Controls on potential iron reduction in soils from diverse ecosystems

Background/Question/Methods

Iron-coupled biogeochemical reactions could be important in many terrestrial ecosystems during episodic low redox events, but there is no general understanding of where high rates of Fe reduction can occur. We hypothesized that soils with larger poorly crystalline Fe pools, with greater soil C availability, and which experience more frequent reducing conditions would exhibit higher potential Fe reduction. We collected surface soil (0-10 cm) from four study sites spanning a range in ecosystem types: humid tropical forest (Luquillo CZO); desert grassland (Sevilleta LTER); and temperate pine forest (Calhoun CZO). We slurried the soils and incubated them in a glovebox with a dinitrogen headspace for 7 days, measuring Fe reduction rates daily. To evaluate the role of C availability in determining potential Fe reduction, we added sodium acetate daily at rates up to 0.55 mg-C g^-1 soil d^-1. To characterize initial soil properties that are possible controls on potential Fe reduction, we extracted total Fe in citrate-dithionite extractions, microbially reducible Fe in 0.5 N HCl, “reducible” poorly crystalline Fe in citrate-ascorbate, and “chelatable” poorly crystalline Fe in acid ammonium oxalate extractions. DI extractions for dissolved organic C (DOC), soil analysis for total organic C, and pH measurements were also performed.

Results/Conclusions

All measures of Fe availability differed among the study sites (F_3,16 > 233, p < 0.001) with the highest Fe consistently in the humid tropical forest soil. In contrast, the order of the other study sites differed by extraction. For example, the desert grassland soil had the lowest total Fe (3.82 ± 0.08 mg-Fe g^-1-soil) but the second highest “chelatable” poorly crystalline Fe (3.20 ± 0.42 mg-Fe g^-1-soil); the temperate pine forest soil had the lowest microbially reducible Fe (0.005 mg-Fe g^-1-soil) but the second highest total Fe (12.1 ± 0.65 mg-Fe g^-1-soil). Labile C addition rates up to 0.1 µg-C g^-1 d^-1 increased Fe reduction in the humid tropical forest soil to as high as 3154 µg-Fe g^-1 d^-1, while higher C addition rates had no effect on or inhibited Fe reduction. Carbon addition did not affect Fe reduction in the other soils. Iron reduction rates in the temperate pine forest were low throughout the incubation and did not exceed 77 µg-Fe g^-1 d^-1. The desert grassland soil exhibited highly variable Fe reduction rates with C addition and anaerobic incubation length, but rates reached as high as 300 µg-Fe g^-1 d^-1. The controls on Fe reduction potential were complex.