Mark L. Wells

Iron chemistry in seawater and its relationship to phytoplankton: a workshop report

. In coastal and shelf waters, substantial external inputs of iron come from riverine sources and bottom sediments, leading to markedly higher dissolved and particulate iron concentrations. Much of the particulate iron in nearshore waters is inorganic and processes that solubilize this reservoir, making it accessible to phytoplankton, are especially relevant (see below). The greater iron inputs to coastal and shelf regions compared to the open ocean are accompanied by high iron requirements of neritic phytoplankton species (Brand et al., 1983; Sunda et al., 1991). These systems therefore might be strongly influenced as the comparatively rich iron resource is diminished by phytoplankton blooms. The single largest reservoir of iron in the surface waters of HNLC regions may be the biota itself. New evidence indicates that this biological pool of iron is recycled on the time scale of days, much like N and P (Hutchins et al., 1993). This “input” of regenerated iron to surface waters is estimated to be more than an order of magnitude greater than the external supply rate of iron (Bruland, this meeting; Morel, this meeting) and may largely satisfy the iron-demand of phytoplankton in these systems. This view is supported by recent results of Price et al. (1994) showing that iron uptake rates of plankton in the equatorial Pacific are sufficient to entirely turn over the dissolved iron pool within half a day or less. Presently, there is no indication of the chemical forms of this regenerated iron or whether these forms are directly reassimilated by phytoplankton. B. What are the sinks of iron? The removal of iron from surface waters is fairly well constrained within a geochemical (i.e. mean residence time) perspective, however, the mechanisms and dynamics of this removal is not well understood. Mechanisms for removing iron from surface waters include: * sorption and precipitation, 0 biological assimilation, aggregation of inorganic or organic colloids, and . sinking of mineral and biogenic particles. While much of the particulate iron introduced via rivers, sediment resuspension, or as mineral aerosols will be removed by settling, ascertaining the underlying basis for the removal of “dissolved” iron forms is much more difficult. In regions with a high sinking flux of inorganic mineral particles (e.g. in some coastal and well mixed shelf waters), dissolved iron may be removed abiotically by sorption to surfaces of these particles. Similarly, sinking organic particles also can scavenge soluble iron from surface waters (Morel and Hudson, 1985). Dissolved iron also is “removed” via direct assimilation by phytoplankton. The subsequent sinking of live cells or fecal matter will transport a portion of this biogenic iron from surface waters. In addition to direct assimilation and sorption onto sinking (mineral and biogenic) particles, iron may sorb to colloidal organic matter which is abundant in surface waters (Wells and Goldberg, 1992, 1994). The stability of this colloidal phase is a topic of much dispute (Honeyman and Santschi, 1989; Bauer et al., 1992; Moran and Buesseler, 1992; Wells and Goldberg, 1993), but the extremely large colloidal surface area combined with the particle reactive nature of iron suggests that aggregation of organic colloids could be important for removing iron. Significant unresolved issues regarding iron removal include (1) identifying the specific mechanisms of iron sorption to abiotic and biotic sinking particles, which will shed light on how changes in iron speciation may affect this removal pathway, (2) changes in the iron “export efficiency” of the assimilation pathway with shifts in primary productivity or species assemblage, and (3) the abundance and reactivity of iron in the colloidal reservoir. The relative importance of abiotic, bioM.L. Wells et al.iMarine Chemistry 48 (199.5) 157-182 163 tic and colloid aggregation removal processes will vary from regime to regime and with season. (2) What are the chemical speciation and forms of iron among the soluble, colloidal and particulate fractions, including the rates and mechanisms of transformations among these forms? There are some large gaps in our knowledge of iron chemistry in seawater. The development of trace metal clean techniques for seawater collection and analysis over the past decade (Bruland et al., 1979; Gordon et al., 1982; Landing and Bruland, 1987) has given us reliable profiles for iron in the traditional categories of particulate (> 0.4 pm) and “dissolved” (< 0.4 pm) fractions; however, the chemical forms of iron within these fractions has been largely speculative. For example, there now is evidence that iron exists, at least partially, in the small colloidal phase (Wells and Goldberg, 1991; Wu and Luther, 1994; Powell and Landing, abstract) which is included in operationally defined “dissolved” fractions. Determining how iron is partitioned among these phases, and among various chemical forms within each phase, is central to understanding iron speciation in seawater.

