Iron mineral dissolution releases iron and associated organic carbon during permafrost thaw

It has been shown that reactive soil minerals, specifically iron(III) (oxyhydr)oxides, can trap organic carbon in soils overlying intact permafrost, and may limit carbon mobilization and degradation as it is observed in other environments. However, the use of iron(III)-bearing minerals as terminal electron acceptors in permafrost environments, and thus their stability and capacity to prevent carbon mobilization during permafrost thaw, is poorly understood. We have followed the dynamic interactions between iron and carbon using a space-for-time approach across a thaw gradient in Abisko (Sweden), where wetlands are expanding rapidly due to permafrost thaw. We show through bulk (selective extractions, EXAFS) and nanoscale analysis (correlative SEM and nanoSIMS) that organic carbon is bound to reactive Fe primarily in the transition between organic and mineral horizons in palsa underlain by intact permafrost (41.8 ± 10.8 mg carbon per g soil, 9.9 to 14.8% of total soil organic carbon). During permafrost thaw, water-logging and O2 limitation lead to reducing conditions and an increase in abundance of Fe(III)-reducing bacteria which favor mineral dissolution and drive mobilization of both iron and carbon along the thaw gradient. By providing a terminal electron acceptor, this rusty carbon sink is effectively destroyed along the thaw gradient and cannot prevent carbon release with thaw.


Supplementary
The green box marks the organic horizon, grey box the transition zone and yellow box the mineral horizon.
Extractable iron was determined via the ferrozine assay. Total organic carbon (TOC) was determined via combustion, whereas the carbon in the dithionite citrate (control corrected) and the control extract (sodium chloride bicarbonate) was determined with the carbon analyzer. Errors of the TOC and 6M hydrochloric acid

Absolute and % values of organic carbon and reactive iron content reported in the main text
Supplementary Table 1. Absolute and % values of iron and carbon in locations Palsa A, Bog C and Fen E, i.e. the cores reported in the main text. In most of the layers, the maximum mass ratio of organic carbon (OC) to iron (Fe) (reactive Fe-associated organic carbon:reactive Fe, OC:Fe (wt:wt)) exceeds 0.22, the maximal sorption capacity of reactive iron oxides for natural organic matter 1,2 . Co-precipitation and/or chelation of organic compounds can generate structures with OC:Fe ratios (wt:wt) above 0.22, as shown in other studies 1,2. Errors of control iron, control carbon, total organic carbon and total extractable iron indicate the range of duplicate analyses of each layer in each thaw stage. Total extractable iron represents the 6M HCl extractable iron (more crystalline Fe phases). Errors of the dithionite/citrate extractable a, iron (reactive Fe, control corrected) and b, carbon (carbon bound to reactive iron, control corrected) represent a combined standard deviation of sodium chloride bicarbonate extractable a, iron and b, carbon, b, citrate blank and dithionite/citrate extractable a, iron and b, carbon (not control corrected).

Fe-associated organic carbon: extraction method and controls
The determination of Fe-associated organic carbon has several well-known difficulties which can only be addressed by combining different approaches.
Considerations for the sodium dithionite-citrate extraction: To prevent hydrolysis of organic matter as well as its protonation and re-adsorption onto sediment particles, which occur under acidic conditions, the sodium dithionite citrate extraction was performed at circumneutral pH (sodium bicarbonate buffered). Therefore, the additional hydroxylamine-HCl extraction (performed below pH 2) can only be a comparison for the sodium dithionite citrate extractable Fe, but not for the sodium dithionite citrate extractable carbon. The control extraction was performed under the same ionic strength (addition of NaCl) and pH (sodium bicarbonate buffered).

(2) Temperature and incubation time
Dithionite citrate bicarbonate extractions have been widely applied in various studies and were previously performed under two different temperatures and incubation times. One is conducted at room temperature at pH 7-8 for 16 hours on a shaker 1,3-6 and the other conducted at 80°C for 15 minutes 7,8 . Due to the high organic carbon content of the soil samples, the standard approach at room temperature at neutral pH for 16 hours was chosen to avoid alteration of carbon during heating of sample to 80°C, which could further influence the amount of extracted iron. The approach conducted at 80°C is suspected to contribute more to nonselective dissolution 9 in organic-rich samples.

