The release of dissolved actinium to the ocean: A global comparison of different end-members

The measurement of short-lived ²²³Ra often involves a second measurement for supported activities, which represents ²²⁷Ac in the sample. Here we exploit this fact, presenting a set of 284 values on the oceanic distribution of ²²⁷Ac, which was collected when analyzing water samples for short-lived radium isotopes by the radium delayed coincidence counting system. The present work compiles ²²⁷Ac data from coastal regions all over the northern hemisphere, including values from ground water, from estuaries and lagoons, and from marine end-members. Deep-sea samples from a continental slope off Puerto Rico and from an active vent site near Hawaii complete the overview of ²²⁷Ac near its potential sources.The average ²²⁷Ac activities of nearshore marine end-members range from 0.4 dpm m^− 3 at the Gulf of Mexico to 3.0 dpm m^− 3 in the coastal waters of the Korean Strait. In analogy to ²²⁸Ra, we find the extension of adjacent shelf regions to play a substantial role for ²²⁷Ac activities, although less pronounced than for radium, due to its weaker shelf source. Based on previously published values, we calculate an open ocean ²²⁷Ac inventory of 1.35 * 10¹⁸ dpm ²²⁷Ac_ex in the ocean, which corresponds to 37 moles, or 8.4 kg. This implies a flux of 127 dpm m⁻² y^− 1 from the deep-sea floor. For the shelf regions, we obtain a global inventory of ²²⁷Ac of 4.5 * 10¹⁵ dpm, which cannot be converted directly into a flux value, as the regional loss term of ²²⁷Ac to the open ocean would have to be included.Ac has so far been considered to behave similarly to Ra in the marine environment, with the exception of a strong Ac source in the deep-sea due to ²³¹Pa_ex. Here, we present evidence of geochemical differences between Ac, which is retained in a warm vent system, and Ra, which is readily released [Moore, W.S., Ussler, W. and Paull, C.K., 2008-this issue. Short-lived radium isotopes in the Hawaiian margin: Evidence for large fluid fluxes through the Puna Ridge. Marine Chemistry]. Another potential mechanism of producing deviations in ²²⁷Ac/²²⁸Ra and daughter isotope ratios from the expected production value of lithogenic material is observed at reducing environments, where enrichment in uranium may occur. The presented data here may serve as a reference for including ²²⁷Ac in circulation models, and the overview provides values for some end-members that contribute to the global Ac distribution.     

Short-lived radium isotopes in the Hawaiian margin: Evidence for large fluid fluxes through the Puna Ridge

We measured significant activities of short-lived radium isotopes, ²²³Ra (half-life = 11 days) and ²²⁴Ra (half-life = 3.7 days), around the margins of the Hawaiian Islands to water depths of 3500 m. These measurements suggest fluid inputs from the basalt to the surrounding ocean. In general ²²³Ra activities were considerably greater than ²²⁴Ra in spite of the expected higher production rate of ²²⁴Ra activity in basalt. The ²²³Ra was not supported by dissolved ²²⁷Ac. The highest enrichments of ²²³Ra were measured over the Puna Ridge (2100 m depth) east of Hawaii. Here ²²³Ra activities reached 19 dpm/m³, similar to activities measured near sites of active submarine groundwater discharge in the South Atlantic Bight. To explain the high activities of ²²³Ra unaccompanied by ²²⁴Ra, we postulate that thermally-driven circulation of seawater through the Puna Ridge deposits ²³¹Pa on basalt surfaces. With time the ²³¹Pa produces ²²⁷Ac and ²²³Ra; and ²²³Ra desorbs into the circulating fluids. These fluids then transport ²²³Ra into the overlying ocean. Based on the inventory of ²²³Ra above the Puna Ridge, we estimate the flow of fluids through the ridge to be on the order of 20–60 cm³ cm^− 2 day^− 1. In less than 1000 years the incoming seawater could provide enough ²³¹Pa to basalt surfaces to balance the inventory of ²²³Ra above the ridge if only 8% of the ²²³Ra was transported to the overlying water. These observations on the flanks of a volcanically-active ocean island have significant implications for quantifying fluid fluxes from the flanks of the mid-ocean ridge system. By mapping ²²³Ra inventories in the ocean above ridge flanks and measuring the activity of ²²³Ra in the emerging fluids, the fluid flux can be obtained.     

