Environ. Sci. Technol. 2002, 36, 460-467
Phytoplankton-Mediated Redox Cycle of Iron in the Epilimnion of Lake Kinneret Y E A L A S H A K E D , † Y I G A L E R E L , * ,† A N D ASSAF SUKENIK‡ Institute of Earth Sciences, The Hebrew University of Jerusalem, Givat Ram, Jerusalem 91904, Israel and The Yigal Allon Kinneret Limnological Laboratory, Israel Oceanographic & Limnological Research Ltd., P.O Box 345, Tiberias 14102, Israel
The biological-mediated redox cycle of Fe was studied in the epilimnion of Lake Kinneret (Sea of Galilee), a mesotrophic lake in Israel. Multi-annual lake water sampling and incubation experiments were carried out to study Fe(III) reduction by natural phytoplankton populations and their possible role in inhibiting Fe(II) oxidation. The reduction characteristics of the dinoflagellate Peridinium gatunense, the dominant lake alga, were further examined in the laboratory. The steady-state concentration of Fe(II) calculated from the assessed reduction and oxidation rates was compared with Fe(II) measured in the lake in order to evaluate the significance of these processes to the lake Fe redox cycle. Nanomolar concentrations of Fe(II) were measured in the oxygenated, high pH, upper water layer of the lake throughout the year. Reduction rates of Fe by natural phytoplankton assemblages ranged between 0.1 and 10 nM/h. The highest reduction rates, determined in dinoflagellate-dominated lake waters, coincided with the highest concentrations of Fe(II) measured simultaneously in the lake. Iron(II) oxidation rates calculated from the measured lake Fe(II) and the obtained reduction rates were significantly slower than published abiotic Fe(II) oxidation rates. Indeed, Fe(II) oxidation rates measured in algalenriched lake water were 30-fold slower than Fe(II) oxidation rates in natural water, demonstrating the potential for Fe(II) stabilization by the lake phytoplankton.
Introduction Iron is the most common essential trace metal for phytoplankton growth in aquatic ecosystems. As opposed to its high biological demand and its great abundance in the earth’s crust, Fe availability to phytoplankton in oxic waters is very low. This is because Fe(III), the thermodynamically stable form of Fe, forms highly insoluble oxides or hydroxides that are not directly available for biological uptake (1). Furthermore, the bioavailability of dissolved Fe (which is very scarce) is determined by the nature and stability of its complexes (2). Iron(II), which is more soluble than Fe(III), is thermodynamically unstable in most natural surface aquatic systems. Several studies have recently described the dynamic redox cycle of Fe in natural waters (3-9). These redox transforma* Corresponding author phone: +972-2-658-6515; fax: +972-2566-2581; e-mail:
[email protected]. † The Hebrew University of Jerusalem. ‡ Israel Oceanographic & Limnological Research Ltd. 460
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 3, 2002
tions have important implications for Fe bioavailability in both seawater and freshwater environments. To attain measurable Fe(II) concentrations in oxygenated waters, Fe(III) reduction rates should be enhanced, and/or Fe(II) oxidation rates should decrease. Significant Fe(II) concentrations in low pH, low temperature, and highly saline environments (such as acidic lakes, atmospheric waters, the cold Southern Ocean, and the Dead Sea) were attributed to slower Fe(II) oxidation rates (3, 10-13). Similarly, a substantial decrease in Fe(II) oxidation rate was imputed to complexation of Fe(II) by natural organic ligands or sulfides (13-15). In addition, elevated Fe(II) concentrations in atmospheric and surface waters were ascribed to enhanced photoreduction rates (4-7, 11). Photoreduction of Fe(III) occurs directly through photolysis of Fe(III) complexes (16, 17) or indirectly by photoproduction of reductants (18). For example, high Fe(II) concentrations observed during phytoplankton bloom in bay waters were explained by enhanced Fe(III) photoreduction by dissolved organic substances (5). The ability of different phytoplankton species to enzymatically reduce Fe(III) at the cell surface has been long recognized (19-21), but only lately its ubiquity in different taxa and its importance for Fe uptake became clear (22-24). In a previous study, high Fe(II) concentrations were detected in the surface waters of Lake Kinneret throughout the year (9). It was found that Fe(III) photoreduction rates, determined both in the laboratory and in the field, were too slow to maintain the observed Fe(II) concentrations (9). Therefore we hypothesized that lake phytoplankton are the key players responsible for both Fe(II) production and Fe(II) stabilization. Because of its high Fe concentrations and dense algal biomass, Lake Kinneret is well-suited to assess the biomediated Fe redox cycle. This study is one of the first to combine field measurements of Fe(II) with in vitro lake water incubations and controlled experiments of laboratory cultures. To the best of our knowledge, this is the first study that investigates the effect of biologically mediated Fe(III) reduction and retardation of Fe(II) oxidation on Fe redox speciation in oxygenated waters.
