High-amplitude redox fluctuations prime tropical forest soils for rapid iron reduction rates

Background/Question/Methods

The fate and bioavailability of carbon and phosphorus in soils undergoing frequent shifts in redox status is tightly coupled to iron cycling. We examined the influence of dynamic redox conditions on soils from the Bisley Site of the Luquillo Critical Zone Observatory in Puerto Rico. These soils contained ~0.14 g kg^-1 P, ~2 g kg^-1 C, and ~62 g kg^-1 Fe that resides predominately in nano-crystalline Fe^III oxyhydroxide phases. We conducted eight-week incubations of triplicate soil slurries subjected to 0 - 21% O₂ redox oscillations with a ratio of time under oxic to anoxic conditions of 1:6 at three frequencies (3.5-d, 7-d, and 14-d) with and without the addition of phosphate. We tracked Fe reduction rates, P availability and Fe mineral composition changes over the course of the experiment and incorporated our data into a numerical model. In addition, we probed the susceptibility of soil Fe phases toward Fe^II-facilitated atom exchange using isotopically-labeled Fe²⁺(aq).

Results/Conclusions

We found the rate of Fe^II reduction (measured by HCl-extractable Fe^II) increased from ~ 3 mmol Fe^II kg^-1 soil d^-1 in the first reducing cycles to > 45 mmol Fe^II kg^-1 soil d^-1 in the later reducing cycles with a concomitant increase in the Fe^II concentration plateaus from 10 to 180 mmol Fe^II kg^-1 soil. This apparent maximum in reducible Fe coincides with the amount of Fe extractable by citrate-ascorbate and the portion of Fe^III that magnetically orders between 77K and 13K in the Mössbauer spectra. The P amended treatments approached this maximum Fe^II concentration at earlier oscillation cycles than the unamended treatments, but by 56 days all treatments exhibited similar Fe^II oscillation amplitudes. We found that aqueous Fe atoms can be exchanged with both the labile (0.5 M HCl-extractable) and bulk (7M HCl-extractable) Fe pools, with turnover times on the order of hours and months, respectively. Synthesis of this experimental data within our preliminary numerical model suggests temporal dynamics of Fe(II) can be explained by minor increases in the population of Fe reducers accompanied by progressive dissolution of recalcitrant Fe(III) solid phases. In addition, it suggests the Fe reduction rates during the onset of anoxia are strongly coupled to recent physiochemical dynamics. Past performance may dictate future results.