Analysis of Ac in seawater by delayed coincidence counting

The activity of ²²⁷Ac in seawater was determined using a delayed coincidence counting technique based on the measurement of the short-lived daughter ²²³Ra (half-life of 11 days). Two 180-l aliquots of Ac free seawater from surface waters of the California borderlands were spiked with Harwell Uraninite standard and the added Ac and Ra were extracted on Mn-fibers. The measured activities on the fibers agreed to within 5% of the equilibrium Ac activity of the spike addition, well within the 2σ counting error of 7% for the measurements. Thus, any recovery error was indistinguishable from counting error. Large volume samples were collected from the water column of the Eastern North Pacific in December of 1998 and processed through Mn-fiber columns. The ²²³Ra activities on the fibers were measured four or more times over a period of 2 months by the technique of [J. Geophys. Res. 101 (1996) 1321] to determine if the activity of ²²³Ra was wholly supported by ²²⁷Ac. In most cases, no significant differences in ²²³Ra activity were observed during repeated counting of the same sample, again implying that ²²³Ra was supported wholly by ²²⁷Ac. In samples where unsupported ²²³Ra were present, samples were counted beyond five half live of the ²²³Ra (>60 days). Two depth profiles of ²²⁷Ac in the Eastern North Pacific showed similar shape and activity to profiles in the Western and Central Pacific measured with alpha-spectrometry by [Nature 310 (1984) 486). Surface waters of the Southern Ocean were relatively enriched in ²²⁷Ac compared to Eastern North Pacific waters, possibly indicating export of the parent ²³¹Pa from Atlantic surface waters or upwelling.     

Biogeochemical transport in the Loxahatchee River estuary, Florida: The role of submarine groundwater discharge

The distributions of dissolved organic carbon (DOC), Ba, U, and a suite of naturally occurring radionuclides in the U/Th decay series (²²²Rn, ^{223,224,226,228}Ra) were studied during high- and low-discharge conditions in the Loxahatchee River estuary, Florida to examine the role of submarine groundwater discharge in estuarine transport. The fresh water endmember of this still relatively pristine estuary may reflect not only river-borne constituents, but also those advected during active groundwater/surface water (hyporheic) exchange. During both discharge conditions, Ba concentrations indicated slight non-conservative mixing. Such Ba excesses could be attributed either to submarine groundwater discharge or particle desorption processes. Estuarine dissolved organic carbon concentrations were highest at salinities closest to zero. Uranium distributions were lowest in the fresh water sites and mixed mostly conservatively with an increase in salinity. Suspended particulate matter (SPM) concentrations were generally lowest (< 5 mg L^− 1) close to zero salinity and increased several-fold ( 18 mg L^− 1; low discharge) toward the seaward endmember, which may be attributed to dynamic resuspension of bottom sediments within Jupiter Inlet.Surface water-column ²²²Rn activities were most elevated (> 28 dpm L^− 1) at the freshwater endmember of the estuary and appear to identify regions of the river most influenced by the discharge of fresh groundwater. Activities of four naturally occurring isotopes of Ra (^{223,224,226,228}Ra) in this estuary and select adjacent shallow groundwater wells yield mean estuarine water-mass transit times of less than 1 day; these values are in close agreement to those calculated by tidal prism and tidal frequency. Submarine groundwater discharge rates to the Loxahatchee River estuary were calculated using a tidal prism approach, an excess ²²⁶Ra mass balance, and an electromagnetic seepage meter. Average SGD rates ranged from 1.0 to 3.8 × 10⁵ m³ d^− 1 (20–74 L m^− 2 d^− 1), depending on river-discharge stage. Such calculated SGD estimates, which must include both a recirculated as well as fresh water component, are in close agreement with results obtained from a first-order watershed mass balance. Average submarine groundwater discharge rates yield NH₄⁺ and PO₄^− 3 flux estimates to the Loxahatchee River estuary that range from 62.7 to 1063.1 and 69.2 to 378.5 μmol m^− 2 d^− 1, respectively, depending on river stage. SGD-derived nutrient flux rates are compared to yearly computed riverine total N and total P load estimates.     

Seasonal changes in submarine groundwater discharge to coastal salt ponds estimated using Ra and Ra as tracers

