Microbial methane cycling in a landfill on a decadal time scale

Landfills are significant sources of methane, contributing an estimated 60–69 Tg CH4 annually between 2000 and 2017, with landfills accounting for up to ~20% of net methane emissions in some high-income countries. Global municipal solid waste (MSW) production is projected to rise from ~2.01 billion tonnes in 2018 to 3.40 billion tonnes by 2050, intensifying the need to understand long-term biogeochemical controls on methane cycling in landfills. MSW degradation proceeds through a five-phase conceptual model from initial aerobic decay (Phase 1) to anaerobic acidogenesis (Phase 2), rapid methanogenesis (Phase 3), slow methanogenesis (Phase 4), and finally partial oxygen intrusion and stabilization (Phase 5). Predicting methane emissions over decadal timescales is challenging due to spatial and temporal heterogeneity in waste composition, moisture, redox conditions, and gas transport. Prior studies have relied heavily on 16S rRNA and functional-gene amplicon surveys to infer methanogen and methanotroph guild dynamics, but these approaches struggle with uncharacterized taxa and often overlook anaerobic or intra-aerobic methane oxidation. This study applies genome-resolved metagenomics across landfill cells spanning the five lifecycle phases to resolve community composition, methanogen and methanotroph diversity, metabolic capacities, and adaptations relevant to methane production and oxidation over nearly four decades of landfill operation.

Multiple 16S rRNA amplicon surveys across landfills have shown that refuse age, environmental parameters (e.g., pH, temperature), nutrient and contaminant concentrations shape microbial communities. Targeted amplicon surveys of methanogenic (mcrA) and methanotrophic (pmoA, mmoX) markers have expanded known diversity of methane-cycling taxa, especially in cover soils, but often reduce functional inference to dichotomies (e.g., hydrogenotrophic vs. acetoclastic; type I vs. type II methanotrophs) that neglect novel or anaerobic pathways (e.g., intra-aerobic and anaerobic methane oxidation). Recent metagenomic studies in landfills have elucidated cellulose and plastic degradation, antibiotic resistance, and broader diversity patterns, but comprehensive, genome-resolved, temporally contextualized surveys of methane-cycling guilds across landfill lifecycle phases are scarce. This study addresses these gaps by integrating phylogenomics and metabolic reconstruction to profile methanogens and methanotrophs, including overlooked anaerobic methane oxidizers and putative novel methanotroph lineages.

Site and sampling: A sanitary landfill in the northeastern United States (anonymized) with leachate collection and biogas capture was sampled in February 2019. Eight leachate samples represented cells of differing ages: A (1980–1982), B (1982–1988), C (1988–1993), D1 (1993–1998), D2 (1995–1998), E (1999–2014), F1 (2014–present), F2 (2016–present). Leachate was collected after purging wells; 1 L was filtered (0.22 µm Sterivex) and filters frozen at −80°C. Historical geochemistry (1983–2019) and flare gas composition/flow (2018–2019) were provided by site owners. Key variables included BOD, COD, pH, ORP, organic acids, bicarbonate, gas volume and composition. Lifecycle phase classifications used these records: A, B (Phase 5); C, D1, D2 (Phase 4); E (Phase 3); F (transition Phase 2→3). DNA extraction and sequencing: DNA from filters was extracted (Qiagen PowerSoil). Shotgun metagenomes were sequenced (Illumina HiSeq, 2×150 bp). Reads were quality trimmed (bbduk, Sickle) and assembled with metaSPAdes (k-mers 33,55,77,99,127); scaffolds ≥2.5 kb were retained. Read mapping used Bowtie2. Binning used CONCOCT, MaxBin2, MetaBAT2 with dereplication by DAS Tool; quality assessed by CheckM. In total, 1,881 MAGs (>70% complete, <10% contamination) were retained. Taxonomy and annotation: MAG taxonomy used GTDB-Tk r89. Metabolic annotation employed DRAM v1.0 (default databases; UniRef90 added for select lineages). Relative abundance was estimated from mean scaffold coverage per MAG normalized to site totals; family- and phylum-level abundances summed across MAGs. Beta diversity used Bray–Curtis NMDS (vegan in R). Guild identification and classification: Putative methanogens were identified by presence of mcrA and/or ≥75% completion of CO2→methane pathway; additional mcr operon genes (mcrBCDG) were verified when mcrA was absent. Methanogens were categorized by substrate use: strictly hydrogenotrophic (mcrA and/or ≥75% hydrogenotrophic pathway; lacking CODH/ACS), acetoclastic (acetate activation genes plus >50% CODH/ACS and >75% hydrogenotrophic pathway), methylotrophic (mcrA plus methyltransferases; lacking CODH/ACS and with <50% hydrogenotrophic pathway), or broad substrate (mcrA plus capacity for multiple substrates and >75% hydrogenotrophic pathway). Putative aerobic methanotrophs were identified by pmoA and/or mmoX; ANME (e.g., Methanoperedenaceae) were classified as anaerobic methanotrophs. Novel putative methanotroph annotations were validated by verifying multiple pmo/mmo genes on ≥3 kb scaffolds. Nitrogen and sulfur reduction pathways and terminal oxidases (high/low-affinity complex IV) were evaluated from DRAM outputs. Meta-analysis: Literature on landfill methanogens/methanotrophs over ~20 years was compiled at family level across methods (16S, mcrA, metagenomics, FISH, qPCR, etc.), with taxonomic names harmonized to GTDB r89. Additional analyses profiled acetate-cycling potential (Wood-Ljungdahl, CODH/ACS, phosphotransacetylase/acetate kinase) and phylogenomics of candidate methanotroph families lacking established methanotrophy (Nevskiaceae, Acetobacteraceae, Mycobacteriaceae), using GToTree and DRAM for gene presence mapping.

