Gregory A. Cutter

Species determination of selenium in natural waters

Abstract A method is described for the determination in natural waters of selenite, selenate, dimethyl selenide, and dimethyl diselenide. The detection limits are in the parts per trillion range. The volatile methyl species are removed from the sample with a stripping gas. The inorganic forms are selectively reduced to the hydride, and also stripped from the sample. A liquid nitrogen trap is used to collect both the selenides and the generated hydrides. Separation of the methyl species is accomplished by gas chromatography. All the species are detected by an atomic absorption spectrometer equipped with a quartz tube furnace.

Selenium in reducing waters.

The analysis of selenium species in reducing waters provides important insight into the elements biogeochemical cycle. The absence of selenate and selenite in reducing waters suggests that some removal mechanism could be operative, but the presence in these waters of about 1 nanomole per liter of dissolved organic selenide indicates that the regeneration of selenium in the form of organic species may be the dominant process. The data demonstrate that the regenerative and biogeochemical cycles of selenium are quite complex.

Behavior of dissolved antimony, arsenic, and selenium in the Atlantic Ocean

Abstract Vertical profiles for dissolved antimony, arsenic, and selenium were obtained at four stations in the eastern basins of the North and South Atlantic Ocean, and on a surface-water transect from 24 ° S to 31 ° N. Total dissolved selenium displays surface-water depletion and deep-water enrichment, with organic selenide (selenium in soluble peptides) being the predominant species in surface waters and selenate predominating in deep waters. Although the concentrations of total selenium in surface waters of the Northern and Southern Hemispheres are similar, the ratio of inorganic to organic selenium is strongly influenced by the intensity of upwelling. Total inorganic arsenic is depleted in the surface waters of all stations, and increases to relatively constant deep-water concentrations (c. 20 nmol/l). In contrast, total inorganic antimony shows surface-water maxima at two stations. Although the average surface-water arsenic and antimony concentrations in the Northern and Southern Hemispheres are identical, there is some evidence for the atmospheric deposition of antimony. Overall, the cycling of the metalloids in the Atlantic is dominated by in situ biotic reactions, and modified by inputs from upwelling and atmospheric deposition.

Antimony and Arsenic Biogeochemistry in the Western Atlantic Ocean

The subtropical to equatorial Atlantic Ocean provides a unique regime in which one can examine the biogeochemical cycles of antimony and arsenic. In particular, this region is strongly affected by inputs from the Amazon River and dust from North Africa at the surface, and horizontal transport at depth from highlatitude northern (e.g., North Atlantic Deep Water) and southern waters (e.g., Antarctic Bottom and Intermediate Waters). As a part of the 1996 Intergovernmental Oceanographic Commission’s Contaminant Baseline Survey, data for dissolved As(III+V), As(III), mono- and dimethyl arsenic, Sb(III+V), Sb(III), and monomethyl antimony were obtained at six vertical profile stations and 44 sites alongthe 11,000 km transect from Montevideo, Uruguay, to Bridgetown, Barbados. The arsenic results were similar to those in other oceans, with moderate surface depletion, deep-water enrichment, a predominance of arsenate (>85% As(V)), and methylated arsenic species and As(III) in surface waters that are likely a result of phytoplankton conversions to mitigate arsenate ‘‘stress’’ (toxicity). Perhaps the most significant discovery in the arsenic results was the extremely low concentrations in the Amazon Plume (as low as 9.8 nmol/l) that appear to extend for considerable distances offshore in the equatorial region. The very low concentration of inorganic arsenic in the Amazon River (2.8 nmol/l; about half those in most rivers) is probably the result of intense iron oxyhydroxide scavenging. Dissolved antimony was also primarily in the pentavalent state (>95% antimonate), but Sb(III) and monomethyl antimony were only detected in surface waters and displayed no correlations with biotic tracers such as nutrients and chlorophyll a. Unlike As(III+V)’s nutrient-type vertical profiles, Sb(III+V) displayed surface maxima and decreased into the deep waters, exhibitingthe behavior of a scavenged element with a strongatmospheric input. While surface water Sb had a slight correlation with dissolved Al, it is likely that atmospheric Sb is delivered with combustion byproducts and not from mineral aerosols. In the Amazon Plume, antimony concentrations dropped substantially, and an Amazon River sample had a concentration (0.25 nmol/l) that was less than one-fourth those found in other major rivers. Usingthese river data, and estimates of atmospheric fluxes based on

