[Home]   [Technology]  [Services]   [Company Info] [News]




Micro to Fine-scale Chemical Gradients and Layers in Stratified Coastal Waters

By Alfred K. Hanson, Jr. and Percy L. Donaghay

Graduate School of Oceanography, University of Rhode Island, Narragansett, RI, 02882, USA

Models for biogeochemical dynamics and chemical variability in coastal waters have traditionally relied upon the assumption that small-scale mixing processes are sufficiently strong and isotropic to rapidly disperse centimeter to meter scale chemical gradients and thin layered vertical structure. Although this assumption of homogeneity may be reasonable in tidally well-mixed estuaries, or within the well-mixed surface and bottom layers of some stratified systems, it may not be generally applicable to stratified coastal waters (Donaghay et al., 1992). A number of naturally occurring physical, biological and chemical processes have the potential to influence the development, maintenance and dissipation of micro (cm-m) to fine-scale (1-10 m) chemical structure in stratified coastal waters. Since micro and fine-scale structure in temperature, salinity and density distributions result from small-scale physical mixing processes (Osborn, 1997), it is plausible that the vertical distributions of dissolved chemicals should have similarly scaled variability. Biogeochemical activity focused within thin plankton layers (Donaghay et al., 1992; Cowles and Desiderio, 1993; Cowles et al, 1997; Donaghay et al, 1997) should also cause, and be influenced by, similarly scaled gradients in chemical concentration and reactivity. This Although this has clearly been demonstrated for methane (Sieburth and Donaghay, 1993), mercury (Mason, et al, 1993) and iron (O'Sullivan, et al, 1997) in the oxic-anoxic transition zone of the lower basin of the Pettaquamscutt Estuary. However, field observations of micro to fine-scale chemical variability associated with physical micro and fine-structure and/or thin plankton layers have been limited in more open systems both by the difficulties in sampling at these scales and by the assumption that such structures cannot persist., however, because the assumption of homogeneity has often influenced our water sampling and analytical chemical protocols.

Sensitive in situ chemical profiling techniques, with the same spatial and temporal resolution as in situ physical and bio-optical profiling techniques, are required to evaluate the influence of small-scale physical mixing and biological processes on chemical distributions, reaction rates and transport. We have developed an electronic profiling and sampling system (Fig. 1a) with the capability for simultaneously sampling at centimeter scales the highly-resolved (cm - m scale) characterization and monitoring of a variety of physical, biological, chemical, and bio-optical structural properties of the water column. The in situ analytical chemical capabilities of the profiling system include electrochemical sensors for dissolved oxygen and pH and a sensitive profiling chemical analyzer (PCA; Fig. 1b) that can measure configured for dissolved nitrite and iron(II) down to sub-nanomolar concentration levels.measurements. Herein, we will first consider the rational for expecting micro to fine-scale structure in nitrite and iron( II), briefly describe our technical approach, to resolve micro to fine-scale chemical gradients and then use high resolution profiles collected present some results from a field investigation conducted in East Sound, WA to illustrate the importance of such measurements to understanding chemical dynamics and impacts in stratified coastal systems.

Fig. 1. Components of (a) the High-Resolution Electronic Profiling and Sampling System and (b) the Profiling Chemical Analyzer configured for dissolved nitrite and iron(II) measurements.

Fig1.jpg (64573 bytes)

Rationale for Nitrite and Iron(II) Fine-structure

High-resolution measurements of dissolved nitrite and iron(II) were selected for this investigation because it was thought that their biological and chemical reactivity would lead to fine-scale chemical structure in stratified coastal waters. Nitrite and iron( II) were also chosen because the spectrophometric determination required a limited number of reagents and had rapid color development rates, simplifying adaptation for in situ vertical profiling.

Nitrite, a dynamic intermediate in the marine nitrogen cycle, is generally a trace constituent (<50 nM) in euphotic surface waters because it is both assimilated by phytoplankton and photochemically reactive (Zafiriou and True, 1979; Grasshoff and Koroleff, 1983). The accumulation of nitrite (>100 - 2000 nM) in mid-depth and deeper coastal waters is thought to be controlled by the microbiological processes of nitrification (i.e., ammonia and nitrite oxidation) and denitrification (i.e., nitrate reduction), occurring within the water column and in surficial sediments (Seitzinger et al., 1980; Lipschultz et al., 1986).

Iron plays an essential role as a micronutrient in phytoplankton photosynthesis (Geider and La Roche, 1994). There is evidence that iron limitation may selectively influence the growth of some species of phytoplankton in coastal waters (Brand, 1991; Sunda and Huntsman, 1995). However, the complexities of the marine chemistry of iron have hindered the elucidation of its micronutrient role (Donaghay et al., 1991; Wells et al., 1995). Most of the iron in oxygenated coastal marine waters is present as insoluble hydrous iron(III) oxide colloids and larger particles, forms of iron that are not readily bioavailable. Dissolved iron(III) is usually present in trace (nanomolar) amounts. Dissolved iron(II) is a more soluble form of iron that is produced by microbial or chemical reduction within oxygen-depleted waters and in suboxic sedimentary porewaters, or by photochemical reduction within the euphotic zone (O'Sullivan et al., 1991; 1997). The rate of iron(II) oxidation back to the less soluble iron(III) form is dependent upon the oxidation rate constant and reactant concentrations (Millero et al., 1987). The redox cycling of iron between less available forms of iron(III) and the more soluble forms of iron(II) results in elevated transient concentrations of reactive forms of iron(II) and iron(III), and can enhance the availability of iron for uptake by phytoplankton (Wells et al., 1995).