Submarine groundwater discharge (SGD) to coastal southern Rhode Island was estimated from measurements of the naturally-occurring radioisotopes ²²⁶Ra (t_1/2 = 1600 y) and ²²⁸Ra (t_1/2 = 5.75 y). Surface water and porewater samples were collected quarterly in Winnapaug, Quonochontaug, Ninigret, Green Hill, and Pt. Judith–Potter Ponds, as well as nearly monthly in the surface water of Rhode Island Sound, from January 2002 to August 2003; additional porewater samples were collected in August 2005. Surface water activities ranged from 12–83 dpm 100 L^− 1 (60 dpm = 1 Bq) and 21–256 dpm 100 L^− 1 for ²²⁶Ra and ²²⁸Ra, respectively. Porewater ²²⁶Ra activities ranged from 16–736 dpm 100 L^− 1 (2002–2003) and 95–815 dpm 100 L^− 1 (2005), while porewater ²²⁸Ra activities ranged from 23–1265 dpm 100 L^− 1. Combining these data with a simple box model provided average ²²⁶Ra-based submarine groundwater fluxes ranging from 11–159 L m^− 2 d^− 1 and average ²²⁸Ra-derived fluxes of 15–259 L m^− 2 d^− 1. Seasonal changes in Ra-derived SGD were apparent in all ponds as well as between ponds, with SGD values of 30–472 L m^− 2 d^− 1 (Winnapaug Pond), 6–20 L m^− 2 d^− 1 (Quonochontaug Pond), 36–273 L m^− 2 d^− 1 (Ninigret Pond), 29–76 L m^− 2 d^− 1 (Green Hill Pond), and 19–83 L m^− 2 d^− 1 (Pt. Judith–Potter Pond). These Ra-derived fluxes are up to two orders of magnitude higher than results predicted by a numerical model of groundwater flow, estimates of aquifer recharge for the study period, and values published in previous Ra-based SGD studies in Rhode Island. This disparity may result from differences in the type of flow (recirculated seawater versus fresh groundwater) determined using each technique, as well as variability in porewater Ra activity.     

Transformation dynamics and reactivity of dissolved and colloidal iron in coastal waters

We have investigated the chemical forms, reactivities and transformation kinetics of Fe(III) species present in coastal water with ion exchange and filtration methods. To simulate coastal water system, a mixture of ferric iron and fulvic acid was added to filtered seawater and incubated for a minute to a week. At each incubation time, the seawater sample was acidified with hydrochloric acid and then applied to anion exchange resin (AER) to separate negatively charged species (such as fulvic acid, its complexes with iron and iron oxyhydroxide coated with fulvic acid) from positively charged inorganic ferric iron (Fe(III)′). By monitoring the acid-induced Fe(III)′ over an hour, it was found that iron complexed by fulvic acid dissociated rapidly to a large extent (86–92% at pH 2), whereas amorphous ferric oxyhydroxide particles associated with fulvic acid (AFO-L) dissociated very slowly with the first-order dissociation rate constants ranging from 6.1 × 10^− 5 for pH 3 to 2.7 × 10^− 4 s^− 1 for pH 2. Therefore, a brief acidification followed by the AER treatment (acidification/AER method) was likely to be able to determine fulvic acid complexes and thus differentiate the complexes from the AFO-L particles (the dissolution of AFO-L was insignificant during the brief acidification). The acidification/AER method coupled with a simple filtration technique suggested that the iron–fulvic acid complexes exist in both the < 0.02 μm and 0.02–0.45 μm size fractions in our coastal water system. The truly dissolved iron (< 0.02 μm) was relatively long-lived with a life-time of 14 days, probably due to the complexation by strong ligands. Such an acid-labile iron may be an important source of bioavailable iron in coastal environments, as a significant relationship between the chemical lability and bioavailability of iron has been well recognised.     

Depth profiles of volatile iodine and bromine-containing halocarbons in coastal Antarctic waters

Measurements of bromoform (CHBr₃), diiodomethane (CH₂I₂), chloroiodomethane (CH₂ICl) and bromoiodomethane (CH₂IBr) were made in the water column (5–100 m depth) of the Southern Ocean within 0–40 km of the Antarctic sea ice during the ANTXX1/2 transect of the German R/V Polarstern, at five locations between 70–72°S and 9–11°W in the Antarctic spring/summer of 2003–2004. Some of the profiles exhibited a very pronounced layer of surface sea-ice meltwater, as evidenced by salinity minima and temperature maxima, along with surface maxima in concentrations of CHBr₃, CH₂I₂, CH₂ICl and CH₂IBr. These results are consistent with in situ surface halocarbon production by ice algae liberated from the sea ice, although production within the sea ice followed by transport cannot be entirely ruled out. Additional sub-surface maxima in halocarbons occurred between 20 and 80 m. At a station further from shore and not affected by surface sea-ice meltwater, surface concentrations of CH₂I₂ were decreased whereas CH₂ICl concentrations were increased compared to the stations influenced by meltwater, consistent with photochemical conversion of CH₂I₂ to CH₂ICl, perhaps during upward mixing from a layer at  70 m enhanced in iodocarbons. Mean surface (5–10 m) water concentrations of halocarbons in these coastal Antarctic waters were 57 pmol l^− 1 CHBr₃ (range 44–78 pmol l^− 1), 4.2 pmol l^− 1 CH₂I₂ (range 1.7–8.2 pmol l^− 1), 0.8 pmol l^− 1 CH₂IBr (range 0.2–1.4 pmol l^− 1), and 0.7 pmol l^− 1 CH₂ICl (range 0.2–2.4 pmol l^− 1). Concurrent measurements in air suggested a sea-air flux of bromoform near the Antarctic coast of between 1 and 100 (mean 32.3, median 10.4) nmol m^− 2 day^− 1 and saturation anomalies of 557–1082% (mean 783%, median 733%), similar in magnitude to global shelf values. In surface samples affected by meltwater, CH₂I₂ fluxes ranged from 0.02 to 6.1 nmol m^− 2 day^− 1, with mean and median values of 1.9 and 1.1 nmol m^− 2 day^− 1, respectively.     

