Subscriber access provided by UNIV OF NEWCASTLE
Environmental Processes
Influence of pO2 on Iron Redox Cycling and Anaerobic Organic Carbon Mineralization in a Humid Tropical Forest Soil Chunmei Chen, Christof Meile, Jared Lee Wilmoth, Diego Barcellos, and Aaron Thompson Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b01368 • Publication Date (Web): 12 Jun 2018 Downloaded from http://pubs.acs.org on June 12, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 31
1
Environmental Science & Technology
To Environ. Sci. & Technol.
2 3 4 5 6
7
8
Influence of pO2 on Iron redox cycling and Anaerobic Organic Carbon Mineralization in a
9
Humid Tropical Forest Soil
10 11 12 13
Chunmei Chen1, Christof Meile2, Jared Wilmoth1, Diego Barcellos1, and Aaron Thompson1*
14 15
1
16
University of Georgia, Crop and Soil Sciences, Athens, Georgia 30602, United States 2
17
University of Georgia, Marine Sciences, Athens, Georgia 30602, United States
18 19 20 21 22
Corresponding author
23
*
Aaron Thompson, AaronT@uga.edu, (01) 706-410-1293
ACS Paragon Plus Environment
Environmental Science & Technology
Abstract
24 25
Ferrous iron (FeII) oxidation is an important pathway for generating reactive FeIII phases
26
in soils, which can affect organic carbon (OC) persistence/decomposition. We explored how pO2
27
concentration influences FeII oxidation rates and FeIII mineral composition, and how this impacts
28
the subsequent FeIII reduction and anaerobic OC mineralization following a transition from oxic
29
to anoxic conditions. We conducted batch soil slurry experiments within a humid tropical forest
30
soil amended with isotopically-labeled 57FeII. The slurries were oxidized with either 21% or 1%
31
pO2 for 9 d and then incubated for 20-d under anoxic conditions. Exposure to 21% pO2 led to
32
faster FeII oxidation rates and greater partitioning of the amended 57Fe into low-crystallinity FeIII-
33
(oxyhydr)oxides (based on Mössbauer analysis) than exposure to 1% pO2. During the subsequent
34
anoxic period, low-crystallinity FeIII-(oxyhydr)oxides were preferentially reduced relative to
35
more crystalline forms with higher net rates of anoxic FeII and CO2 production—which were
36
well correlated—following exposure to 21% pO2 than to 1% pO2. This study illustrates that in
37
redox-dynamic systems, the magnitude of O2 fluctuations can influence the coupled iron and
38
organic carbon cycling in soils and more broadly, that reaction rates during periods of anoxia
39
depend on the characteristics of prior oxidation events.
40 41
TOC Art
42 43 44 45 46
ACS Paragon Plus Environment
Page 2 of 31
Page 3 of 31
Environmental Science & Technology
Introduction
47
Iron (Fe) is an abundant redox-active element in soils 1-3, and conversions between FeII
48 49
and FeIII can be mediated by both microbial and abiotic processes 4. Fe redox cycling is coupled
50
to multiple biogeochemical cycles 5, 6. In natural environments, FeII can be oxidized via biotic or
51
abiotic mechanisms coupled to O2 or NO3- 4, 7, 8, which contributes to the formation of FeIII
52
(oxyhydr)oxides (hereafter referred to as FeIII oxides) of varying size, crystallinity, and purity in
53
soils and sediments 3, 9-14. At circumneutral pH, abiotic oxidation dominates at high O2 levels,
54
while abiotic and biotic oxidation may occur at comparable rates at low O2 levels.15, 16 The FeII
55
oxidation and resultant precipitation of FeIII oxides in environments transitioning from anoxic to
56
oxic conditions is a key process for retaining organic matter (OM), nutrients and contaminants 5,
57
17-19
58
the capacity to adsorb and/or incorporate a variety of constituents19-21. However, when O2
59
becomes depleted in soils and sediments, FeIII oxides serve as important electron acceptors for
60
microbial respiration 22-24, leading to the dissolution and transformation of solid phases and the
61
release of sorbed and incorporated constituents 25-28. In addition, microbial FeIII reduction plays a
62
central role in anaerobic OM mineralization in soils and sediments 23, 29-30. Microbial Fe
63
reduction rates are influenced by the crystallinity of FeIII oxides 31-33, with short-range-ordered
64
(SRO) phases (e.g., ferrihydrite, nano-goethite) favored over more crystalline forms 34, 35.
