Background/Question/Methods We report on a novel pathway in the terrestrial nitrogen (N) cycle that couples anaerobic ammonium oxidation with iron(III) reduction. The pathway, called Feammox, results in the production of dinitrogen (N₂) gas or nitrite (NO₂^-). Nitrite produced via Feammox can be subsequently reduced to nitrous oxide (N₂O) gas, N₂, or ammonium (NH₄⁺) via dissimilatory NO₂^- reduction or abiotic oxidation of iron(II). We conducted three experiments. As a general test of anaerobic NH₄⁺ oxidation we added equimolar ¹⁵NH₄⁺ and natural abundance NO₂^- (¹⁴NO₂^-) to iron (Fe)-rich tropical forest soils. Acetylene (C₂H₂) was added as an inhibitor of N₂O reduction to N₂ and of anammox to a second set of isotopically labeled samples. This method can distinguish direct N₂ production via Feammox or anammox from N₂ produced from denitrification of Feammox-generated NO₂^-. Changes in the ^15-14N₂ mole fraction could result from anammox or Feammox, whereas changes in the ^15-15N₂ mole fraction indicated Feammox. In the presence of C₂H₂, increases in the ¹⁵N enrichment of N₂ suggest direct N₂ production via Feammox. In the second experiment we added NH₄⁺ only or NH₄⁺ with Fe(III) to pre-incubated anaerobic slurries to test for NO₂^- production. We explored the potential for Feammox to produce N₂ in a third experiment. Pre-incubated soil slurries were amended with ¹⁵NH₄⁺ (24 h only) or ¹⁵NH₄⁺ with Fe(III) (6 and 24 h) in stoichiometric equivalency. Dinitrogen was measured immediately after collection on an isotope ratio mass spectrometer.

Results/Conclusions

There was no ^15-14N₂ produced during the 96 h incubation with or without C₂H₂ added. However, we measured considerable ^15-15N₂ production. Acetylene addition (24 h only) increased the ^15-15N₂ mole fraction suggesting that some N₂ was produced directly via Feammox and did not result from anammox or biological denitrification of NO₂^-. In the second experiment, aqueous NO₂^- concentrations were significantly higher in samples that had received NH₄⁺ and Fe(III) than soils that had received only NH₄⁺ after 22 hours. A similar trend occurred at 3 h, but the differences were not statistically significant. In the third experiment, we measured significantly higher ^15-15N₂ mole fraction in the ¹⁵NH₄⁺ and Fe(III) treatment than in the ¹⁵NH₄⁺ only or blank (soil free) treatments. Using mass balance calculations we determined that over 20% of the added NH₄⁺ was lost as N₂ via Feammox. Our results suggest that Feammox occurs in these upland soils, and has the potential to produce significant N₂ and NO₂^-.