(3) Leaching of carbon which is not associated with iron
The carbon measured in the sodium dithionite citrate extraction was corrected by subtraction of the measured DOC values in a citrate blank and in a control extraction, performed under the same ionic strength and pH. This control extraction determines how much carbon would be leached from the soil without any reduction (see Table S1 and Supplementary Figure 7). The effect of a reducing agent, which potentially reductively transforms certain organic functional groups, is not considered in this control (see point (4)).

(4) Dithionite as strong reducing agent
Dithionite is a strong reducing agent which can reductively transform certain organic functional groups and could lead to organic carbon release which is not associated with reactive Fe.
Nevertheless, we consider this to be negligible for our extractions as the concentration of sodium dithionite citrate extractable carbon of a horizon containing primarily organic material and no mineral phase (Palsa A, organic horizon) is very low (0.3±0.1 to 0.4±0.1 mg extractable carbon per g soil; see also Supplementary Table 1). Additionally, a sodium pyrophosphate extraction was performed to determine the colloidal/OM-Fe. The sodium pyrophosphate extraction carbon yielded similar concentrations and showed similar trends to the dithionitecitrate extractable carbon across the thaw gradient (Supplementary Figure 7). Variation between the absolute values can occur due to heterogeneity in the samples and the alkaline conditions of the sodium pyrophosphate extraction (pH 10).

(5) Citrate as strong metal complexing agent
Citrate is a strong metal complexing agent that can influence the amount of extractable iron during dithionite extraction. A test-run was performed with the same experimental conditions (same ionic strength and pH), but no citrate addition. Without citrate, we obtained 64±3% less iron and 57±28% less carbon after sodium dithionite reductive dissolution. We therefore concluded, that the metal ion complexing agent citrate is necessary to avoid under-estimation of iron and organic carbon as a result of complexing or mineral precipitation during extraction.

13
Since extractions have well-known limitations as mentioned above, additional approaches were used and combined to characterize Fe-C associations in the solid phase along the thaw gradient.
Additional approaches used were: (A) Extended X-ray absorption fine structure (EXAFS) with reference for reactive Fe    with stock in mg/cm -2 , bulk density in g cm -3 , content in mg g -1 and actual layer thickness in the core in cm.
For the horizons where compaction was assumed (Supplementary Figure 9), a compaction factor was calculated from the difference between the actual core depth (as reported in the text) and the core hole (compaction factor for palsa organic horizon and palsa transition zone 2.31, for bog organic horizon 5.00 and for fen organic horizon 8.25). The compaction-corrected stock was calculated as follows: with stock in mg/cm -2 , bulk density in g cm -3 , content in mg g -1 and actual layer thickness in the core in cm.

Different Fe analysis of the extracts to rule out matrix effects
Different analytical approaches (ferrozine assay, MP-AES, ICP-MS) have been used to determine Fe in the extracts, to rule out matrix effects and to determine additional elements in the extracts (Supplementary Figure 12). ICP-MS was also used to measure sulphur (S) and phosphorous (P) (Supplementary Figure 8). MP-AES was also used to determine aluminum concentrations in the extracts (Al) (Supplementary Figure 8). For the ferrozine assay, the calibration curves were r 2 > 0.999, and the standard deviations of the triplicate analyses were <1%. For the ICP-MS, the calibration curves were r 2 > 0.999, and the standard deviations of the triplicate analyses were <5%. For the MP-AES, the calibration curves were r 2 > 0.993 and the standard deviations of the triplicate analysis were <10%. We are aware of differences between the iron values. However, because the values only vary slightly, the ferrozine values had a higher accuracy (<1%) and sodium dithionite citrate was not measured with ICP-MS and MP-AES due to citric formation after acidification, we decided to use the data from the ferrozine assay.