65
Repeated reductive dissolution and transformation of and precipitation of FeIII oxide minerals
66
can influence the crystallinity of those phases.3, 11, 36, 37Such redox cycling can be caused by
67
periodic rainfall events or fluctuating water tables, which lead to temporal variations in soil O2
68
concentrations.38-42 Thus, it has been suggested that in tropical humid forest soils subject to such
69
conditions and rich in SRO FeIII oxides, microbial FeIII reduction may account for up to 44% of
70
anaerobic organic carbon (OC) mineralization.22
. This is due to the physico-chemical properties of freshly formed FeIII oxides, which have
ACS Paragon Plus Environment
Environmental Science & Technology
The influence of different patterns of O2 fluctuations (magnitude, duration and frequency)
71 72
on Fe redox cycling and associated OC cycling is only beginning to be understood1, 11, 36, 43, 44.
73
Ginn et al.1 showed that repetitive oscillations can lead to an overall increase in Fe reduction
74
rates, which appears to be tied to the repeated formation (and dissolution) of rapidly reducible
75
FeIII phases. During oxic periods, O2 concentrations are expected to control the oxidation rates of
76
aqueous FeII 7 and influence the mineral composition of the resulting FeIII oxides 9, 45, 46. In a
77
synthetic system of pure goethite, γ-Al2O3 and Suwannee River fulvic acid, Chen and Thompson
78
47
79
exposed to low O2 conditions than when exposed to high O2. The degree to which O2 impacts the
80
characteristics of FeIII precipitates depends on the type of mineral surface and OM present 47.
81
However, the availability of the recently precipitated FeIII phases formed at varying O2 levels
82
toward microbial reduction has not been directly evaluated in any system (synthetic or natural)
83
that approximates soil or sediments by containing a mixture of aluminosilicates, Fe/Al oxides or
84
other mineral phases and a variety of organic compounds 48, 49. Considering the range of
85
naturally occurring environments such as wetlands and humid tropical forest that exhibit O2
86
variations of 0-21% at the time scale of days to weeks driven by rainfall or changes in the water
87
table38, 41, 50-52, a comprehensive understanding of the impacts of O2 concentration on the coupled
88
structural and reactivity of FeIII-precipitates formed by FeII oxidation in soils and their impacts
89
on C cycling is needed.