Kinetic controls on metalloid speciation in seawater

Mechanisms for the production of thermodynamically unstable species of three metalloids, antimony, arsenic, and selenium, and the rates of their transformation to stable forms have been critically reviewed. The occurrence of thermodynamically unstable species of these metalloids falls into two categories: reduced species in oxic waters and oxidized species in anoxic water. In surface waters biotic processes produce reduced species (e.g. Sb(III), arsenite, selenite), while in deep anoxic waters lateral advection and sinking detritus deliver oxidized species (e.g. arsenate, antimonate) from surface waters. In both cases, slow rates of conversion allow unstable species to persist. Dissolved and particulate data for antimony, arsenic, and selenium speciation in the Black Sea are used to illustrate the processes involved in this kinetic stabilization.

Assessing selenium cycling and accumulation in aquatic ecosystems

We conducted a joint experimental research and modeling study to develop a methodology for assessing selenium (Se) toxicity in aquatic ecosystems. The first phase of the research focused on Se cycling and accumulation. In the laboratory, we measured the rates and mechanisms of accumulation, transformation, and food web transfer of the various chemical forms of Se that occur in freshwater ecosystems. Analytical developments helped define important Se forms. We investigated lower trophic levels (phytoplankton and bacteria) first before proceeding to experiments for each successive trophic component (invertebrates and fish). The lower trophic levels play critical roles in both the biogeochemical cycling and transfer of Se to upper trophic levels. The experimental research provided the scientific basis and rate parameters for a computer simulation model developed in conjunction with the experiments. The model includes components to predict the biogeochemical cycling of Se in the water column and sediments, as well as the accumulation and transformations that occur as Se moves through the food web. The modeled processes include biological uptake, transformation, excretion, and volatilization; oxidation and reduction reactions; adsorption; detrital cycling and decomposition processes; and various physical transport processes within the water body and between the water column and sediments. When applied to a Se-contaminated system (Hyco Reservoir), the model predicted Se dynamics and speciation consistent with existing measurements, and examined both the long-term fate of Se loadings and the major processes and fluxes driving its biogeochemical cycle.

The estuarine behaviour of selenium in San Francisco Bay

Abstract In April and September 1986 concentrations of dissolved selenate, selenite and Se(−II+O), suspended particulate selenium, nutrients, chlorophyll a and total suspended matter, were determined in the San Francisco Bay estuarine system. In addition, dissolved selenium speciation was determined in the Sacramento and San Joaquin Rivers between 1984 and 1987. The April 1986 estuarine sampling occurred during high river discharge, and within the Northern Reach of San Francisco Bay mid-estuarine input of selenite and Se(−II+O) is apparent, while selenate appears to be removed. During September 1986 river discharge rates were approximately four orders of magnitude lower than in April, and the mid-estuarine production of all selenium species is apparent. In contrast, dissolved selenium in the South San Francisco Bay generally shows conservative mixing behaviour during April and September 1986. The source of dissolved selenium in the South Bay appears to be effluents from sewage treatment plants. In the Northern Reach effluents from oil refineries located in the mid-estuary may be major sources of selenium input during low river discharge periods. However, during periods of high river discharge the sources and sinks of dissolved selenium species within the Northern Reach remain unidentified.