TIMS measurements of Ra and Ra in the Gulf of Lion, an attempt to quantify submarine groundwater discharge

Submarine groundwater discharge (SGD) is now recognized as an important pathway for water and chemical species fluxes to the coastal ocean. In order to determinate SGD to the Gulf of Lion (France), we measured the activities of ²²⁶Ra and ²²⁸Ra by thermal ionization mass spectrometry (TIMS) in coastal waters and in the deep aquifer waters of the Rhone deltaic plain after pre-concentration of radium by MnO₂. Compared to conventional counting techniques, TIMS requires lower quantities of water for the analyses, and leads to higher analytical precision. Radium isotopes were thus measured on 0.25–2 L water samples containing as little as 20 fg of ²²⁶Ra and 0.2–0.4 fg of ²²⁸Ra with precision equal to 2%. We demonstrate that coastal surface waters samples are enriched in ²²⁶Ra and ²²⁸Ra compared to the samples further offshore. The high precision radium measurements display a small but significant ²²⁶Ra and ²²⁸Ra enrichment within a strip of circa 30 km from the coast. Radium activities decrease beyond this region, entrained in the northern current along the shelf break or controlled by eddy diffusion. The radium excess in the first 30 km cannot be accounted for by the river nor by the early diagenesis. The primary source of the radium enrichment must therefore be ascribed to the discharge of submarine groundwater. Using a mass-balance model, we estimated the advective fluxes of ²²⁶Ra and ²²⁸Ra through SGD to be 5.2 × 10¹⁰ and 21 × 10¹⁰ dpm/d respectively. The ²²⁶Ra activities measured in the groundwater from the Rhone deltaic plain aquifer are comparable to those from other coastal groundwater studies throughout the world. By contrast, ²²⁸Ra activities are higher by up to one order of magnitude. Taking those groundwater radium activities as typical of the submarine groundwater end-member, a minimum volume of 0.24–4.5 × 10¹⁰ l/d is required to support the excess radium isotopes on the inner shelf. This has to be compared with the average rivers water runoff of 15.4 × 10¹⁰ l/d during the study period (1.6 to 29% of the river flow).     

Uncertainties in the preparation of Ra Mn fiber standards

Delayed coincidence counters (RaDeCC), used for measuring ²²³Ra and ²²⁴Ra preconcentrated from water onto MnO₂-impregnated acrylic fiber (“Mn-fiber”), require a standard Mn-fiber column that has a precisely known activity of ²²⁴Ra for calibration. This may be done by adding an aged ²²⁸Th standard solution to adsorb both ²²⁸Th and its daughter ²²⁴Ra quantitatively onto a Mn fiber. We used both seawater and deionized water (DIW) for testing the adsorption efficiency of Th and Ra onto Mn fibers. Our experimental results show that more than 50% of thorium (²³²Th and ²²⁸Th) breaks through the Mn-fiber column when DIW is used as a medium. However, near quantitative recoveries are obtained if filtered (0.45 μm) seawater is used to prepare the standard. In the case of pure DIW, the pH (initial pH  5.3) rises to > 10 after passing through the column while seawater (initial pH  7.8) changes to  7.2. Thus, the lack of thorium adsorption in DIW may be attributed to this huge increase of pH and the consequent formation of Th(OH)₄ and polyhydroxyl colloids. Based on these observations, we recommend that one should use either artificial seawater or natural seawater (which has negligible ²²⁴Ra and ²²⁸Th) as a loading solution after 0.45 μm filtration. In addition, the thorium adsorption efficiency should be confirmed either by thorium analysis of the effluent solution or long-term monitoring of the supported ²²⁴Ra on the Mn fiber using the RaDeCC. Similar cautions are likely necessary for making ²²³Ra standards by adsorption of ²²⁷Ac onto Mn fibers.     