Geochemistry and lifecycle: Cells A, B (oldest) were Phase 5 with low gas and evidence of oxygen intrusion; C was Phase 4; D1 and D2 Phase 4 but with heterogeneity (D2 higher BOD/COD and acetate ~74 mg L−1 vs 2.3 mg L−1 at D1); E was Phase 3 with high methane but transiently low bicarbonate on sampling day; F (F1/F2) was transitioning from Phase 2 to Phase 3 with variable methane output and increasing bicarbonate. Community diversity: 1,881 MAGs were recovered; per-cell MAG counts: A 93, B 188, C 134, D1 220, D2 269, E 210, F1 294, F2 239. Newer cells (D2, E, F1, F2) had more similar community composition and higher richness than older cells (A, B, C). Phylum-level patterns: Proteobacteria dominated older cells (34–46%) but <5% in newer cells; Bacteroidota (15–39%) enriched in methane-producing/transitioning cells (D2, E, F1, F2); Campylobacterota dominated C (48%) and were high in A (28%) and E (22%). Family-level shifts: Gallionellaceae (microaerophilic Fe-oxidizing autotrophs) were 10–26% in older cells, <5% in newer cells; Sulfurimonadaceae reached 26% in A; Arcobacteraceae dominated C (44%); Dysgonomonadaceae and Cloacimonadaceae (fermentative/acetogenic contributors) increased (12–20%) in methane-producing newer cells (E, F1, F2). Methanogens: 74 methanogen MAGs across Halobacterota (Halobacteriota), Thermoplasmatota, and Euryarchaeota, spanning 10 families: Methanobacteriaceae, Methanocorpusculaceae, Methanocullaceae, Methanofollaceae, Methanomethylophilaceae, Methanomicrobiaceae, Methanoregulaceae, Methanosarcinaceae, Methanospirillaceae, Methanotrichaceae; plus Methanofastidiosales (one MAG). Methanogens were generally low abundance; Methanocorpusculaceae and Methanocullaceae reached ~6% in newer sites (D2, F1, F2). D2 was a hotspot: a Methanocorpusculaceae MAG (STD2_64) was the most abundant genome at 6.66%. Substrate use varied by cell: A/B largely acetoclastic; C mixed (hydrogenotrophic, acetoclastic, methylotrophic); D1 resembled older cells (acetoclastic focus), while D2 supported broader pathways including strictly hydrogenotrophic (Methanocorpusculaceae, Methanofollaceae, Methanomicrobiaceae), acetoclastic (Methanocullaceae, Methanoregulaceae), methylotrophic (Methanomethylophilaceae), and broad-substrate Methanosarcinaceae. E (classified geochemically as Phase 3) had unexpectedly low methanogen abundance/diversity, suggesting a transition toward Phase 4 and sensitivity to bicarbonate fluctuations. F1/F2 harbored diverse methanogens with Methanosarcinaceae generalists possessing acetoclastic, hydrogenotrophic, and methylotrophic pathways. Methanotrophs and AOM: 31 putative aerobic methanotroph MAGs (15 pmoA-only, 5 mmoX-only, 11 both) plus 2 ANME (Methanoperedenaceae) were found mostly in cells with oxygen detected/suspected (A, B, C, D1; none in D2, F1, F2). Families included Methylomonadaceae (20 MAGs), Methylococcaceae (2), Methylacidiphilaceae (2), Methylomirabilaceae (1). Six MAGs belonged to lineages with poorly characterized methanotrophy (Acetobacteraceae, Nevskiaceae, Elusimicrobiota, Actinobacteriota/Mycobacteriaceae, Chloroflexota). Methanotrophs were low abundance (typically 0.1–4%). Methylacidiphilaceae MAGs (cells B, C) had urease genes, suggesting landfill-specific adaptations for nitrogen and inorganic carbon acquisition. D1 housed both Methanoperedenaceae and Methylomirabilaceae, implicating anaerobic methane oxidation (AOM) coupled to nitrogen redox; Methanoperedenaceae MAGs displayed differing capacities (one with nitric oxide reduction; the other lacking nitrogen/sulfur reduction), while Methylomirabilaceae encoded nitrate-to-N2O reduction, potentially providing NOx intermediates for syntrophy. Broadly, 27/31 methanotroph MAGs encoded nitrite→NO reduction, 8/31 nitrate→N2O reduction, 24/31 high-affinity cytochrome bd-type oxidase, and 26/31 low-affinity complex IV, supporting survival and methane oxidation under oxygen limitation. Novel methanotroph candidates: Phylogenomics showed PMMO/SMMO gene clusters in multiple genomes within Nevskiaceae, Acetobacteraceae, and Mycobacteriaceae, indicating independent acquisitions and overlooked methanotrophic potential across diverse environments. Meta-analysis: Across 21 landfill studies, Methanosarcinaceae were most frequently detected (17/21), followed by Methanotrichaceae (15/21), Methanocullaceae (13/21), and Methanobacteriaceae (13/21). This study detected 11 methanogenic families, among the highest reported. Strict hydrogenotrophs (Methanofollaceae, Methanocorpusculaceae) and methylotrophs (Methanomethylophilaceae) were less frequently reported in literature but were important in specific cells here (e.g., D2, F).

