Anaerobic microsites have an unaccounted role in soil carbon stabilization

Abstract

Soils represent the largest carbon reservoir within terrestrial ecosystems. The mechanisms controlling the amount of carbon stored and its feedback to the climate system, however, remain poorly resolved. Global carbon models assume that carbon cycling in upland soils is entirely driven by aerobic respiration; the impact of anaerobic microsites prevalent even within well-drained soils is missed within this conception. Here, we show that anaerobic microsites are important regulators of soil carbon persistence, shifting microbial metabolism to less efficient anaerobic respiration, and selectively protecting otherwise bioavailable, reduced organic compounds such as lipids and waxes from decomposition. Further, shifting from anaerobic to aerobic conditions leads to a 10-fold increase in volume-specific mineralization rate, illustrating the sensitivity of anaerobically protected carbon to disturbance. The vulnerability of anaerobically protected carbon to future climate or land use change thus constitutes a yet unrecognized soil carbon-climate feedback that should be incorporated into terrestrial ecosystem models.

Fig. 1

Microsite impacts on metabolic rates in flow reactor experiments. a Particle size-induced variations in the anaerobic pore volume. Oxygen profiles in reactor experiments conducted with soil amended with quartz grains of different particle sizes (coarse = 150–250 µm and fine = 25–45 µm). Oxygen profiles are the average of three replicate profiles recorded with microelectrodes on day 35 of the incubation. b Particle size-induced variations in CO₂ production (i.e., mineralization) over the incubation period. Linear fits indicate the incubation period used to calculate mineralization rates. Error bars denote the standard error of the mean calculated from four replicate reactors

Fig. 2

The effect of particle size-induced variations in the extent of anaerobic microsites on microbial metabolic rates and pathways. a Overall mineralization rate within the reactor, b comparison of the total aerobic and anaerobic respiration rates, c volume-specific aerobic and anaerobic respiration rates, and d contribution of anaerobic respiration pathways to the overall rates. The overall mineralization rates were calculated using linear fits, as shown in Fig. 1a. A combination of anaerobic incubations, solid- and solution-phase measurements, and a mass balance approach was used to quantify aerobic and anaerobic respiration rates across the diffusion-limited domain (a detailed description of our approach is provided in the “Methods” section). Particle size variations to manipulate the extent of anaerobic microsites are shown in red (coarse = 150–250 µm) and blue (fine = 25–45 µm). Error bars denote the standard error of the mean calculated from four replicate reactors

Fig. 3

Bioenergetic projections for the mineralization of organic compounds under aerobic and anaerobic conditions. The thermodynamic driving force, F _T, is given for the oxidation of organic compounds spanning a range of nominal oxidation states when coupled to the reduction of oxygen (aerobic respiration) or Fe(OH)₃ (anaerobic respiration). When coupled to oxygen (gray line), F _T for the oxidation of an organic compound is close to 1. Mineralization (R _min) is thus expected to proceed uninhibited for compounds spanning the full range of oxidation states. Under Fe(OH)₃-reducing conditions (orange line), mineralization of more oxidized compounds, such as sugars, yields projected rates comparable to those under aerobic conditions (60–80%). By contrast, mineralization of reduced organic compounds, such as lipids, would proceed at limited rates or may even be completely inhibited (0–30%), resulting in a relative enrichment, and ultimately preservation, of these abundant compound classes in anaerobic microsites

Fig. 4

The effect of particle size-induced variations in the extent of anaerobic microsites on the microbial oxidation of different carbon pools. a Changes in total carbon across the diffusion-limited domain within the reactor. b Changes in dissolved, particulate, or mineral-associated organic carbon pools in the aerobic and anaerobic zone. Dissolved (DOC), particulate (POC), and mineral-associated organic carbon (MAOC) are listed in order of decreasing bioavailability. c Changes in carbon oxidation state in reactor experiments. The changes in oxidation state are calculated as the ratio of the absorbance of carboxyl (288.35 eV) to that of aromatic C (285.05 eV) in spectra obtained by near-edge X-ray absorption fine structure (NEXAFS) spectroscopy. Changes in carbon pools and oxidation state are shown relative to a time-zero control (t ₀). Particle size variations to manipulate the extent of anaerobic microsites are shown in red (coarse = 150–250 µm) and blue (25–45 µm). Error bars denote the standard error of the mean calculated from four replicate reactors

Fig. 5

Impact of soil particle size distribution, and the associated variations in anaerobic soil volume, on carbon bioavailability and chemistry. a Differences in the dissolved (DOC), particulate (POC), or mineral-associated organic carbon (MAOC) pools between coarse- and fine-textured soil. DOC, POC, and MOAC are shown as the relative differences between the two soils. b Differences in carbon oxidation state between fine- and coarse-textured soil. Carbon oxidation state is calculated as the ratio of the NEXAFS absorbance of carboxyl (288.35 eV) to that of aromatic C (285.05 eV). c Differences in the abundance of reduced carbon functional groups between fine- and coarse-textured soil and their accumulation with depth. Changes in the abundance of NEXAFS-detectable aromatic and aliphatic C in the subsoil (100 cm) are shown relative to subsoil horizons (20 cm) (see the “Methods” section for details on NEXAFS analysis). Error bars denote the standard error of the mean calculated from the analysis of three replicate soil cores

Fig. 6

Nominal oxidation state of carbon in fine- and coarse-textured upland soil varying in the abundance of anaerobic microsites. The boxes enclose the interquartile range, the whiskers indicate minimum and maximum values, horizontal bars dissecting the boxes represent the median values, and the symbols (dot) represent the mean. Carbon in fine-textured soil showed a lower oxidation state (i.e., it consisted of more reduced carbon compounds) than the coarse-textured soil when extracted with water (H₂O) or methanol (MeOH). The average nominal oxidation state of carbon (NOSC) was calculated based on molecular formulae provided for water and methanol extracts analyzed by FT-ICR-MS. Details on the FT-ICR-MS analysis and NOSC calculations can be found in the “Methods” section