Photochemical production of molecular hydrogen in lake water and coastal seawater

Samples of lake water and coastal seawater from Nova Scotia, Canada, were irradiated with natural or artificial sunlight to investigate the potential for photochemical hydrogen production. Hydrogen photo-production was observed in all natural water samples. Rates of hydrogen formation were highest in coloured lake water (range: 98–163 pmol L^− 1h^− 1) and lower in seawater (range: 19–45 pmol L^− 1 h ^− 1). Dilutions of the most highly coloured lake sample (Kejimkujik Lake) showed a positive linear relationship between H₂ production rates and CDOM concentration. Photo-production rates normalised to UV absorption coefficients at 350 nm indicated that the photochemical efficiency of hydrogen formation varied between samples, perhaps due to differences in the CDOM composition. Photochemical hydrogen formation was also seen in solutions of syringic acid and acetaldehyde: two low-molecular-weight carbonyl compounds found in natural waters. Photochemistry may therefore offer least a partial explanation for the persistently high levels of hydrogen observed in the low-latitude surface ocean.     

Water mass ages of coastal ponds estimated using Ra and Ra as tracers

Measurements of the naturally occurring radioisotopes ²²³Ra (t_1/2 = 11.4 days) and ²²⁴Ra (t_1/2 = 3.66 days) in southern Rhode Island salt ponds were combined with a simple model to obtain independent estimates of the age of these coastal waters. Surface water and porewater samples were collected quarterly in Winnapaug, Quonochontaug, Ninigret, Green Hill, and Pt. Judith-Potter Ponds, as well as nearly monthly in the surface water of Rhode Island Sound, beginning January 2002 through August 2003. Surface water activities ranged from 1–78 dpm 100 L^− 1 and 5–885 dpm 100 L^− 1 for ²²³Ra and ²²⁴Ra, respectively. Porewater radium activities ranged from 3 to 715 dpm 100 L^− 1 for ²²³Ra, and 57–4926 dpm 100 L^− 1 for ²²⁴Ra. Results indicate seasonally varying water mass ages for Ninigret (5–12 days), Winnapaug (2–6 days) and Pt. Judith-Potter Ponds (1–9 days) and, in contrast, relatively constant ages for Green Hill (5–7 days) and Quonochontaug Ponds (3–6 days).     

Estimating submarine groundwater discharge around Isola La Cura, northern Venice Lagoon (Italy), by using the radium quartet

The four naturally-occurring radium isotopes (²²³Ra, ²²⁴Ra, ²²⁶Ra and ²²⁸Ra) were used to estimate the submarine groundwater discharge (SGD) in the Isola La Cura marsh area in the northern Venice Lagoon (Italy). By determining the radium contributors to the study area (river, coastal ocean and sediments) the radium excess in the lagoon water was quantified through a mass balance model. This radium excess is attributed to a submarine groundwater discharge source and represents the most important input of radium. Possible endmembers were considered from analysis of groundwater samples (subtidal and marsh piezometers, marsh wells and seepage meters) that were enriched in Ra by one to two orders of magnitude relative to surface waters. In particular, a permeable layer at 80 cm depth in the surrounding marsh is considered to be representative of the most likely SGD source, although similar radium activities were measured in other subtidal porewater samples collected in the Isola La Cura area. The estimated SGD flux to the study area ranged from 1 · 10⁹ to 6 · 10⁹ L·d^− 1, the same order of magnitude as the overall riverine input to the lagoon (3 · 10⁹ L·d^− 1). A major fraction of this SGD flux is likely recirculated seawater, as evidenced by the endmember salinity. The water residence time of 2 days was estimated by both using the shortest-lived radium isotope and estimating the volume of water exchanged between the lagoon and the open sea during a tidal cycle (tidal prism approach). This SGD flux could be used to estimate the input of other chemical species (metals, nutrients, etc.) via SGD which might affect the Venice Lagoon ecosystem.     

Comparison of β-dimethylsulfoniopropionate (DMSP) levels in two mediterranean ecosystems with different trophic levels

