the presence of phytoplankton blooms. In August, higher pCO₂ than in July between 68.5 and 69°N along 169°W was associated with higher sea-surface temperatures (SST), possibly as an influence of the ACC. In August in the MIZ, pCO₂ was observed to increase along with the temperature, indicating that SST plays an important role when the pack ice melts and recedes.     

Ionization of water in seawater

The ion product of water in seawater and the total activity coefficients of hydroxide and hydrogen ions were determined over the temperature range 2° to 35°C, and the salinity range 20‰ to 44‰. At 25°C and 35‰ salinity, the measured values are pK_W^SW = 13.20, f_OH = 0.22, f_H = 0.71 on the molar concentration scale.     

Doubled CO2 impact on subarctic marine carbon cycle: a case study at Ocean Station P

Along with meteorological observations, complementary and systematic oceanographic observations of various physical, biological and chemical parameters have been made at Ocean Station P (OSP) (50°N, 145°W) since the early 1950s. These decadal time scale data have contributed to a better understanding of the physical, biological and chemical processes in the surface layer of the northeastern subarctic region of the Pacific Ocean. These data have demonstrated the importance of the North Pacific in the global carbon cycle and, in particular, the role of biological/chemical processes in the net exchange of CO₂ across the air–sea interface. Although we do not fully comprehend how climatic variations influence marine communities or marine biogeochemistry, previous studies have provided some basic understanding of the mechanisms controlling the seasonal and inter-annual variations of biological and chemical parameters (such as phytoplankton, bacteria, nitrate/ammonium concentration) at OSP, and how they affect the carbon cycling in the subarctic North Pacific. In this study, we investigate how these mechanisms might alter the seasonal variations of these parameters at OSP under a 2XCO₂ condition. We examine these influences using a new biological model calibrated by the climatological data from OSP. For the 2XCO₂ simulation, the biological model is driven off line (i.e., no feedback to the ocean/atmospheric model components) by the climatology plus 2XCO₂−1XCO₂ outputs from a global surface ocean model and the Canadian GCM. Under the 2XCO₂ condition, the upper layer ocean shows an increase in the entrainment rate at the bottom of the mixed layer for OSP during the late autumn and winter seasons, resulting in an increase in the f-ratio. Although there is an overall increase in the primary production (PP) by 3–18%, a decrease in the biomass of small phytoplankton and microzooplankton (due to mesozooplankton grazing) lowers the concentration of dissolved organic matter (DOM) by 4–25%. The model also predicts a significant increase in the concentrations of nitrate and ammonium, and in bacterial production during July and August. Doubling of the atmospheric CO₂ from 330 to 660 ppm forces the marine pCO₂ to increase by about 63%, much of which is driven by an increased flux of CO₂ from the atmosphere to the oceans.     

Increase of anthropogenic CO2 in the Pacific Ocean over the last two decades

The multiple-parameter linear regression method (Monitoring global ocean carbon inventories. Ocean Observing System Development Panel, Texas A&M University, College Station, TX, 1995, 54pp; Global Biogeochem. Cycles 13 (1999) 179) is used to compare inorganic carbon data from the GEOSECS CO₂ survey in the Pacific Ocean in 1973 to the WOCE/JGOFS global CO₂ survey in the 1990s. A model of total dissolved inorganic carbon (DIC) as a function of five variables (AOU, θ, S, Si, and PO₄) has been developed from the recent CO₂ survey data (namely CGC91 and CGC96) in the Pacific Ocean. After correcting for a systematic DIC offset of −30.3±7 μmol kg⁻¹ from the GEOSECS data, the residual DIC based on this model as computed from GEOSECS data has been used to estimate the anthropogenic CO₂ penetration in the Pacific Ocean. In the Northeast Pacific, we obtained an increase of CO₂ of 21.3±7.9 mol m⁻² over the period from GEOSECS in 1973 to CGC91 in 1991. This gives a mean anthropogenic CO₂ uptake rate of 1.3±0.5 mol m⁻² yr⁻¹ over this 17 year time period. In the South Pacific, north of 50°S between 180° and 120°W region, the integrated anthropogenic CO₂ inventory is estimated to be 19.7±5.7 mol m⁻² over the period from GEOSECS in 1974 to CGC96 in 1996. The equivalent mean CO₂ uptake rate is estimated to be 0.9±0.3 mol m⁻² yr⁻¹ over the 22 years. These results are compared with the isopycnal method (Nature 396 (1998) 560) to estimate the anthropogenic CO₂ signal in the Northeast Pacific (30°N, 152°W) at the crossover region between CGC91 and GEOSECS. The results of the isopycnal method are consistent with those derived from the MLR method. Both methods show an increase in anthropogenic CO₂ inventory in the ocean over two decades that is consistent with the increase expected if the ocean uptake has kept pace with the atmospheric CO₂ increase.     

