Iron chemistry of Hawaiian rainforest soil solution: Biogeochemical implications of multiple Fe redox cycles

Iron (Fe)-oxides are important sorbents for nutrients, pollutants and natural organic matter (NOM). When flucutations in soil oxygen status exist, Fe can cycle through reduced and oxidized forms and thus greatly affect the aqueous conc. of nutrients and metals. We are examining the influence of oscillating oxic/anoxic conditions on Fe-oxide formation and biogeochemical processes (microbial community composition, and carbon, nutrient and trace metal availability). Our work makes use of a natural rainfall gradient ranging from 2.2 to 4.2 m mean annual precipitation (MAP) on the island of Maui, Hawaii, USA. All sites developed on a 400ky basaltic lava flow and comprise soils under similar vegetation. Solid phase Fe concentration and oxidation state vary systematically across this rainfall gradient with a sharp decrease in pedogenic Fe between 2.8 m and 3.5 m MAP that corresponds with an Eh of 330 mV (1-yr ave.). Fe isotopic composition and Fe-oxide associated rare earth elements (REE) also suggest a shift from ligand-promoted to redutive Fe dissolution with increasing rainfall. To examine the effects of multiple Fe oxidation/reduction cycles, we constructed a set of redox-stat reactors that maintain Eh values within a set range by small Eh-triggered additions of oxygen. Triplicate soil slurry reactors are subjected to redox (Eh) oscillations such that Fe is repeatedly cycled from oxidized to reduced forms. During our current experiment, we measure pH and Eh dynamics and monitor the distribution of Fe(II) and Fe(III), major ion and anion concentrations, a range of trace metals including the REE, and total organic carbon (TOC) in three Stokes-effective particle size fractions (<0.45 mm, <0.1 mm, and <0.02 mm) by cascade centrifugation and a <3000 MW fraction isolated via ultra-filtration. Each sample is then sequentially extracted in dilute (0.5 M) HCl and acid-ammonium oxalate. Concurrently, CO2 release is measured and DNA fingerprinting is used to track changes in the microbial community. Prior to implementing the rigorous sampling procedure above, we completed two preliminary reactor experiments focusing only on Fe distribution between aqueous, HCl, and oxalate extractions. These experiments illustrated (1) a distinct threshold for Fe oxidation at ~ 350 mV in the soils (pH 5) and (2) multiple redox cycles increased the HCl-extractable Fe(III) fraction relative to initial conditions. Unexpectedly, this increase occurred predominantly during reducing cycles-perhaps indicating a weakening of Fe-oxide structures during initiation of reducing conditions or oxidation of Fe(II) by NO3. By integrating Fe analysis with trace metal and microbial characterization in triplicate reactors, we will verify this increase in HCl-extractable Fe(III), and assess the impacts of Fe redox oscillation on biogeochemical processes.