β-dimethylsulfoniopropionate (DMSP) and dimethylsulfide (DMS) concentrations were recorded from September 1999 to September 2000 in two geographically close ecosystems, differently affected by eutrophication: the Little Bay of Toulon and the Niel Bay (N.W. Mediterranean Sea, France). Little Bay had higher nutrient levels ([NO₃⁻]_max. = 30.3 μM; [PO₄³⁻]_max. = 0.46 μM) and higher chlorophyll a concentrations ([chl a]_mean = 2.4 μg/L) compared to Niel Bay ([NO₃⁻]_max. = 19.7 μM; [PO₄³⁻]_max. = 0.17 μM; [chl a]_mean = 0.4 μg/L). In the two sites, we measured dissolved (DMSP_d < 0.2 μm) and particulate DMSP (DMSP_p > 0.2 μm) concentrations. The DMSP_p was particularly analysed in the 0.2–5, 5–90 and > 90 μm fractions. In the eutrophicated Little Bay, DMSP_d concentrations showed a clear seasonality with high values from January to March (124–148 nM). The temporal profile of the DMSP_p concentrations was similar, peaking in February–March (38–59 nM). In the less eutrophic Niel Bay, DMSP_p concentrations were much lower (6–9 nM in March–April), whereas DMSP_d concentrations were relatively high (110–92 nM in February–March). DMS concentrations were elevated from the end of the winter to the spring in Little Bay, ranging from 3 nM in October to 134 nM in March. In the less eutrophic Niel Bay, lower DMS levels were observed, generally not exceeding 20 nM. Each particulate fraction (0.2–5; 5–90; > 90 μm) contained less DMSP in Niel Bay than in Little Bay. At both sites, the 5–90 μm fraction made up most of the DMSP_p. This 5–90 μm fraction consisted of microphytoplankton, principally Dinophyceae and Bacillariophyceae. The 5–90 μm biomass calculated from cell biovolumes, was more abundant in Little Bay where the bloom at the end of the winter (165 μg/L in March) occurred at the same time as the DMSP peaks. The estimated DMSP_p to biomass ratio for the 5–90 μm fraction was always higher in Little Bay than in Niel Bay. This suggests that the high DMSP levels recorded in Little Bay were not only due to a large Dinophyceae presence in this ecosystem. Indeed, the peak of DMSP_p to biomass ratio obtained from cell biovolumes (0.23 nmol/μg in March) was consistent with the proliferation of Alexandrium minutum. This Dinophyceae species may account for between 50% (2894 cells/L) and 63% (4914 cells/L) of the total phytoplankton abundance in the Little Bay of Toulon.     

Kinetic and equilibrium studies of copper-dissolved organic matter complexation in water column of the stratified Krka River estuary (Croatia)

An interaction of dissolved natural organic matter (DNOM) with copper ions in the water column of the stratified Krka River estuary (Croatia) was studied. The experimental methodology was based on the differential pulse anodic stripping voltammetric (DPASV) determination of labile copper species by titrating the sample using increments of copper additions uniformly distributed on the logarithmic scale. A classical at-equilibrium approach (determination of copper complexing capacity, CuCC) and a kinetic approach (tracing of equilibrium reconstitution) of copper complexation were considered and compared. A model of discrete distribution of organic ligands forming inert copper complexes was applied. For both approaches, a home-written fitting program was used for the determination of apparent stability constants (K_i^equ), total ligands concentration (L_iT) and association/dissociation rate constants (k_i¹,k_i^- 1).A non-conservative behaviour of dissolved organic matter (DOC) and total copper concentration in a water column was registered. An enhanced biological activity at the freshwater–seawater interface (FSI) triggered an increase of total copper concentration and total ligand concentration in this water layer. The copper complexation in fresh water of Krka River was characterised by one type of binding ligands, while in most of the estuarine and marine samples two classes of ligands were identified. The distribution of apparent stability constants (log K₁^equ: 11.2–13.0, log K₂^equ:8.8–10.0) showed increasing trend towards higher salinities, indicating stronger copper complexation by autochthonous seawater organic matter.Copper complexation parameters (ligand concentrations and apparent stability constants) obtained by at-equilibrium model are in very good accordance with those of kinetic model. Calculated association rate constants (k₁¹:6.1–20 × 10³ (M s)^− 1, k₂¹: 1.3–6.3 × 10³ (M s)^− 1) indicate that copper complexation by DNOM takes place relatively slowly. The time needed to achieve a new pseudo-equilibrium induced by an increase of copper concentration (which is common for Krka River estuary during summer period due to the nautical traffic), is estimated to be from 2 to 4 h.It is found that in such oligotrophic environment (dissolved organic carbon content under 83 µM_C, i.e. 1 mg_CL^− 1) an increase of the total copper concentration above 12 nM could enhance a free copper concentration exceeding the level considered as potentially toxic for microorganisms (10 pM).     