Experimental Methods Study Area. Lake Kinneret is a warm monomictic freshwater lake, located in northern Israel, with a surface area of 160 km2 and a total volume of 4 × 109 m3 (25). Average and maximum depths are 24 and 42 m, respectively (25). The lake is mesotrophic to eutrophic with dissolved organic carbon (DOC) of 200-300 µM, alkalinity of 1.6-2.1 mequiv, and ionic strength of 10-15 mM (25). The lake is thermally stratified from March to September, and the depth of the thermocline varies between 15 and 20 m (25). During the stratified period, the epilimnion retains high pH values (8.59.1) and is well oxygenated, but its major nutrient content is relatively low (nitrate ) 0.6-11 µM, total phosphorus ) 0.4-0.9 µM) (25). The hypolimnion, on the other hand, becomes anaerobic, with lower pH values (7.6 ( 0.4) and high concentrations of sulfides, ammonia, Fe(II), and phosphorus (25). Lake Kinneret phytoplankton is dominated by Peridinium gatunense, which blooms during the spring (February-June) (26). This large dinoflagellate (diameter: 35-70 µm) comprises more than 95% of the total phytoplankton biomass during its bloom and over 60% of the annual primary productivity in the lake (27). During the summer and autumn, nanoplanktonic forms of chlorophytes, cyanophytes, diatoms, and dinoflagellates dominate the phytoplankton population. With the winter turnover and runoff, nutrient 10.1021/es010896n CCC: $22.00
2002 American Chemical Society Published on Web 01/04/2002
concentration increases, and species diversity becomes greatest. Large filamentous algae such as the diatom Melosira granulata are common during this period. Clean Laboratory Protocols. All of the ware and reagents used in the laboratory and in the field were carefully cleaned and checked for blanks. Pre- and post-sampling manipulations were conducted in a clean laboratory. Polyethylene bottles for sample collection, Teflon beakers for total digestions, and Teflon filter holders and filters were cleaned by three sequential treatments with distilled HNO3 according to Patterson and Settle (28). All reagents were prepared with double-distilled water, distilled acids, and analytical grade salts. Field Sampling. Lake water was sampled monthly from February 1998 to March 2001 with detailed profiles taken at the deepest part of the lake. All samples, except the surface water, were collected with Niskin bottles. Temperature, pH, and light intensity were recorded simultaneously in situ. Chlorophyll a (Chl a), algal-species composition, and major ions were determined on subsamples. Lake water was filtered through 0.025-µm pore size filters in the laboratory within 1 h of sampling. In addition to thorough cleaning of the filters and the filtration equipment, the first 30 mL of the filtrate was discarded. Iron(II) Determination. Iron(II) concentrations ([Fe(II)]) were determined by the colorimetric reagent Ferrozine (FZ) (Sigma Chemical Co., 562 nm ) 27 900 M-1 cm-1). During 1998, the concentration of Fe(II) in lake water was determined directly by the FZ method (29). However, this method was not sensitive enough for our purposes since it led to an overestimation of [Fe(II)]. Hence in 1999 we used a method that included a preconcentration step (6, 7, 30). Calibration between the two methods shows that [Fe(II)] determined directly was 2.2-fold higher than [Fe(II)] determined with a preconcentration step (R2 ) 0.82, n ) 12). The preconcentration step was adjusted to the lake’s low ionic strength water (10-15 mM) as follows. Upon sampling, 1 L of lake water was fixed on-board with 40 µM FZ. In the laboratory, each FZ-fixed sample was split into three acid-cleaned 200mL Nalgene bottles. One aliquot was then spiked with 10 nM Fe(II) standard. All three aliquots were pumped simultaneously by a peristaltic pump through a 0.2-µm filter and a precleaned C18 Sep-Pak cartridge (30) at a flow rate of 15 mL/min. NaCl solution was mixed with the water sample during pumping to reach a final concentration of 35 g/L NaCl. From each depth, a water sample without FZ was manipulated in the same way and was used as a blank. The molar extinction coefficients of the Fe(II)(FZ)3 complex () and the Fe(II) recovery efficiency for each sample were determined using the standard addition method. All samples were extracted within 4 h of sampling. The Fe(II)(FZ)3 complex was eluted from the Sep-Pak cartridge with 6 mL of methanol and was measured in a 5-cm quartz cell at 562 nm using a Spectronic 20 Genysis spectrophotometer. RSD of the analysis was e15%, and the detection limit was 1 nM. Iron Reduction and Oxidation Experiments. Fe reduction experiments were preformed according to the BPDS