Integrating geochemical records with genome-resolved metagenomics across cells spanning 39 years of landfill operation revealed that methane-cycling guilds respond to substrate availability and redox gradients over decadal timescales. Newer waste supports diverse fermenters and methanogens, while older waste favors autotrophic, microaerophilic, and versatile redox metabolisms that can inhibit methanogenesis via oxygen intrusion. The strong divergence between D1 and D2, despite close ages, underscores spatial heterogeneity and the importance of local hydrology and gas transport in shaping microbial processes. Methanotrophs occur at low abundance and are more restricted spatially than methanogens, yet exhibit adaptations (high-affinity oxidases, nitrogen/sulfur reduction) enabling methane oxidation in oxygen-limited habitats. The co-occurrence of Methanoperedenaceae and Methylomirabilaceae at D1, alongside widespread denitrification potential among putative methanotrophs, indicates that anaerobic and intra-aerobic methane oxidation pathways are relevant in landfill interiors. These findings address the research need to resolve long-term microbial controls on methane cycling and suggest current emission models likely underestimate methane sinks from anaerobic/intra-aerobic oxidation and overlook novel methanotrophic lineages. Incorporating microbial functional data can refine predictions of methane production and oxidation across landfill lifecycle phases and guide targeted biostimulation to enhance methane attenuation when energy recovery is no longer viable.

This work provides a decadal, genome-resolved perspective on landfill microbial succession and methane cycling. Key contributions include: (1) demonstrating that newer cells host richer, compositionally similar communities dominated by fermenters and diverse methanogens, while older cells shift toward autotrophic, microaerophilic guilds; (2) resolving methanogen diversity and metabolic breadth across cells and identifying hotspots (e.g., D2) where multiple methanogenic pathways coexist; (3) documenting widespread adaptations among methanotrophs for oxygen-limited conditions, and detecting anaerobic methane oxidizers (Methanoperedenaceae, Methylomirabilaceae) within landfill interiors; and (4) revealing overlooked methanotrophic potential in Nevskiaceae, Acetobacteraceae, and Mycobacteriaceae. These insights argue for integrating microbial functional data into landfill models to better manage methane emissions. Future research should validate predicted methanotrophy in candidate lineages, quantify in situ AOM and intra-aerobic oxidation rates, link activity to geochemical drivers (e.g., bicarbonate, nitrogen oxides), and test biostimulation strategies to promote methane oxidation in anoxic landfill zones.

Metagenomics infers potential but not activity, limiting discrimination between carbon assimilation and energy conservation roles of shared pathways. Classification of substrate use (e.g., acetoclastic vs. hydrogenotrophic) can be ambiguous for some families (e.g., Methanocullaceae, Methanobacteriaceae) without physiological evidence. Sampling represents leachate at specific wells and times and may not capture spatial heterogeneity or align with aggregate gas data; transient geochemical shifts (e.g., bicarbonate decline in cell E) may decouple abundance from activity. The study focuses on one landfill site; broader generalization requires multi-site validation. Lack of metatranscriptomic, proteomic, or stable isotope labeling limits direct rate estimates and confirmation of anaerobic or intra-aerobic methane oxidation pathways.