The location and intensity of the expected fine-structure should vary between nitrite and iron(II) because of differences in their transient nature (i.e., oxidation or consumption half-lives), sources and sinks, and sensitivities to other factors that vary on fine-scales. For example, the relatively slow rates of microbial production and loss of nitrite (Vaccaro, 1965) should lead to weak gradients except in areas where nitrogen assimilation by intense layers of phytoplankton rapidly generate nutriclines. In contrast, the relatively rapid rates and chemical control of iron(II) oxidation Such effects should be particularly strong for iron(II), because its oxidation rate is dynamically controlled by ambient pH and oxygen levels, temperature, and salinity, should lead to vertical iron(II) structure being highly sensitive to recent input events and to the fine structure in these physico-chemical factors. Testing these hypotheses presented a major analytical challenge since it required; measuring the concentrations of nitrite and iron(II) in situ, at centimeter scales over the 3-4 orders of magnitude range of concentrations (nM to 然) expected in coastal waters. In addition, it  also required simultaneously measuring chlorophyll, temperature, salinity, oxygen and pH at the same scales in order to evaluate the influence of these gradients on the chemical structure, reaction rates and fluxes, reaction rates and biological productivity.

The Profiling Chemical Analyzer

The PCA is comprised of a continuous-flow, reagent-injection module connected to a dual-channel, nine wavelength, submersible spectral absorption meter (Fig. 1b). A WET Lab's model ac-9 spectral absorption and attenuation meter (Moore et al., 1992) was customized to contain twin absorption channels (denoted as model aa-9). The dual-channel reagent addition module was sized (i.e., pump, tubing sizes, flowrates, etc.) to match the geometry of the optical flow cells (25 cm pathlength, 1.2 cm id, ~25 ml void volume) of the aa-9 detector. Seawater and reagents were drawn continuously through each channel of the PCA at(flowrate = 250 ml/min per channel) by a gear pump located down-stream. A 0.22 痠 pore size in-line filter with high-filtration surface area (Gelman maxi-capsule)0.22 痠 pore size) was fixed horizontally, at the seawater inlet to the system. The continuous flow injection of Rreagents (flowrate range 1-5 ml/min), stored in 1.5 liter plastic bags were continuously injected at constant flowrates (3-5 ml/min) into the flowing seawater. The flow rates were, was controlled hydrodynamically by means of adjustable restricting orifices (Dial-A-Flo, Abbott Laboratories). After flow injection, the reagents and seawater were homogeneously mixed by passage through in-line static mixing tubes. . After mixing,After mixing, the reagents and analytes continued continued to react for 30 s during passage through the reaction coils (~30 s residence time) before passing into the WET Labs aa-9 flowcells (25 cm pathlength) for detection.

Dissolved nitrite (NO2-) and iron(II) (Fe+2) were determined in situ spectrophotometrically using classical color-forming reagents: N-1 naphthylethylene - diamine and sulphanilamide for nitrite (Grasshoff and Koroleff, 1993) and Ferrozine reagent for iron(II) (Stookey, 1970). Differential absorption (units = m-1) measurements were utilized to determine nitrite and iron(II) using 540 nm and 560 nm as the respective analytical wavelengths and 650 nm as the reference wavelength. The details of the PCA operation, analytical calibration and intercomparisons with laboratory autoanalyser results on preserved water samples will be presented elsewhere. The improved optics, stable electronics, and longer pathlength of the WET Labs' model aa-9 resulted in the PCA being 10-100 times more sensitive, than conventional laboratory autoanalyzers and spectrophotometers. The limits of detection for nitrite and iron(II) were 0.5 nM and 0.3 nM, respectively (estimated from 4 times the standard deviation of replicate blank readings). The fast data acquisition rate (5.8 Hz) of the WET Lab's aa-9 allowed greater than 5000 concentration readings to be acquired during a 15 minute cast to 25 m depth (2 readings/cm). The PCA had the sensitivity and resolution to detect chemical gradients, at nanomolar to micromolar concentration levels, that occurred within a depth range of 10-20 centimeters with an instrument package drop-rate of 1-2 m/min.

The PCA was deployed as a component of a high-resolution electronic profiling system (Fig. 1a) that included 1) a Sea-Bird Electronics, Inc. CTD911plus CTD with fast response thermister, ducted flow cell, centimeter-scale resolving Digiquartz pressure sensor, and electrodes for+ with pH, redox potential, and dissolved oxygen electrodes;, 2) a WET Labs, Inc. WETStar miniature chlorophyll fluorometer;, 3) a Biospherical Instruments model PUV500 spectral UV and PAR radiometer;, 4) a WET Labs model ac-9 spectral absorption and attenuation meter;, and 5) pump-to-surface and micro-gradient sampling systems. A data acquisition, time stamping and display system, the WET Labs, Inc. Modular Ocean Data Acquisition and Power System (MODAPS) was used to, provide DC power for the instrumentation, acquire and time stamp data from the in situ instruments, then transmit the data to the surface where it was archived in a common file and displayed in real time.