The complexation of 'dissolved' Cu, Zn, Cd and Pb by soluble and colloidal organic matter in Narragansett Bay, RI

It is widely accepted that the speciation of most bioactive metals in seawater is regulated by natural organic ligands, but the nature of these molecules has remained a mystery. We used a combination of physical and chemical separation schemes to better characterize organic molecules complexing Cu, Zn, Cd, and Pb in Narragansett Bay, RI. Conventionally filtered ( 1 kDa), with the bulk of it found in the 1–8 kDa colloidal size range. Of the three Cu-binding ligand classes measured, the strongest class occurred mainly in the soluble fraction while the weaker Cu-binding ligand classes were predominantly colloidal (>1 kDa). Approximately 40% of chelated Pb was colloidal but, in contrast to Cu, the bulk of these ligands resided in the larger colloidal size range (8 kDa–0.2 μm). Thus, a continuum of metal complexing ligand size exists, spanning from truly soluble to colloidal, the nature of which differs for individual bioactive metals. These findings support the hypothesis that metal complexation in seawater is dominated by distinct, metal-specific ligand molecules. A central question that now emerges is whether these ligand molecules function predominantly to buffer metal ion activities in seawater, thereby decreasing metal sorption to particulates, or to facilitate metal removal by sweeping organically bound metals into particulate phases via colloid aggregation.

Colloid aggregation in seawater

Aggregation of small colloids (< 0.2 μm) is common in ocean waters and leads to agglomerates that are several microns in size. These aggregates are the most abundant macroparticles (1.0 μm) in mid-depth and deep waters of the Atlantic and Pacific Oceans, with concentrations in the order of 105 colloid aggregates ml−1. The fractal structures of these aggregates are characteristic of both reaction-limited and diffusion-limited processes, indicating that colloidal dynamics vary widely in seawater. Aggregation can greatly enhance the involvement of marine colloidal matter in biological and sedimentation processes.

Marine submicron particles

Abstract The qualitative and quantitative analyses of submicron particles from coastal surface seawaters off California involved their separation by ultracentrifugation, visual examination by transmission electron microscopy (TEM), and the determination of elemental composition by energy dispersive spectroscopy. Particles fell into three classes: organic, both living and detrital; inorganic, such as iron colloids; clay minerals. The greatest abundance of particles occurred in the smaller than 120 nm size fraction. The mass concentration for particles sized between 5 and 120 nm is estimated to range between 0.03 and 0.09 mg l −1 . Particle numbers in this size range were greater than 10 9 particles ml −1 with the total surface area being 8 m 2 per cubic meter of seawater, or more. Particle size distributions in the smaller than 120 nm fraction showed numbers to increase nearly logarithmically with decreasing size. TEM examination indicates that many of these particles are in fact aggregates of granules 2–5 nm in size.

Algae as nutritional and functional food sources: revisiting our understanding

Global demand for macroalgal and microalgal foods is growing, and algae are increasingly being consumed for functional benefits beyond the traditional considerations of nutrition and health. There is substantial evidence for the health benefits of algal-derived food products, but there remain considerable challenges in quantifying these benefits, as well as possible adverse effects. First, there is a limited understanding of nutritional composition across algal species, geographical regions, and seasons, all of which can substantially affect their dietary value. The second issue is quantifying which fractions of algal foods are bioavailable to humans, and which factors influence how food constituents are released, ranging from food preparation through genetic differentiation in the gut microbiome. Third is understanding how algal nutritional and functional constituents interact in human metabolism. Superimposed considerations are the effects of harvesting, storage, and food processing techniques that can dramatically influence the potential nutritive value of algal-derived foods. We highlight this rapidly advancing area of algal science with a particular focus on the key research required to assess better the health benefits of an alga or algal product. There are rich opportunities for phycologists in this emerging field, requiring exciting new experimental and collaborative approaches.

The phttoconversion of colloidal iron oxyhydroxides in seawater

Abstract The marine photochemistry of colloidal Fe oxyhydroxides was studied in order to determine its role in modifying the availability of Fe for phytoplankton. Using a new technique that measures a labile portion of total Fe in seawater, the lability of synthetic crystalline and non-crystalline colloidal Fe oxyhydroxides was found to increase in pH8 seawater upon irradiation with UV light, and simulated and natural full spectrum sunlight. The increased lability appeared to result from an organic-dependent Fe photoreduction followed by very rapid Fe(II) oxidation and Fe(III) reprecipitation as ferrihydrite. Aging in the dark caused significant reductions in the lability of the photoproduced ferrihydrites, with both synthetic iron oxides and natural iron from sunlit surface seawater. The results suggest that the lability of Fe in surface seawater varies in response to competing processes of photoredox and aging of Fe on oxyhydroxide surfaces. Because the lability of colloidal Fe correlates positively with its availability to marine algae, the photolysis of colloidal Fe has direct implications for phytoplankton.

