Comparative assessment of a restored and natural wetland using 13C-DNA SIP reveals a higher potential for methane production in the restored wetland

Abstract

Wetlands are the largest natural source of methane (CH₄), a potent greenhouse gas produced by methanogens. Methanogenesis rates are controlled by environmental factors such as redox potential, temperature, and carbon and electron acceptor availability and are presumably dependent on the composition of the active methanogen community. We collected intact soil cores from a restored and natural freshwater depressional wetland on Maryland's Delmarva Peninsula (USA) to assess the effects of wetland restoration and redox shifts on microbial processes. Intact soil cores were incubated under either saturated (anoxic) or unsaturated (oxic) conditions and amended with ¹³C-acetate for quantitative stable isotope probing (qSIP) of the 16S rRNA gene. Restored wetland cores supported a distinct community of methanogens compared to natural cores, and acetoclastic methanogens putatively identified in the genus Methanosarcina were among the most abundant taxa in restored anoxic and oxic cores. The active microbial communities in the restored wetland cores were also distinguished by the unique presence of facultatively anaerobic bacteria belonging to the orders Firmicutes and Bacteroidetes. In natural wetland incubations, methanogen populations were not among the most abundant taxa, and these communities were instead distinguished by the unique presence of aerobic bacteria in the phyla Acidobacteria, Actinobacteria, and class Alphaproteobacteria. Iron-reducing bacteria, in the genus Geobacter, were active across all redox conditions in both the restored and the natural cores, except the natural oxic-anoxic condition. These findings suggest an overall higher potential for methanogenesis in the restored wetland site compared to the natural wetland site, even when there is evidence of Fe reduction.IMPORTANCEMethane (CH₄) is a potent greenhouse gas with an atmospheric half-life of ~10 years. Wetlands are the largest natural emitters of CH₄, but CH₄ dynamics are difficult to constrain due to high spatial and temporal variability. In the past, wetlands were drained for agriculture. Now, restoration is an important strategy to increase these ecosystems' potential for sequestering carbon. However, the consequences of wetland restoration on carbon biogeochemistry are under-evaluated, and a thorough assessment of the active microbial community as a driver of biogeochemical changes is needed. Particularly, the effects of seasonal flooding/drying cycles in geographically isolated wetlands might have implications for CH₄ emissions in both natural and restored wetlands. Here, we found that active microbial communities in natural and restored wetlands responded differently to flooding and drying regimes, resulting in differences in CH₄ production potentials. Restored wetlands had a higher potential for CH₄ production compared to natural wetlands. Our results show that controls on CH₄ production in a restored wetland are complex, and dynamics of active microbial communities are linked to seasonal dry-wet cycles.

Keywords: methanogenesis; stable isotope probing; wetland soil microbial ecology.

Fig 1

(Left) Map showing wetland sites located on Delmarva Peninsula in Maryland, USA. (Data from Google, Landsat/Copernicus, SIO, NOAA, U.S. Navy, NGA, GEBCO, LDEO-Columbia, NSF, 14 December 2015–1 January 2021.) (Center) Wetlands within sampling location at wet forest edge (white boxes) highlighted (data from Airbus, 9 November 2023). (Right) Example of transect design for soil core collection. Cores were collected every 0.5 m along five 3 m transects radiating from a randomly chosen location in the forested edge of the wetlands. Twenty-one and 27 cores were collected from the restored and natural sites, respectively.

Fig 2

Schematic showing the differences between the intact soil cores obtained from the natural (left) and restored (right) wetlands. Natural wetland cores were shallower as a deeper organic layer (0–6 cm) on average was removed prior to incubation (green, Oe/Oi) compared to the restored wetland cores. The yellow circles represent the approximate locations of the redox probes: three probes were inserted in the restored wetland cores corresponding to 5.0, 7.5, and 15.0 cm depth below surface (bls). Approximate depths of redox probes for the natural wetland cores were 5 and 10 cm bls; there was no third depth due to the shallow nature of these cores. The equivalent depth range of comparison using DNA–SIP presented in this study is shown by red dotted lines. A photo of the cores placed in incubation design is shown.

Fig 3

Change in redox potential (mV) during the soil core incubations. Data are presented as averages (±standard deviations) for replicate cores (n = 6) in each redox treatment group. Redox readings were collected at three depths within the core, as indicated by color. For all cores, the first redox measurement is reported from the pre-incubations (days −14 or −7), followed by a measurement made at the initiation (day 0) of the incubations. Vertical line at day 0 indicates the beginning of incubation. Midpoint redox measurements were made on day 8 in restored cores, followed by a final reading on day 13. The redox reading for all natural cores is reported for day 12. Natural anoxic and oxic–anoxic cores continued through day 21, when a final redox measurement was made. Data are facetted by wetland and redox conditions on the x axis.

Fig 4

Non-metric multidimensional scaling (NMDS) of the active microbial communities in natural and restored wetland cores under different redox conditions. Points represent the Bray–Curtis distance between active communities in replicate cores. Community data reported as total absolute abundance of each active taxa in each ¹³C-treated core. n = 3 for the restored anoxic, restored oxic, and natural oxic cores, while n = 2 for the natural anoxic and natural oxic–anoxic cores.

Fig 5

Average total absolute abundance of all active taxa grouped by phylum and by genera for methanogens in the phylum Euryarchaeota. Taxa are listed alphabetically by phylum and facetted by kingdom (Archaea and Bacteria). Size of the circle corresponds to average total absolute abundance in the wetland’s different redox treatment communities. Colors indicate wetland type and redox condition. Conditions are listed left to right: natural oxic, natural oxic–anoxic, natural anoxic, restored oxic, and restored anoxic.

Fig 6

Venn diagrams showing the number of active taxa that co-occur or are independent in (a) active communities from the restored or natural wetland; (b) the natural oxic, anoxic, and oxic–anoxic treatments; and (c) the restored oxic and anoxic treatments. Numbers in the Venn diagrams report the number of taxa shared or unique to the groups compared. The accompanying table (d) lists the number of unique taxa over the total number of taxa (no. of unique taxa/no. of total taxa) in each wetland redox treatment group organized by phyla. The restored anoxic active communities had 42 unique taxa and 162 total taxa. The restored oxic active communities had 47 unique taxa and 182 total taxa. The natural anoxic active communities had 48 unique taxa and 166 total taxa. The natural oxic active communities had 22 unique taxa and 98 total taxa. The natural oxic–anoxic active communities had two unique taxa and 58 total taxa.