Evidence for Fine-scale Chemical Structure in East Sound

Fig2.gif (8383 bytes)Fig. 2. Station locations and bathymetry of the coastal fjord study site, East Sound, WA.

East Sound is a shallow fjord-like basin penetrating Orcas Island, located within the San Juan Archipelago of western Washington. The basin (Fig. 2) is about 13 km long, 1-2 km wide and has a fairly uniform depth of about 30 m. A partial sill rises 10-15 m above surrounding bottom depths and extends about half-way across the southern entrance to East Sound. The physical dynamics of the primarily wind-driven circulation of East Sound are complex (Rattray, 1967). The stratified waters of East Sound, like many other seasonally-stratified coastal environments, can have its vertical structure strongly modified by episodic wind events -that can cause upwelling, downwelling, lateral advection and/or mixed layer deepening. We have observed that such intermittent wind events can result in altered nutrient availability and lead to intense phytoplankton blooms. These blooms may be dispersed throughout the surface mixed layer and/or focused into thin layers within the pycnocline.

Vertical profiles, acquired at Station A during flood tide in the late-afternoon, revealed considerable micro to fine-scale structure and regions of strong gradients for temperature, salinity, density, chlorophyll a, dissolved oxygen, nitrite, iron(II), and pH and some of the co-acquired physical and bio-optical parameters (Fig. 3). The physical structure of the water column was typified by a 2 m thick weakly stratified surface layer, a sharp pycnocline extending from ~2 to 4 m, a more gradual increase in density (with fine-scale step-structure) through the mid-depth waters and a well-mixed bottom layer (density gradients of less than 0.001 sigma theta m-1). The vertical phytoplankton structure chlorophyll fluorescence distribution was dominated by an intense thin chlorophyll a layer positioned within the pycnocline near 3 m depth. This thin plankton layer had a maximum thickness of about one meter and was comprised of several peaks of 10 to 40 cm thickness. The estimated chlorophyll contents ranged from background levels of 3 痢 l-1 to peak values greater than 20 痢 l-1.

Fig. 3. Micro to fine-scale chemical gradients associated with similarly scaled physical structures and thin chlorophyll fluorescence layers during flood tide at Station A.

Fig3.jpg (75313 bytes)

Dissolved Oxygen, pH and Nitrite Gradients

The enhanced photosynthetic activity associated with the thin plankton layer caused the Oxygen and pH levels to be higher and nitrite levels lower in surface than in deeper waters (Fig. 3). Near surface levels of oxygen and pH were 340 然 and 8.51, respectively. Subsurface microscale maxima in dissolved oxygen (>340 - 381 然) were located at the same depths as maxima in chlorophyll a. The supersaturated oxygen and elevated pH values at the depth of the thin chlorophyll a layer indicate that the layer was a region of intense photosynthetic activity. Oxygen and pH levels decreased sharply below the thin plankton layer (265 然 O2, pH=8.15 at 5 m depth), and then continued to decrease gradually towards near bottom values (146 然 O2, pH=7.74). The nitrite profile was characterized by low surface concentrations (~10 nM nitrite), a steep gradient to higher levels just below the pycnocline (220 nM NO2- at 5 m) and a more gradual increase to 400 nM NO2- near the bottom. The variations in the dissolved oxygen, pH and nitrite profiles were highly correlated (linear correlation coefficients for profiles in Fig. 3b: O2-pH: r2 = 0.974, O2-NO2-: r2 = 0.960 and pH-NO2-: r2 = 0.976). This high degree of correspondence is attributed to the chemical stoichiometry of the photosynthetic production and organic matter remineralization reactions.

In an effort to quantify the nitrite gradient and its relationship to the thin chlorophyll a layer, we used the centimeter-scale nitrite data to calculate the nitrite concentration gradient as d[NO2-]/dz. As indicated in Fig. 4b, the nitrite gradient was extremely steep (d[NO2-]/dz = >300 nM NO2- m-1) at the same depth as the thin chlorophyll a layer. This supports supports the hypothesis that phytoplankton uptake was controlling the position and intensity of the nitrite gradient. It is also consistent with the hypothesis that the vertical position and intensity of the thin phytoplankton layer was being controlled by the position and steepness of the nutricline. Although macronutrient determinations of water samples collected from discrete depths indicated that nitrate and ammonia were extremely low in the surface waters and increased rapidly in the region of the chlorophyll a layer, these discrete samples did not have the spatial resolution needed to accurately estimate the steepness and location of peak gradients. Rigorous testing of the hypothesis would require measuring these other potentially limiting macronutrients on the same centimeter scales as the nitrite.are actively taking up nitrite within the thin plankton layer.