The distribution of colloidal and particulate bioactive metals in Narragansett Bay, RI

Abstract The bioactive metals Fe, Mn, Cu, Zn and Ni in Narragansett Bay, RI, were partitioned into soluble, colloidal and particulate size fractions using a combination of conventional and cross-flow filtrations. Particulate samples (0.2–8.0 μm; >8 μm) were chemically fractionated into acetic-acid reactive and non-reactive metals. Conventional “dissolved” samples ( Mn>Zn>Cu>Ni with concentrations in the 0.2–0.8 μm fraction being generally higher than in the >8.0 μm fraction. The acid leachable fraction of the particulate phase increased from ∼32%–80% in the order Fe 8.0 μm) being generally less labile than small particulates (0.2–8.0 μm). The colloidal phase represented an average 4%–96% of the “dissolved” metals, ranging in importance from Fe (96%)>Cu (44%)>Ni (25%)>Zn (7%)>Mn (4%). Although generally small, the colloidal fraction of Zn and Mn was highest in a region of the bay where biomass typically is high. Changes in soluble and colloidal fractions along a transect through the bay indicate that a significant proportion of Fe, Cu and Ni were transferred from “dissolved” to particulate size fractions via colloid aggregation. Predicting colloidal metal concentrations from measurements of particulate mass ( C p ) and literature values of colloid metal partition coefficients ( K c ) underestimated the measured concentrations by 5–50×. Acetic acid leachable metal concentrations in the small particle (0.2–8 μm) phase correlated well with metal concentrations in the larger (8 kDa–0.2 μm) colloid fraction ( r =0.91–0.99). In contrast, metals in the smaller colloid fraction (1–8 kDa) were for the most part independent of any measured parameters. Metals were not distributed equally between colloidal size classes; colloidal Zn was associated with larger colloids (>90%), Fe and Ni were associated primarily with larger colloids (∼70–85%) but also with the smaller colloid fraction, while ≳70% of colloidal Cu was associated with smaller colloids. The non-uniform distribution of metals within colloidal size classes indicates that metal:colloid associations are regulated by specific interactions. These findings suggest that it is inappropriate to employ single, non-specific sorbing metal tracers (e.g. Th) to delimit the pathways and kinetics of bioactive metal interactions with marine colloids.

Tectonic processes in Papua New Guinea and past productivity in the eastern equatorial Pacific Ocean

Phytoplankton growth in the eastern equatorial Pacific Ocean today accounts for about half of the ‘new’ production—the fraction of primary production fuelled by externally supplied nutrients—in the global ocean. The recent demonstration that an inadequate supply of iron limits primary production in this region supports earlier speculation that, in the past, fluctuations in the atmospheric deposition of iron-bearing dust may have driven large changes in productivity. But we argue here that only small (∼2 nM) increases in the iron concentration in source waters of the upwelling Equatorial Undercurrent are needed to fuel intense diatom production across the entire eastern equatorial Pacific Ocean. Episodic increases in iron concentrations of this magnitude or larger were probably frequent in the past because a large component of the undercurrent originates in the convergent island-arc region of Papua New Guinea, which has experienced intensive volcanic, erosional and seismic activity over the past 16 million years. Cycles of plankton productivity recorded in eastern equatorial Pacific sediments may therefore reflect the influence of tectonic processes in the Papua New Guinea region superimposed on the effects of global climate forcing.

Marine colloids: A neglected dimension

In the ocean, colloids lie at the boundary between soluble chemical species and sinking particles. They account for 30-50% of the ‘dissolved’ organic carbon in sea water, and so are highly important players in marine chemistry. Work invoking polymer gel theory adds a fresh element to understanding colloid behaviour in sea water by quantifying colloid aggregation rates.

The level of iron enrichment required to initiate diatom blooms in HNLC waters

The chemical speciation of iron in seawater is controlled by complexation with organic ligands, the character of which likely regulates which phytoplankton groups can readily access this resource. Evidence from the IronEx II mesoscale iron enrichment experiment provides some insight to this regulation. Dissolved iron ( 1 kDa–0.4 μm) and soluble (<1 kDa) size fractions using cross flow filtration to better delineate iron dynamics with respect to phytoplankton growth. While dissolved concentrations in the patch increased by two orders of magnitude to ∼2 nM Fe, soluble concentrations only roughly doubled to ∼35 pM Fe, and this change occurred only during the first few days of the experiment. The subsequent decrease in soluble iron to ambient levels coincided with a dramatic increase in chlorophyll a, indicating that biological demand was responsible for the disappearance of soluble iron. This decrease also coincided with preferential drawdown of silicic acid over nitrate in the patch, indicating that diatoms were experiencing iron stress even as the bloom developed. Even so, calculations of the diffusional flux and iron uptake kinetics at the cellular level reveal that the concentrations of soluble iron during the later stages of the bloom were above that required to support continued rapid growth of the long, narrow pennate diatoms. This observation implies that the bulk of the organically bound soluble and colloidal iron was unavailable to meet the diatom growth requirements, and further that the measured increase in concentrations of strong Fe(III) complexing ligands impaired iron acquisition by diatoms. These findings indicate that excess macronutrients in HNLC waters might not be fully converted into diatom biomass even with repeated infusions of iron.