method (31). BPDS (bathophenanthroline disulfonic acid, Sigma Chemical Co., 535 nm ) 22 140 M-1 cm-1) is a strong Fe(II) chelating agent that prevents Fe(II) oxidation and uptake by phytoplankton (32). In most experiments, the lake algae were concentrated and were resuspended in duplicates in 50 mL of aged, 0.2-µm filtered lake water. This was done in order to render a stronger signal and to weaken the background noise. The experiment started when 0.2-20 µM FeEDTA (1:2 molar ratio of Fe:EDTA), Fe(NO3)3, or FeCl3 and 200 µM BPDS were added to the samples. BPDS itself is capable of a substantial Fe reduction (32); hence, 0.2-µm filtered lake water, aged for at least 1 yr, was used as a blank and was subtracted from all other treatments. Every hour a subsample
was withdrawn, filtered through a 0.2-µm filter, and measured in a 5-cm quartz cell at 535 nm. Iron(III) reduction rates were computed from the linear increase in Fe(II) concentrations over the first 6 h of the experiment. In 1998, all experimental bottles were incubated in the lake, while in subsequent years incubations were carried out in the laboratory at constant conditions of 20 °C and 30 µmol (quanta s)-1 m-2 provided by cool white fluorescence bulbs. Subsamples were taken to measure pH, Chl a, and algalspecies composition. pH was determined with a WTW 196 pH meter using a senTix-97 combination electrode. In an attempt to mimic the lake conditions, experiments were conducted at the lake pH values (8.3-9.3). In a single experiment, the pH never varied by more than 0.3 unit throughout the duration of the experiment. The iron chelate added to most of the experiments was FeEDTA, since its reduction rate was less effected by the annual lake water pH changes than Fe(NO3)3 or FeCl3. The use of FeEDTA was further supported by our speciation calculation (Hydraql (33)), which showed that, in order to maintain the observed lake water levels of dissolved Fe (10-20 nM), the presence of strong organic chelator is needed (Shaked, unpublished data). Laboratory experiments with Fe-replete and Fe-deplete cells were conducted with cultures of P. gatunense. The cultures were placed in 1-L Erlenmeyer flasks with 8-fold enriched macronutrient L16 medium (34) and maintained at exponential growth in a growth chamber under a 12-h light (30 µmol (quanta s)-1 m-2) and 12-h dark regime at 20 °C. The growth media was adapted to Fe studies by addition of 0.1 mM EDTA (added to buffer [Fe′]). The algae were harvested and resuspended in a fresh 100 mL of medium at pH of 9.0 ( 0.3. EDTA was omitted from the medium when other Fe complexes were tested. In all the experiments, an aliquot was taken for cell count. Another set of experiments with P. gatunense was conducted under identical conditions with K3Fe(CN)6 as an Fe(III) source (Sigma Chemical Co., 420 nm ) 1017 ( 8 M-1 cm-1). The rate of K3Fe(CN)6 reduction was determined from the decrease in its absorbance at 420 nm (20). Experiments were initiated with 10-100 µM fresh K3Fe(CN)6, and every 15 min a subsample was withdrawn, filtered, and measured. Oxidation experiments were conducted in 0.2-µm filtered, unfiltered, and biomass-enriched lake water. P. gatunense culture was supplemented to lake water at 100 µg/L Chl a, which represents its actual density in the lake during a bloom. The 0.1-10 µM Fe(II) standard (ammonium ferrous sulfate, Sigma Chemical Co.) was added to the experiment medium at room temperature under fluorescence light. A subsample was withdrawn every few seconds, fixed with FZ, filtered, and recorded. The same sample without FZ was used as a blank. pH was monitored throughout the experiment. Total Fe and Chlorophyll Determination. Samples for Fe analysis were acidified (0.1 M distilled HNO3) and stored at 4 °C. Prior to analysis, the water samples were digested at 100 °C with 10% distilled HNO3 and were preconcentrated when needed. Total and dissolved Fe were analyzed with a Perkin-Elmer 5100-PC atomic absorption spectrometer, equipped with a 5100 ZL Zeeman graphite furnace module. The RSD of the analysis was less than 10%, and the detection limit was 1 nM. Chl a was measured in a Turner Designs 10400 fluorometer following an extraction of the GFC filters in 90% acetone for 24 h in the dark (35).
Results and Discussion Fe(II) in Lake Kinneret Epilimnion. Total Fe concentrations in Lake Kinneret epilimnion are 50-70 nM during the summer and autumn and 150-300 nM during the winter and spring (Figure 1). Fe(II) accounts at most for 20% of the total Fe but usually ranges between 1 and 8%. The different VOL. 36, NO. 3, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
461
FIGURE 1. Total Fe, Fe(II), and dissolved Fe (