Distinct submeter-scaled chemical gradients were also observed for nitrite, dissolved oxygen, and pH within the mid-depth and deeper waters for this flood tide cast. Two examples are the oxygen and pH maxima and corresponding nitrite minima present near 16.5 and 20 m depth (Fig. 3b). It is unlikely that these structures are the result of locally enhanced primary production because 1) they are not associated with peaks in chlorophyll a, 2) the light levels are low (1-10 mE m-2 s-1), and 3) micromolar concentration levels of nitrate ( 7-15 然) and ammonia (4-6 然) were found at these depths. It is possible that fine-scale maxima and minima were produced within the water column by locally elevated rates of microbial remineralization of dissolved and particulate organic matter. An alternative explanation is that the corresponding mid-water oxygen and pH minima, and nitrite maxima are chemical signals that originated near the sediment-water interface (via microbial remineralization) and were horizontally transported by physical processes (i.e. isopycnal transport) to these depths at station A. Although not readily discernible at the density scale shown in Fig. 3a, these mid-depth waters had characteristic density step-structures, thin (cm-scale) "sheets", with stronger gradients, separated by thicker (m-scale) "layers" with weaker or negligible gradients (Osborn and Cox, 1972; Gregg, 1975). The observed chemical fine-structure appears to be systematically related to these micro-scale and fine-scale physical discontinuities that Osborn (1997) attributes to localized vertical mixing and lateral intrusions, respectively. These observations are also consistent with the results of dye studies conducted in the Baltic Sea (Kullenberg 1974; 1983). A slow rate of vertical mixing, relative to the rate of horizontal transport, may allow the chemical fine-structure to propagate along isopycnals and persist in the stratified waters.

Fig. 4. Positioning of the thin chlorophyll fluorescence layer within the steep nitricline and the agreement between the fine-scale vertical distributions of AOU , nitrite and model-predicted nitrite for the flood tide cast shown in Fig. 3.

                            Fig4.jpg (55063 bytes)

Prediction of Nitrite Fine Structure : If either of the remineralization explanations discussed above are correct, then the classic "Redfield model" predicts a very strong correlation in subsurface waters between nitrite concentrations and apparent oxygen utilization (AOU= theoretical O2 solubility - observed O2 concentration) (Redfield, 1942). This correspondence is very evident when one compares vertical profiles of AOU and PCA-nitrite (Fig. 4). To make sure these relationships were not the result of some peculiarity of a particular cast, we collected water samples from discrete depths on two casts (10 hr separation), determined nitrite concentration levels in the laboratory using standard autoanalysis methods, and then linearly regressed these data against the AOU value for the appropriate collection depth. The resulting regression, NO2- = 1.61(AOU) + 164.2; r2 = 0.912, was then used to predict the continuous PCA-nitrite profile for other profiler casts. Application of this linear relationship to the AOU values calculated from CTD-oxygen data from an earlier profiler cast (Fig. 3b) allowed prediction of the approximate vertical position and slope of the primary nitricline and much of the fine-scale variability in the continuous nitrite profiles, that were obtained with the PCA during the earlier cast (compare predicted- NO2- and PCA- NO2- profiles in Fig. 4; Predicted- NO2- = 0.998(PCA- NO2-) - 3.18, r2 = 0.974). Such a predictive capability is a testament to the sensitivity and accuracy of the PCA instrumentation for continuous, high-resolution measurements of dissolved nitrite in coastal waters. Perhaps more importantly, these results indicate that we should expect fine structure in nitrite profiles anytime lateral transport or locally enhanced microbial activity lead to fine-scale structure in the AOU.

Observed and Predicted Iron(II) Fine Structure

The vertical structure of iron(II) was quite different at station A from the nitrite structure discussed above. Dissolved iron(II) was near the detection limit of the PCA (<0.3 nM) in the upper water column, then increased  rapidly at about 20 m to anomalously higher concentrations (up to 440 nM) of iron(II)  in the near bottom waters. The resulting micronutrient gradient was extremely steep (>20 m depth; d[Fe(II)]/dz = >500 nM Fe(II) m-1; see Fig. 5). In contrast to the tight correlation of the nitrite gradient with the thin chlorophyll layer in the upper water column, this steep gradient in dissolved iron(II) was unrelated to changes in chlorophyll fluorescence. Instead, it corresponded to the top of the well-mixed bottom boundary layer, suggesting that the elevated bottom water iron(II) levels might be the result of resuspension and mixing by bottom currents during this flood-tide cast. We made no attempt to collect samples for comparative laboratory analyses of dissolved iron(II) during this study because iron(II) oxidation rates were expected to be so fast that significant losses of iron(II) would occur during sample collection and preservation (see below for a discussion of iron(II) oxidation rates). However, replicate samples were collected from discrete depths by the pump-to-deck and microgradient sampling systems and preserved by acidification for subsequent analytical determinations of dissolved and total acid reducible iron (O’Sullivan et al., 1997). Taken together, the in situ and laboratory results indicated that there was a substantial reservoir of iron in East Sound (ranges: 10-440 nM dissolved Fe(II+III) and 50-2929 nM total Fe(II+III)), with the highest levels in the near bottom waters.

Fig. 5. Agreement between the steep iron(II) gradients detected in the bottom waters and model predictions of the vertical variability in the pseudo first-order rate constants and half-lives for iron(II) oxidation, and the iron(II) oxidation rate for the flood tide cast shown in Fig. 3.