Submarine groundwater discharge: A large, previously unrecognized source of dissolved iron to the South Atlantic Ocean

This paper reports the initial results of a study of groundwater and coastal waters of southern Brazil adjacent to a 240 km barrier spit separating the Patos Lagoon, the largest coastal lagoon in South America, from the South Atlantic Ocean. The objective of this research is to assess the chemical alteration of freshwater and freshwater–seawater mixtures advecting through coastal permeable sands, and the influence of the submarine discharge of these fluids (SGD) on the chemistry of coastal waters. Here we focus on dissolved iron in this system and use radium isotopic tracers to quantify SGD and cross-shelf fluxes. Iron concentrations in groundwaters vary between 0.6 and 180 μM. The influence of the submarine discharge of these fluids into the surf zone produces dissolved Fe concentrations as high as several micromolar in coastal surface waters. The offshore gradient of dissolved Fe, coupled with results for Ra isotopes, is used to quantify the SGD flux of dissolved Fe from this coastline. We estimate the SGD flux to be 2 × 10⁶ mol day^− 1 and the cross-shelf flux to be 3.2 × 10⁵ mol day^− 1. This latter flux is equal to about 10% of the soluble atmospheric Fe flux to the entire South Atlantic Ocean. We speculate on the importance of this previously unrecognized iron input to regional ocean production and on the potential significance of this source to understanding variations in glacial–interglacial ocean production.     

Macrozoobenthic biomass in the Bay of Seine (eastern English Channel)

The benthic biomass values of various trophic groups were studied for the first time at the scale of the entire Bay of Seine (50 × 100 km) in the eastern English Channel. Sampling was carried out before and after the winter of 1999. In both cruises the suspension feeders dominated (66% of the 12 gAFDW m^− 2 in 1998 and 56% of the 10 gAFDW m^− 2 in 1999).The common European ophiuroid Ophiothrix fragilis was the most important contributor to total biomass. The repartition of its patches (sometimes > 20 gAFDW m^− 2) cannot be explained by the environmental parameters recorded (viz., granulometry, organic matter and pigment content).     

The impact of groundwater discharges on mercury partitioning, speciation and bioavailability to mussels in a coastal zone

The Mussel Watch program conducted along the French coasts for the last 20 years indicates that the highest mercury concentrations in the soft tissue of the blue mussel (Mytilus edulis) occur in animals from the eastern part of Seine Bay on the south coast of the English Channel, the “Pays de Caux”. This region is characterized by the presence of intertidal and submarine groundwater discharges, and no particular mercury effluent has been reported in its vicinity. Two groundwater emergence systems in the karstic coastal zone of the Pays de Caux (Etretat and Yport with slow and fast water percolation pathways respectively) were seasonally sampled to study mercury distribution, partitioning and speciation in water. Samples were also collected in the freshwater–seawater mixing zones in order to compare mercury concentrations and speciation between these “subterranean” or “groundwater” estuaries and the adjacent macrotidal Seine estuary, characterized by a high turbidity zone (HTZ). The mercury concentrations in the soft tissue of mussels from the same areas were monitored at the same time.The means of the “dissolved” (< 0.45 μm) mercury concentrations (HgT_D) in the groundwater springs were 0.99 ± 0.15 ng l^− 1 (n = 18) and 0.44 ± 0.17 ng l^− 1 (n = 17) at Etretat and Yport respectively. High HgT_D concentrations were associated with strong runoff over short water pathways during storm periods, while low concentrations were associated with long groundwater pathways. Mean particulate mercury concentrations were 0.22 ± 0.05 ng mg^− 1 (n = 16) and 0.16 ± 0.10 ng mg^− 1 (n = 17) at Etretat and Yport respectively, and decreased with increasing particle concentration probably as a result of dilution by particles from soil erosion. Groundwater mercury speciation was characterized by high reactive-to-total mercury ratios in the dissolved phase (HgR_D/HgT_D: 44–95%), and very low total monomethylmercury concentrations (MMHg < 8 pg l^− 1). The HgT_D distributions in the Yport and Etretat mixing zones were similar (overall mean concentration of 0.73 ± 0.21 ng l^− 1, n = 43), but higher than those measured in the adjacent industrialized Seine estuary (mean: 0.31 ± 0.11 ng l^− 1, n = 67). In the coastal waters along the Pays de Caux dissolved monomethylmercury (MMHg_D) concentrations varied from 9.5 to 13.5 pg l^− 1 (2 to 8% of the HgT_D). Comparable levels were measured in the Seine estuary (range: 12.2– 21.1 pg l⁻¹; 6–12% of the HgT_D). These groundwater karstic estuaries seem to be mostly characterized by the higher HgT_D and HgR_D concentrations than in the adjacent HTZ Seine estuary. While the HTZ of the Seine estuary acts as a dissolved mercury removal system, the low turbid mixing zone of the Pays de Caux receives the dissolved mercury inputs from the groundwater seepage with an apparent Hg transfer from the particulate phase to the “dissolved” phase (< 0.45 μm). In parallel, the soft tissue of mussels collected near the groundwater discharges, at Etretat and Yport, exhibited significantly higher values than those found in the mussel from the mouth of the Seine estuary. We observe that this difference mimics the differences found in the mercury distribution in the water, and argue that the dissolved phase of the groundwater estuaries and coastal particles are significant sources of bioavailable mercury for mussels.     