Increased selenium threat as a result of invasion of the exotic bivalve Potamocorbula amurensis into the San Francisco Bay-Delta

Following the aggressive invasion of the bivalve, Potamocorbula amurensis, in the San Francisco Bay-Delta in 1986, selenium contamination in the benthic food web increased. Concentrations in this dominant (exotic) bivalve in North Bay were three times higher in 1995-1997 than in earlier studies, and 1990 concentrations in benthic predators (sturgeon and diving ducks) were also higher than in 1986. The contamination was widespread, varied seasonally and was greater in P. amurensis than in co-occurring and transplanted species. Selenium concentrations in the water column of the Bay were enriched relative to the Sacramento River but were not as high as observed in many contaminated aquatic environments. Total Se concentrations in the dissolved phase never exceeded 0.3 microg Se per l in 1995 and 1996; Se concentrations on particulate material ranged from 0.5 to 2.0 microg Se per g dry weight (dw) in the Bay. Nevertheless, concentrations in P. amurensis reached as high as 20 microg Se per g dw in October 1996. The enriched concentrations in bivalves (6-20 microg Se per g dw) were widespread throughout North San Francisco Bay in October 1995 and October 1996. Concentrations varied seasonally from 5 to 20 microg Se per g dw, and were highest during the periods of lowest river inflows and lowest after extended high river inflows. Transplanted bivalves (oysters, mussels or clams) were not effective indicators of either the degree of Se contamination in P. amurensis or the seasonal increases in contamination in the resident benthos. Se is a potent environmental toxin that threatens higher trophic level species because of its reproductive toxicity and efficient food web transfer. Bivalves concentrate selenium effectively because they bioaccumulate the element strongly and lose it slowly; and they are a direct link in the exposure of predaceous benthivore species. Biological invasions of estuaries are increasing worldwide. Changes in ecological structure and function are well known in response to invasions. This study shows that changes in processes such as cycling and effects of contaminants can accompany such invasions.

Analytical intercomparison results from the 1990 Intergovernmental Oceanographic Commission open-ocean baseline survey for trace metals: Atlantic Ocean

“Dissolved” (< 0.4 μm filtered) and “total dissolvable” (unfiltered) trace element samples were collected using “clean” sampling techniques from four vertical profiles in the eastern Atlantic Ocean on the first IOC Trace Metals Baseline expedition. The analytical results obtained by 9 participating laboratories for Mn, Fe, Co, Ni, Cu, Zn, Cd, Pb, and Se on samples from station 4 in the northeast Atlantic have been evaluated with respect to accuracy and precision (intercomparability). The data variability among the reporting laboratories was expressed as 2 × SD for a given element and depth, and was comparable to the 95% confidence interval reported for the NASS seawater reference standards (representing analytical variability only). The discrepancies between reporting laboratories appear to be due to inaccuracies in standardization (analytical calibration), blank correction, and/or extraction efficiency corrections.Several of the sampling bottles used at this station were not adequately pre-cleaned (anomalous Pb results). The sample filtration process did not appear to have been a source of contamination for either dissolved or particulate trace elements. The trace metal profiles agree in general with previously reported profiles from the Atlantic Ocean. We conclude that the sampling and analytical methods we have employed for this effort, while still in need of improvement, are sufficient for obtaining accurate concentration data on most trace metals in the major water masses of the oceans, and to enable some evaluation of the biogeochemical cycling of the metals.

In situ measurements of calcium carbonate dissolution rates in deep-sea sediments

Benthic fluxes of alkalinity, carbon dioxide, and oxygen have been measured using an in situ incubation chamber at three sites (MANOP Sites C and S and PACFLUX Site SC) in the central equatorial north Pacific. At two carbonate-rich sites (C and SC), a budget for oxygen, alkalinity, and TCO2 fluxes indicate a net CaCO3 dissolution rate of approximately 0.4 mmol m−2 day−1. This rate is only 20% of previous estimates but is consistent with a dissolution rate constant predicted from laboratory experiments with deep-sea sediments and derived from in situ flux measurements in sediments of the southern California borderland. Organic matter oxidation in the sediment column provides the acid for 60–100% of the calcium carbonate dissolution occurring at these sites with the remainder derived from undersaturated bottom water. At a low-carbonate site (S), no carbonate dissolution in the sediment or on the sea floor is apparent, although sediment traps indicate a rain of CaCO3 through 3400 m. The rates of remineralization relative to resuspension or erosion at this site must differ for organic carbon and calcium carbonate, so that carbonate grains reaching the sea floor are physically removed by erosion or resuspension before they dissolve, while organic carbon is largely oxidized before it can be physically removed.