                         Fig5.jpg (60493 bytes)

Distinct submeter scaled chemical gradients were also observed for dissolved oxygen, pH, nitrite and iron(II) within the mid-depth and deeper waters for this flood tide cast. Two examples are the oxygen and pH maxima and corresponding nitrite minima present near 16.5 and 20 m depth (Fig. 3b). Although not readily discernible at the density scale shown in Fig. 3a, these mid-depth waters also had characteristic density step-structures, thin (cm-scale) "sheets", with stronger gradients, separated by thicker (m-scale) "layers" with weaker or negligible gradients (Osborn and Cox, 1972; Gregg, 1975). The observed chemical fine-structure appears to be systematically related to these micro-scale and fine-scale physical discontinuities that Osborn (1997) attributes to localized vertical mixing and lateral intrusions, respectively. These observations are also consistent with the results of dye studies conducted in the Baltic Sea (Kullenberg 1974; 1983) and suggest that the corresponding mid-water chemical maxima and minima for nitrite, iron(II), oxygen and pH are chemical signals that originate near the sediment-water interface (via microbiological remineralization of organic matter) and are horizontally propagated by localized mixing and lateral intrusion. A slow rate of vertical mixing, relative to the rate of horizontal transport, may allow the chemical fine-structure to propagate and persist in the stratified waters. Biogeochemical activity focused within thin plankton layers or in surficial sediments can cause fluxes of chemicals (i.e. nutrient, CO2, O2, and iron(II) uptake and release), which are similarly influenced by these physical stirring mechanisms, adding to the complexity of the micro to fine-scale chemical distributions observed in the water column.

Predicted and Observed Chemical Fine-structure

The strong biogeochemical coupling between oxygen and nitrite in East Sound allowed us to apply a ‘Redfield Model’ to test the validity of the observed nitrite fine-structure. The variations in the dissolved oxygen, pH and nitrite profiles were highly correlated (linear correlation coefficients for profiles in Fig. 3b: O2-pH: r2 = 0.974, O2-NO2-: r2 = 0.960 and pH-NO2-: r2 = 0.976). This high degree of correspondence is attributed to the biogeochemical stoichiometry between the photosynthetic and organic matter degradation reactions that have occurred over time in the waters sampled. The apparent production or utilization of oxygen should be related to the nutrient changes which co-occur with these processes (Redfield, 1942). The linear relationship between the apparent oxygen utilization (AOU = theoretical O2 solubility - observed O2 concentration) and dissolved nitrite for the discrete-depth water samples, collected from two separate pump-to-surface casts and independently analyzed by laboratory autoanalysis, was Lab-NO2- = 1.61(AOU) + 164.2; r2 = 0.912. Application of this linear relationship to the CTD-O2 data from an earlier profiler cast (Fig. 3b) allowed accurate prediction of the vertical position and slope of the primary nitricline and much of the fine-scale variability in the continuous nitrite profiles that were obtained with the PCA during the earlier cast (compare predicted- NO2- and PCA- NO2- profiles in Fig. 4; Predicted- NO2- = 0.998(PCA- NO2-) - 3.18, r2 = 0.974). Such a predictive capability is a testament to the sensitivity and accuracy of the PCA instrumentation for continuous, high-resolution measurements of dissolved nitrite in coastal waters.

In the introduction, we hypothesized that the rapid rates and chemical control of iron(II) oxidation by ambient pH, oxygen, temperature, and salinity levels should lead to vertical iron(II) structure being highly sensitivity to recent input events and to fine structure in the physico-chemical factors that control the oxidation rate. If theThe detection by the PCA of variably elevated levels of PCA-iron(II) detected in near bottom waters was the result of a physical mixing event providing a was evidence that benthic remineralization was supplying a considerable flux of iron(II) into the overlying waters, then tThe iron(II) released would eventually be oxidized, transported and accumulated as colloidal and particulate iron oxides within the mid-depth and deeper waters. The kinetics of the oxidation of iron(II) have been well studied in the laboratory, as a function of pH, dissolved oxygen, temperature and salinity (Millero et al., 1987). The pseudo-first-order rate constant, k1, for the rate equation -d[Fe(II)]/dt = k1 [Fe(II)], encompasses the effects of these four parameters and is a first degree function of the oxygen concentration and second degree function of the pH. The effects of salinity (expressed as ionic strength: I = 0.0199S) and temperature (oK) on the overall rate constant, k , for the rate equation -d[Fe(II)]/dt = k [Fe(II)] [O2] [OH-]2 can be estimated from the relationship log k = 21.56 -1545/T - 3.29I(1/2) + 1.52I. These algorithms conveniently allow the estimation of k and k1 from field data. The simultaneous measurement of all five parameters (S, T, pH, O2 and Fe(II)) in East Sound provided us with a unique opportunity to test the real-world applicability of this kinetic model for iron(II) oxidation, developed by Millero and co-workers, and an independent check on the validity of the Fe(II) fine-structure detected by the PCA.

The continuous salinity, temperature, oxygen and pH(NBS) data from Fig. 3 were used to estimate the vertical variability for the pseudo-first-order rate constant (k1) for iron(II) oxidation in the East Sound water column (Fig 5). Potential effects due to the complexation of iron(II) by organic ligands or the oxidation of iron(II) by hydrogen peroxide (Millero et al., 1987; Millero and Sotolongo, 1989) were ignored for these approximations. The estimated pseudo-first-order rate constant decreased systematically with depth by over two orders of magnitude for this cast. The calculated half-livestimes for iron(II) oxidation (t(1/2) = ln 2.k1-1) ranged from about 1 minute within the well oxygenated surface waters to over 200 minutes near the benthic boundary. If iron(II) was introduced (e.g. pore water fluxes, sediment-water interactions or photochemical reduction) into waters characterized by slower oxidation rates (longer oxidation half-lives), then the iron(II) signal would persist for a longer time period than in the surface waters. The longer half-live estimates for iron(II) oxidation in the near bottom waters are the combined result of the cooler temperatures, reduced pH values and lower dissolved oxygen contents. These oxidation rate estimates also suggest that there was micro to fine-scale variability in the iron(II) oxidation potential of the stratified waters of East Sound.