90
have shown that FeII is oxidized at slower rates and results in more crystalline FeIII solids when
In this study, we hypothesized that in redox-fluctuating soils, higher O2 levels will lead to
91
faster FeII oxidation rates and lower crystallinity FeIII solids, as well as greater subsequent
92
microbial FeIII reduction and anaerobic OC mineralization following a transition from oxic to
93
anoxic conditions. To test this, we amended isotopically-labeled 57FeII to soil slurries, oxidized
ACS Paragon Plus Environment
Page 4 of 31
Page 5 of 31
Environmental Science & Technology
94
soil slurries with continous 21% and 1% O2 respectively, and then re-exposed soil slurries to
95
anoxic (N2-headspace) environment to promote Fe reduction. We tracked aqueous- and solid
96
phase-FeII and 57Fe throughout the experiment, measured CO2 production during the anoxic
97
period, and documented the solid-phase speciation of the spiked 57Fe using 57Fe Mössbauer
98
spectroscopy. Materials and Methods
99 100 101
Field site and soil characteristics Soil was collected from an upland valley site in the Bisley Research Watersheds of the
102
Luquillo Experimental Forest, Puerto Rico, a NSF Long-term Ecological Research and Critical
103
Zone Observatory (CZO) site. Soils are classified as Ultisols formed from volcanic parent
104
material with quartz diorite intrusions53, 54. The abundant rainfall and warm temperatures,
105
coupled with the high biological activity typical of humid tropical forests, lead to temporal
106
fluctuations in bulk soil O2 concentrations, which vary from 0% to 21% over timescales of hours
107
to weeks 38, 41. This can stimulate fluctuating FeII concentrations on similar timescales
108
comprising gross Fe reduction and oxidation 55. We collected the upper 10 cm of the soil from
109
the lower valley portion of Bisley Research Watershed site described elsewhere 1, 56, 57. Soils
110
were placed in polypropylene ziplock bags, transported at field temperature to the University of
111
Georgia and air-dried at 20 °C followed by dry sieving (< 2 mm) and homogenizing. The total
112
soil Fe concentration, measured by ICP-MS following Li-metaborate fusion58 (Acme Labs,
113
Vancouver, BC Canada), was 1152 mmol kg-1 soil. Total OC content measured via a Carlo Erba
114
Elemental Analyzer was 4.8% (4000 mmol kg-1 soil). The concentration of short-range-ordered
115
(SRO) FeIII-oxides (an index of the FeIII potentially reducible by microbes) was ~148 mmol Fe
116
kg-1 soil (based on an ascorbic acid/citrate extraction with air-dried soil) 48. XRD revealed quartz
ACS Paragon Plus Environment
Environmental Science & Technology
117
(SiO2) as the major soil mineral, and a kaolinite group phase, a 2:1 layered aluminosilicate
118
(likely chlorite), and goethite present in the soil clay fraction (< 2 mm).48
119
Fe Oxidation and Reduction Experiments
120
In soils, moisture and microbial activity largely control bulk pO2. While bulk soil pO2
121
typically decreases from 21% up to 1% with soil moisture above field capacity52, 55, 59, 60, O2-
122
limited (18 MO-cm) that was deoxygenated by sparging with N2 gas for 2 h and
146
exposed to the glovebox atmosphere (95%/5% N2/H2 mix) for 48 h.
147
All sampling was performed in the anoxic glove box. During oxic conditions, samples
148
were moved temporarily (< 5 min) into the anoxic glove box for sampling. We monitored FeII
149
and Fe isotopic compositions in both the aqueous (dissolved) phase and 0.5 M HCl-extract of
150
soil slurry (See SI section 2 for detailed sampling). FeII in the aqueous sub-samples and the HCl-
151
extracts was analyzed using a modified ferrozine protocol 1, 3, 44, in which the ammonium acetate
152
buffer solution with pH 8.2 was used to accommodate the higher acid content. We selected HCl
153
over organic-acid extractions because HCl has been shown to not introduce any isotopic
154
fractionation of Fe during solubilization63, 64 and preserves FeII from oxidation during the
155
analysis. The 0.5 M HCl extraction includes dissolved FeII in the aqueous phase, desorbs the
156
majority of sorbed FeII and may also dissolve a portion of FeII in minerals such as siderite and
157
magnetite1, 44, 65-67, although siderite formation is unlikely given the acidic pH (6) and a low
158
carbonate concentration in our samples. We did not detect magnetite in our Mössbuaer analysis
159
and therefore we assume the difference between HCl-extractable FeII and aqueous (dissolved)
160
FeII can be attributed to sorbed FeII.
