Microbial drivers of methane emissions from unrestored industrial salt ponds

Abstract

Wetlands are important carbon (C) sinks, yet many have been destroyed and converted to other uses over the past few centuries, including industrial salt making. A renewed focus on wetland ecosystem services (e.g., flood control, and habitat) has resulted in numerous restoration efforts whose effect on microbial communities is largely unexplored. We investigated the impact of restoration on microbial community composition, metabolic functional potential, and methane flux by analyzing sediment cores from two unrestored former industrial salt ponds, a restored former industrial salt pond, and a reference wetland. We observed elevated methane emissions from unrestored salt ponds compared to the restored and reference wetlands, which was positively correlated with salinity and sulfate across all samples. 16S rRNA gene amplicon and shotgun metagenomic data revealed that the restored salt pond harbored communities more phylogenetically and functionally similar to the reference wetland than to unrestored ponds. Archaeal methanogenesis genes were positively correlated with methane flux, as were genes encoding enzymes for bacterial methylphosphonate degradation, suggesting methane is generated both from bacterial methylphosphonate degradation and archaeal methanogenesis in these sites. These observations demonstrate that restoration effectively converted industrial salt pond microbial communities back to compositions more similar to reference wetlands and lowered salinities, sulfate concentrations, and methane emissions.

Methanogenesis and Salt Tolerance Genes of a Novel Halophilic Methanosarcinaceae Metagenome-Assembled Genome from a Former Solar Saltern

“…This agrees with experimental data from a sediment core from this pond (but note that this was not a pure culture of Methanosalis sp. SBSPR1A) in which methane emissions increased more, and more rapidly, when trimethylamine was added compared to methanol or acetate [30]. Methanol is a product of pectin degradation [77], which proceeds slowly in lake sediments [77]; perhaps for this reason methanol was only a marginal methanogenic precursor in a previous lake sediment study [78] and may not be as important as trimethylamine in the salterns or wetland studied here.…”

“…The genome was assembled from metagenomic sample R1_A_D2 (IMG ID 3300026157, NCBI BioSample accession SAMN06266236) as part of the study published by Zhou et al [30] on South San Francisco Bay salt ponds, who initially classified the MAG as Methanolobus with Bin Annotation Tool (BAT) [31]. The genome name SBSPR1A is based on the sample from which it is derived (South Bay Salt Pond R1 replicate A).…”

“…Industrial salt production occurred in the pond from the 1850s to 2003. The sample was characterized by high salinity (97.2 total ppt Cl − ), a temperature of 24.7 • C, low dissolved oxygen content (0.96 mg L −1 ) and a pH of 7.75 [30]. With salinity levels over twice that of seawater (~35 ppt), the pond can be considered hypersaline [25].…”

“…With salinity levels over twice that of seawater (~35 ppt), the pond can be considered hypersaline [25]. Elevated methane emissions were measured from this core in the field (737 µmol CH 4 m −2 d −1 ) and in a lab incubation using sediments collected from the same pond [30]. Approximately 10 Gb of shotgun sequence data were generated from this sample on the HiSeq 2500 (Illumina, San Diego, CA, USA) platform in a 2 × 150 paired-end run mode.…”

“…Abundance estimates were calculated by mapping the raw metagenomic reads to the MAG, taking the average read depth, and then transforming to counts per million of assembled reads to take into account differences in library sizes across the 24 metagenome samples [30]. The effect of site (4 separate wetlands) on MAG abundance was calculated with a Kruskal-Wallis test (R package stats) followed by Nemenyi posthoc (R package PMCMR [39]), as assumptions for ANOVA and Tukey HSD were not met (Levene Test, p < 0.05, Shapiro-Wilk test, p < 0.05).…”