Comparison of the half-lives for iron(II) oxidation with the PCA-iron(II) distribution showed a remarkable degree of correspondence (Fig. 5). Higher concentrations of dissolved iron(II) were only detected in waters with slower iron(II) oxidation rates. The estimated oxidation rate for iron(II), -d[Fe(II)]/dt, was greatest (1.1 to 1.4 nM min-1) within the iron(II) containing waters that comprised the benthic boundary layer. The estimated oxidation rate for iron(II) increased very sharply near the top of the benthic boundary layer (~20 m) and associated steep iron(II) gradient (d[Fe(II)]/dz =~300 nM/m)), indicating that iron(II) was more effectively oxidized within this region, than in the near bottom waters containing more iron(II). Since the most likely source of the high iron(II) in the bottom waters is from sediment pore-water fluxes and tidal resuspension of sediments, it should not be surprising that another region characterized by a steep iron(II) gradient is near the sediment interface. The iron(II) fine-structure and strong gradients, detected by the PCA at this station, are predicted to be present by the underlying iron(II) oxidation kinetics.

Spatial and Temporal Variation in Fine-scale Structure

There was considerable spatial and temporal variability in chemical compositions, predicted reaction rates and fine-scale gradients in the waters of East Sound. This was particularly evident for a late afternoon profiler deployment at Station B (Fig. 6) that was characterized by a broad mid-depth minima (12-22 m) in dissolved oxygen and pH and corresponding maxima in dissolved nitrite and iron(II). The PCA-iron(II) levels increased from sub-nanomolar surface values to about 7 nM within the mid-depth waters with lower oxygen contents, then declined rapidly to sub-nanomolar levels in deep waters. In contrast to Station A, the increases in iron(II) at Station B cannot be attributed to iron(II) rich bottom waters being recently mixed upward by tidal flows. The rapid decline in iron(II) below 20 m corresponds to a rapid increase in oxygen and pH at that depth. The calculated rate constants for iron(II) oxidation indicated that the lower pH and oxygen contents within the mid-water minimum extended the half-lives for iron(II) oxidation considerably (to >800 min.). This not only suggests that this feature should persist for some time, but that it could be derived from an interaction with the benthos (or other source) that occurred several days earlier, and then was advected as a mid-water layer to this location. In addition to the steep gradients and broad iron(II) maximum, several submeter-scale maxima were also evident. As at Station A, the calculated half-live for iron(II) oxidation shows submeter-scale maxima that precisely correspond to Fe(II) concentration maxima. Again, the fine-scale structure and sharp gradients are qualitatively consistent with the underlying iron oxidation kinetics.

Fig. 6. Fine-scale vertical structure of physical, bio-optical and chemical properties associated with a broad mid-depth oxygen minimum present during low tide at station B.

                        Fig6.jpg (63385 bytes)

Biological Control of Iron(II) GradientsFig7.jpg (37142 bytes)

The chlorophyll fluorescence, oxygen and PCA-iron(II) profiles, from a late afternoon deployment at Station C (Fig. 7), demonstrate that thin plankton layers (located near 1.5, 6 and 13 m depth) can exert a similarly scaled influence on dissolved oxygen levels, and both the estimated oxidation rate and the trace concentration levels of dissolved iron(II). The fine-scale minima for estimated iron(II) oxidation half-lives can be attributed to focused photosynthetic carbon dioxide uptake (higher pH) and oxygen evolution within the thin phytoplankton layers. The corresponding minima in iron(II) concentrations may be due to locally enhanced rates for iron(II) oxidation and/or direct bioassimilation of dissolved iron(II). Conversely, the mid-water (9-11 m) maximum in iron(II) corresponds both to a maximum in the expected half-live for iron(II) oxidation and to a minimum in phytoplankton chlorophyll and thus potentially lower direct bioassimilation. Iron redox cycling and phytoplankton growth appear to be closely coupled in East Sound.

Fig 7. Association of thin chlorophyll fluorescence layers with micro to fine-scale gradients of oxygen, nitrite, iron(II), and model-predicted half-lives for iron(II) oxidation during flood tide at Station C.

All the iron(II) fine-structure detected by the PCA in the upper portions of the East Sound water column was below the detection limit (<20 nM iron(II)) of the sampling and analytical methodology used by Murray and Gill (1978) in their earlier investigation of the geochemistry of iron in Puget Sound waters. These observations underscore the importance of being able to determine the concentrations of reactive chemicals at the trace levels that influence chemical and biological rates, and with sufficient resolution to define the gradients as they really occur in nature.