161 162
Aqueous and HCl-extractable Fe isotope compositions were measured by inductively Coupled Plasma Mass Spectrometry (ICP-MS, Perkin Elmer, Elan 9000). Samples analyzed by
ACS Paragon Plus Environment
Environmental Science & Technology
163
ICP-MS were measured in an optimized DRC mode using reactive NH3 gas to minimize
164
potential interferences from Ar, O, and H containing complexes that could convolute the mass-
165
to-charge signal of Fe isotopes. The accuracy of this method was tested with Fe isotopic
166
reference material IRMM-014. Repetitive measurements of isotopic fractions in IRMM-014
167
were stable (4.56±0.11%, 93.12±0.22%, 2.02±0.12%, and 0.29±0.02% for f54Fe, f56Fe, f57Fe and
168
f58Fe, respectively, n = 90). Total CO2 production was measured during the anoxic period. Reactors were crimp-
169 170
capped with gas-tight lids inside the N2 chamber. Headspace gas was sampled immediately after
171
capping and after 1 d for sampling during days 10-26 and after capping for 3 d during days 27-30
172
(the last time point) to measured CO2 concentrations on a gas chromatograph with a thermal
173
conductivity detector (Shimadzu, Columbia, MD, USA). The averaged CO2 production rate
174
during the sampling periods was calculated on a soil mass (oven dry equivalent) basis as the
175
difference of initial and final headspace CO2 mass concentration.
176
57
177
Fe Mössbauer Analysis Use of 57Fe isotopes allows us to track the amended 57Fe using Mössbauer spectroscopy,
178
which detects only 57Fe atoms and not other Fe isotopes. To quantify the partitioning of the
179
amended 57Fe via Mössbauer spectroscopy, we established a FeII-amended treatment with natural
180
Fe isotope abundance prepared in the same exact way as the 57FeII-addition treatment described
181
above, except that the FeII-bearing stock solution contained standard FeCl2·4H2O at natural
182
isotopic abundance (~2.1% of 57Fe). The Mössbauer spectra of the amended 57Fe was then
183
calculated as the difference between the spectra from the 57FeII-enriched treatment minus the
184
baseline spectra obtained when adding FeII with natural isotopic abundance, after taking into
185
account the different total 57Fe concentrations in the 57FeII-enriched treatment and the treatment
ACS Paragon Plus Environment
Page 8 of 31
Page 9 of 31
Environmental Science & Technology
186
adding FeII with natural isotopic abundance (see SI section 3 for further details). Therefore, the
187
resulting Mössbauer spectra of the amended 57Fe excluded the spectral signal from the native soil
188
Fe atoms. To prevent FeII oxidation, solid samples for 57Fe Mössbauer analysis were collected in
189
the anoxic glove box following centrifugation at 20,000 g for 10 min, preserved between two
190
layers of O2-impermeable Kapton tape (this step took 0.9) of FeII concentration over time during the anoxic period (SI
217
Figure S3). Quantitatively, HCl-extractable FeII production rate ranged from 4.8– 7.2 mmol kg-1
218
d-1, with dissolved and sorbed FeII being produced at rates of 1.2–2.2 and 4.0–5.3 mmol kg-1 d-1,
219
respectively (SI Table S1). 57FeII-amended soils displayed greater HCl-extractable FeII
220
production and thus higher FeIII reduction rates (5.7 – 7.2 mmol kg-1 d-1) than the control
221
treatments containing only soil (~ 4.8 mmol kg-1 d-1) (SI Table S1). Oxidation of the control
222
treatment (initially oxic soil with no FeII addition) with 21% and 1% O2 showed no difference in
223
FeII production (and therefore FeIII reduction) during the subsequent anoxic period. In contrast,
224
oxidation of 57FeII -amended soils with 21% O2 resulted in a greater FeIII reduction rate following
225
a transition from oxic to anoxic period, compared to oxidation with 1% O2.