A seasonal carbon budget for the sub-Antarctic Ocean, South of Australia

Changes from winter (July) to summer (February) in mixed layer carbon tracers and nutrients measured in the sub-Antarctic zone (SAZ), south of Australia, were used to derive a seasonal carbon budget. The region showed a strong winter to summer decrease in dissolved inorganic carbon (DIC;  45 µmol/kg) and fugacity of carbon dioxide (fCO₂;  25 µatm), and an increase in stable carbon isotopic composition of DIC (δ¹³C_DIC;  0.5‰), based on data collected between November 1997 and July 1999.The observed mixed layer changes are due to a combination of ocean mixing, air–sea exchange of CO₂, and biological carbon production and export. After correction for mixing, we find that DIC decreases by up to 42 ± 3 µmol/kg from winter (July) to summer (February), with δ¹³C_DIC enriched by up to 0.45 ± 0.05‰ for the same period. The enrichment of δ¹³C_DIC between winter and summer is due to the preferential uptake of ¹²CO₂ by marine phytoplankton during photosynthesis. Biological processes dominate the seasonal carbon budget (≈ 80%), while air–sea exchange of CO₂ (≈ 10%) and mixing (≈ 10%) have smaller effects. We found the seasonal amplitude of fCO₂ to be about half that of a study undertaken during 1991–1995 [Metzl, N., Tilbrook, B. and Poisson, A., 1999. The annual fCO₂ cycle and the air–sea CO₂ flux in the sub-Antarctic Ocean. Tellus Series B—Chemical and Physical Meteorology, 51(4): 849–861.] for the same region, indicating that SAZ may undergo significant inter-annual variations in surface fCO₂. The seasonal DIC depletion implies a minimum biological carbon export of 3400 mmol C/ m² from July to February. A comparison with nutrient changes indicates that organic carbon export occurs close to Redfield values (ΔP:ΔN:ΔC = 1:16:119). Extrapolating our estimates to the circumpolar sub-Antarctic Ocean implies a minimum organic carbon export of 0.65 GtC from the July to February period, about 5–7% of estimates of global export flux. Our estimate for biological carbon export is an order of magnitude greater than anthropogenic CO₂ uptake in the same region and suggests that changes in biological export in the region may have large implications for future CO₂ uptake by the ocean.     

Alkalinity titration in seawater: How accurately can the data be fitted by an equilibrium model?

Displaying “calculated minus observed” data for precise titrations of seawater with strong acid permits direct evaluation of important parameters and detection of systematic errors.At least two data sets from the GEOSECS (Geochemical Ocean Sections) program fit an equilibrium model (which includes carbonate, borate, sulfate, silicate, fluoride, and phosphate) within the most stringent experimental error, less than 2 μmol kg⁻¹. The effect of various parameters on the fit of calculated to observed values depends strongly on pH. Although standard potential E₀, total alkalinity A_t, total carbonate C_t, and first acidity constant of carbon dioxide pK₁ are nearly independent, and can be determined for each data set, other parameters are strongly correlated. Within such groups, all but one parameter must be determined from data other than the titration curve.Adding an acid-base pair to the theoretical model (e.g. C_x=20 μmol kg⁻¹, pK_x=6.2) produces a deviation approaching 20 μmol kg⁻¹ at constant C_t; however, adjustment of C_t by about −18 μmol kg⁻¹ to produce a good fit leaves only ± 1.5 μmol kg⁻¹ residual deviation from the reference values. Thus, at current standards of precision, an unidentified weak acid cannot be distinguished from carbonate purely on the basis of the titration curve shape.There are few full sets of numerical data published, and most show larger systematic errors (3–12 μmol l⁻¹) than the above; one well-defined source is experiments performed in unsealed vessels. Total carbonate can be explicitly obtained as a function of pH by a rearrangement of the titration curve equation; this can reveal a systematic decrease in C_t in the pH range 5–6, as a result of CO₂ gas loss from the titration vessel. Attempts to compensate for this by adjustment of A_t, C_t, or pK₁ produce deviations which mimic those produced by an additional acid-base pair.Changing from the free H⁺ scale (for which [HSO₄⁻] and [HF] are explicit terms in the alkalinity) to the seawater scale (SWS) (where those terms are part of a constant factor multiplying [H⁺]) requires modification of the titration curve equation as well as adjustment of acidity constants. Even with this change, however, omission of pH-dependent terms in [HSO₄⁻] and [HF] produces small systematic errors at low pH.Shifts in liquid junction potential also introduce small systematic errors, but are significant only at pH <3. High-pH errors due to response of the glass electrode to Na⁺ as well as H⁺ can be adequately compensated to pH 9.5 by a linear selectivity expression.