The continuous vertical profiles obtained with the high-resolution electronic profiling system and the profiling chemical analyzer in East Sound demonstrate that fine-scale chemical gradients and chemically-distinct thin layers can exist in stratified coastal environments. The fine-scale gradients detected for nitrite were quantitatively related to similarly scaled gradients in apparent oxygen utilization. The fine-scale structure detected for iron(II) was consistent with the similarly scaled structure predicted for the pseudo-first order rate constant for iron(II) oxidation. The mechanisms for the micro to fine-scale gradients observed in the water column included 1) chemical uptake or release (i.e. NO2, Fe(II), CO2, O2) that is focused within thin phytoplankton layers and 2) the interaction of physical mixing and transport processes with biogeochemical fluxes from the sediment-water interface.

It is evident from the East Sound results that the vertical depth resolution of the PCA, for changes in dissolved nitrite and iron(II) concentrations, was approaching the depth-resolution of the electrochemical pH and oxygen sensors that were co-deployed with the electronic profiling system. Whereas many electrochemical sensors and some submersible chemical analyzers may have the required response times and sensitivity to detect micro to fine-scale chemical gradients in stratified coastal waters, they generally have not been applied for that purpose (Johnson et al., 1986a; 1990; 1992; Taylor et al., 1993; Johnson and Jannasch, 1994; Klinkhammer, 1994). The faster package descent rates that have commonly been used during coastal and offshore investigations with CTD's and submersible chemical analyzers (10-40 m min-1 versus the 1-2 m min-1 used in this study) have typically precluded the detection of submeter-scale variability in the water column by the submersible chemical instrumentation.

The enhanced sensitivity of the PCA for nitrite and iron(II) allowed us to document the chemical fine-structure and the presence of transient iron(II) within the water column that has passed undetected by traditional water sampling and analytical methodologies. For example, all the iron(II) fine-structure detected by the PCA in the upper portions of the East Sound water column was below the detection limit (<20 nM iron(II)) of the sampling and analytical methodology used by Murray and Gill (1978) in their earlier investigation of the geochemistry of iron in Puget Sound waters. This underscores the importance of being able to determine the concentrations of reactive chemicals at the trace levels that influence chemical and biological rates, and with sufficient resolution to define the gradients as they really occur in nature. With the high-resolution electronic profiling and PCA technologies we now have the capability to resolve chemical gradients at the small spatial and temporal scales required to investigate the relative importance of chemical processes in the creation and maintenance of thin biologically productive layers and conversely, the influence of such layers on chemical distributions, fluxes and reaction rates.

Future Directions

There are important chemistry-related questions that need to be addressed concerning the dynamics of thin plankton layers. What mechanistic roles do nutrient recycling and availability play in thin plankton layer dynamics? Are episodic variations in nutrient gradients, driven by and coupled to dynamic meteorological and physical processes, critically important to the development, persistence and behavior of thin plankton layers? How significant are the influences of thin plankton layers on the distributions of inorganic and organic chemicals and on their biologically-mediated reaction rates within coastal waters? Our present difficulty in adequately answering these questions illustrates the sparseness of our chemical data on the thin plankton layers phenomena. This short-coming is partly due to the limited availability of high-resolution submersible chemical sensors and analyzers that can interface with contemporary CTD and bio-optical profiling systems. The development of a high-resolution, multi-nutrient profiling analyzer, would allow us to more adequately characterize the nutrient supply to high-biomass plankton layers and provide the parameterization of the nutrient field that is required to develop and test predictive models for thin plankton layer behavior in coastal waters.


Funding for this research was provided by the Office of Naval Research grants N-00014-94-1-0561 to AKH and N-00014-95-1-0225 to PLD. The technical assistance of Robert Griffin with PCA calibration, iron and nutrient determinations and data processing is gratefully acknowledged. We thank Casey Moore (WET Labs, Inc.), Ron Zaneveld, Jennifer Prentice, James Sullivan, and Mike Twardowski for their assistance in this work. We also thank Tim Cowles, Frank Millero and anonymous reviewers for their valuable comments.


Brand, L.E. 1991. Minimum iron requirement of marine phytoplankton and the implications for the biogeochemical control of new production. Limnology and Oceanography, 36: 1756-1771.

Cowles, T. J. and R. A. Desiderio, 1993. Resolution of biological micro-structure through in situ fluorescence emission spectra. Oceanography 6(3): 105-111.

Cowles, T.J. et al., 1997. this issue

Donaghay, P.L., P.S. Liss, R.A. Duce, D.R. Kester, A.K. Hanson, T. Villareal, N.W. Tindale and D.J. Gifford. 1991. The role of episodic atmospheric nutrient inputs in the chemical and biological dynamics of oceanic ecosystems. Oceanography. 4(2): 62-70.

Donaghay, P. L., H. M. Rines and J. M. Sieburth, 1992. Simultaneous sampling of fine-scale biological, chemical and physical structure in stratified waters. Arch. Hydrobiol. 36:1-14.

Donaghay 1997, this issue

Geider, R.J. and J. La Roche, 1994. The role of iron in phytoplankton photosynthesis, and the potential for iron-limitation of primary productivity in the sea. Photosynthesis Research. 39:275-301.

Grasshoff, K., and F. Koroleff, 1983. Determination of Nutrients, In (K. Grasshoff, M. Ehrhardt and K. Kremling eds.) Methods of Seawater Analysis, Verlag Chemie, 125-187.

Gregg, M.C., 1975. Oceanic fine and micro-structure. Reviews of Geophysics and Space Physics, 13(3): 586-591.