Aggregation of riverine colloidal iron in estuaries: A new kinetic study using stopped-flow mixing

The kinetics of aggregation of riverine colloidal iron have been examined using a stopped-flow technique which probes the first few seconds of mixing between river and sea water end members. A significant fraction, up to 80%, of the colloidal iron is aggregated during the first 1–2 s of mixing, indicating that the aggregation process is much faster than previously thought. Most of the aggregation induced by seawater results from the divalent cations Mg²⁺ and Ca²⁺, with the overall ionic strength having little influence. At equal concentrations of 27 mM, the rate of aggregation by alkaline earth cations increased with ionic size Mg²⁺ < Ca²⁺ < Sr²⁺ < Ba²⁺. The aggregation rates were indifferent to the anion (Cl⁻ or SO₄²⁻) present. Very high aggregation rates were also induced by the common water treatment coagulant Al₂(SO₄)₃ at concentrations in the range 20–30 µM Al(III), several orders of magnitude lower that those used for the seawater cations. Our results support the view that specific chemical interactions between the cations and the colloid particle surface, rather than simple electrical effects, control the colloid stability.     

Environmental controls on phytoplankton community composition in the Thames plume, U.K.

The aim of this study was to investigate controls on the phytoplankton community composition and biogeochemistry of the estuarine plume zone of the River Thames, U.K. using an instrumented moored buoy for in situ measurements and preserved sample collection, and laboratory-based measurements from samples collected at the same site. Instrumentation on the moored buoy enabled high frequency measurements of a suite of environmental variables including in situ chlorophyll, water-column integrated irradiance, macronutrients throughout an annual cycle for 2001 e.g. nitrate and silicate, and phytoplankton biomass and species composition. The Thames plume region acts as a conduit for fluvial nutrients into the wider southern North Sea with typical winter concentrations of 45 μM nitrate, 17 μM silicate and 2 μM phosphate measured. The spring bloom resulted from water-column integrated irradiance increasing above 60 W h m^− 2 d^− 1 and was initially dominated by a diatom bloom mainly composed of Nitzschia sp. and Odontella sinesis. The spring bloom then switched after  30 days to become dominated by the flagellate Phaeocystis reaching a maximum chlorophyll concentration of 37.8 μg L^− 1. During the spring bloom there were high numbers of the heterotrophic dinoflagellates Gyrodinium spirale and Katodinium glaucum that potentially grazed the phytoplankton bloom. This diatom–flagellate switch was predicted to be due to a combination of further increasing water-column integrated irradiance > 100 W h m^− 2 d^− 1 and/or silicate reaching potentially limiting concentrations (< 1 μM). Post spring bloom, diatom dominance of the lower continuous summer phytoplankton biomass occurred despite the low silicate concentrations (Av. 0.7 μM from June–August). Summer diatom dominance, generally due to Guinardia delicatula, was expected to be as a result of microzooplankton grazing, dominated by the heterotrophic dinoflagellate Noctiluca scintillans, controlling 0.7–5.0 μm ‘flagellate’ fraction of the phytoplankton community with grazing rates up to 178% of ‘flagellate’ growth rate. The Thames plume region was therefore shown to be an active region of nutrient and phytoplankton processing and transport to the southern North Sea. The use of a combination of moorings and ship-based sampling was essential in understanding the factors influencing nutrient transport, phytoplankton biomass and species composition in this shelf sea plume region.     

Origin of iron and aluminium in large particles (> 53 µm) in the Crozet region, Southern Ocean

Natural iron fertilization processes are occurring around the Crozet Islands (46°26′S–52°18′E), thus relieving the water masses from the normally encountered High Nutrients Low Chlorophyll (HNLC) conditions of the Southern Ocean. During austral summers 2004/2005 and 2005/2006, iron and aluminium concentrations were investigated in large particles (> 53 µm) collected from just below the mixed layer at stations under the influence of island inputs, and also in adjacent HNLC waters. These large particles are anticipated to sink out of the mixed layer, and to reflect the net effects of input and cycling of these elements in the overlying mixed layer. Labile and refractory fractions were determined by a two-stage leaching technique. Data showed that water masses downstream of the islands were enriched in total iron and aluminium (0.25–2.68 nmol L^− 1 and 0.34–3.28 nmol L^− 1 respectively), relative to the southern HNLC control sites (0.15–0.29 nmol L^− 1 for Fe and 0.12–0.29 nmol L^− 1 for Al), with only a small fraction (typically < 1%) being acid leachable in both environments. Particulate iron predominantly derived from the island system represents a significant fraction of the total water column iron inventory and may complement dissolved Fe inputs that help support the high summer productivity around the Crozet islands.