226
57
227
Fe dynamics during the oxic-anoxic cycle The dissolved 57Fe/total Fe (FeT) ratio in the aqueous phase dropped from an initial value
228
of 0.97 to 0.55 (first data point in Figure 2c) over the initial anoxic 1-day equilibration,
229
indicating that a large proportion of the amended 57Fe underwent rapid atom exchange with
230
native Fe in the soil prior to the oxidation event. Over the first day of oxidation with 21% or 1%
231
O2, 57Fe fell below detection in the aqueous phase (Figure 2a) and remained undetectable until
ACS Paragon Plus Environment
Page 10 of 31
Page 11 of 31
Environmental Science & Technology
232
the subsequent anoxic period when the aqueous 57Fe/FeT ratio increased up to ~0.4, presumably
233
due to reductive dissolution of 57FeIII formed either via oxidation/ precipitation of 57FeII or
234
reductive dissolution of 57FeIII formed via atom exchange between 57FeII and native soil FeIII.
235
Relative to oxidation with 1% O2, oxidation with 21% O2 resulted in greater solubilization of
236
57
237
higher FeII production rate. However, the aqueous 57Fe/FeT ratio during the anoxic period was
238
similar for both of the 57Fe-amended soil treatments (21% and 1% O2, Figure 2c).
239
CO2 production during the anoxic period
240
Fluxes of CO2 during the 20-d anoxic period varied over time, but showed similar temporal
241
trends across all treatments (Figure 3a). CO2 production peaked at day 16 (after 6 days of anoxic
242
incubation) and declined afterwards. Generally, more CO2 was produced during the anoxic
243
incubation from the 57FeII-amended soils than from the soil-only treatments (Figure 3a and 3b).
244
While oxidation of the control (soil-only) treatment with 21 and 1 % O2 showed no differences in
245
anoxic CO2 production, oxidation of the 57FeII-amended soils with 21% O2 led to more anoxic
246
CO2 production than oxidation with 1%-O2 (Figure 3a and 3b). However, higher CO2 production
247
was only evident between day 16 and 26, and from day 26 to 30 the 21%- and 1%-O2 oxidation
248
treatments exhibited no differences in anoxic CO2 production in the 57FeII-amended soils (Figure
249
3a). Overall, the cumulative anoxic CO2 production—which had a similar trend as FeII
250
production across all the treatments—decreased in the order: 57FeII-amended soil with oxidation
251
at 21% O2 > 57FeII-amended soil with oxidation at 1% O2 > the soil-only control treatments. The
252
total CO2 production during the 20-d anoxic period accounted for 4.1-6.2% of total OC in the
253
studied soil. In addition, rates of CO2 production were strongly correlated with FeII production
254
under anoxic conditions (R2 = 0.94, P < 0.0001), with a ratio of produced FeII to CO2 of 5. This
Fe and total FeII during the subsequent anoxic cycle (Figure 2a and 2b), consistent with a
ACS Paragon Plus Environment
Environmental Science & Technology
255
is slightly higher than the stoichiometric ratio (~4) assuming microbial FeIII reduction is coupled
256
to anaerobic mineralization of OC with a nominal oxidation state (NOSC) of zero30. It is possible
257
that Fe-reducing microbes utilized OC with negative NOSC values68 or some FeII was generated
258
by fermentative organisms that carry out incomplete C mineralization. Solid-phase partitioning
259
of the added 57FeII following the initial 1-day anoxic period Collection of Mössbauer spectra at 140K, 77K, 12K, and 5K allows us to characterize the
260 261
crystallinity-continuum of the FeIII-oxide solid phases in the sample. As temperature is reduced,
262
portions of the FeIII-oxide populations are resolved as they magnetically order into a sextet
263
(Figure 4 and SI Figure S2-4), with the more crystalline portions of the population ordering at
264
higher temperatures10. In the initial unreacted soil sample, the spectral area assigned to the full
265
FeIII-oxide sextets was 30.7%, 44.0%, 58.6%, 65.4% at 140K, 77K, 12K, and 5K, respectively
266
(SI Figure S5 and Table S2). At 5K, we resolved the following Fe components in the initial
267
unreacted soil: ~38% SRO- (e.g., nano-)goethite, ~27% ferrihydrite, ~14% most-disordered FeIII
268
oxides (remain unordered at 5K), ~15% clay/OM-FeIII, as well as minor clay- FeII and ilmenite
269
(