Johnson, K.S., C.L. Beehler, and C.M. Sakamoto-Arnold. 1986a. A submersible flow analysis system. Analytica Chimica Acta. 179:245-257.

Johnson, K.S., Sakamoto-Arnold, C.M. and Beehler, C.L. 1990. Continuous determination of nitrate concentrations in situ.. Deep-Sea Research, 36:1407-1413.

Johnson, K.S., Coale, K. and Jannasch, H.W. 1992. Analytical Chemistry in Oceanography. Analytical Chemistry. 64:1065A-1075A.

Johnson, K.S. and H.W. Jannasch 1994. Analytical Chemistry Under the Sea Surface: Monitoring Ocean Chemistry In Situ.. Naval Research Reviews. 3: 4-12.

Klinkhammer, G.P. 1994. Fiber optics spectrometers for insitu measurements in the oceans: the ZAPS Probe. Marine Chemistry, 47:13-20.

Kullenberg, G. 1974. Investigation of small-scale vertical mixing in relation to the temperature structure in stably stratified waters. Advances in Geophysics 18: 339-351.

Kullenberg, G. 1983. Physical Processes. In: Pollutant Transfer and Transport in the Sea. G. Kullenberg (Ed.). CRC Press, Boca Raton, Florida. pp. 2-89.

Lipschultz, F., S.C. Wofsy and L.E. Fox. 1986. Nitrogen metabolism of the eutrophic Delaware River ecosystem. Limnology and Oceanography, 31:701-716.

Mason, R.P., W.F. Fitzgerald, J. Hurley, A.K. Hanson, Jr., P.L. Donaghay and J.M. Sieburth. 1993. Mercury biogeochemical cycling in a stratified estuary. Limnology and Oceanography. 38(6): 1227-1241.

Millero, F.J., S. Sotolongo and M. Izaguirre. 1987. The oxidation kinetics of Fe(II) in seawater. Geochimica et Cosmochimica Acta 51:793-801

Millero, F.J. and S. Sotolongo. 1989. The oxidation of Fe(II) with H2O2 in seawater. Geochimica et Cosmochimica Acta 53:1867-1873.

Moore, C. , Zaneveld, J. R. V. and J. C. Kitchen. 1992. Preliminary results from an in situ spectral absorption meter. In: Ocean Optics XI, G. D. Gilbert, (Ed.), Proc. SPIE 1750, 330-337.

Murray, J.W. and G. Gill, 1978. The geochemistry of iron in Puget Sound. Geochimica et Cosmochimica Acta. 42:9-99.

Osborn, T.R. 1997. This issue

Osborn, T.R. and C.S. Cox, 1972. Oceanic Fine-structure. Geophysical Fluid Dynamics. 3:321-345.

O'Sullivan, D.W., A.K. Hanson, Jr., W. Miller and D.R. Kester. 1991. Measurement of Fe(II) in Equatorial Pacific Surface Seawater. Limnology and Oceanography, 36(8):1727-1741.

O'Sullivan, D.W., A.K. Hanson, Jr. and D.R. Kester, 1997. The distribution and redox chemistry of iron in the Pettaquamscutt Estuary. Estuarine and Coastal Shelf Science, In Press.

Rattray, M., Jr. 1967. Some aspects of the dynamics of circulation in fjords. In: Lauff, G.H. (ed.) Estuaries, AAAS, Washington, D.C., p. 52-62.

Redfield. 1942. The processes determining the concentration of oxygen, phosphate and other organic derivatives within the depths of the Atlantic Ocean. Papers in Physical Oceanography and Meteorology, 9(2):1-22.

Seitzinger, S.P., S. W. Nixon and M.E.Q. Pilson, 1980. Denitrification and N2O production in nearshore marine sediments. Geochimica et Cosmochimica Acta, 44:1853-1860.

Sieburth, J.M. and P.L. Donaghay , 1993. Planktonic Methane Production and Oxidation Within the Algal Maximum of the Pycnocline: Seasonal Fine-scale Observations in an Anoxic Estuarine Basin. Marine Ecology Progress Series. 100: 3-15.

Stookey, L.L. 1970. Ferrozine - A new spectrophotometric reagent for iron. Analytical Chemistry 42(7) 779-781.

Sunda, W.G. and S. Huntsman. 1995. Iron uptake and growth limitation in oceanic and coastal phytoplankton. Marine Chemistry 50:189-206.

Taylor, C.D., B.L. Howes and K.W. Doherty, 1993. Automated instrumentation for time-series measurement of primary production and nutrient status in production-platform accessible environments. MTS Journal, 27(2): 32-44.

Vaccaro, R.F. 1965. Inorganic nitrogen in seawater. In Chemical Oceanography, Vol.1. J.P. Riley and G. Skirrow (eds.) Academic Press, London, 365-408.

Wells, M.L., N.M. Price and K.W. Bruland. 1995. Iron chemistry in seawater and its relationship to phytoplankton: a workshop report. Marine Chemistry. 48: 157-182.

Zafiriou, O.C and M.B. True. 1979. Nitrite photolysis in seawater by sunlight. Marine Chemistry, 8:9-13.


        [Home]     [Technology]    [Services]   [Company Info]  [News]

    Information: Hanson@subchem.com
    Copyright 1998 SubChem Sensor Systems, Inc.
     Last modified: May 25, 2016