Elevated methane flux in a tropical peatland post-fire is linked to depth-dependent changes in peat microbiome assembly

Abstract

Fires in tropical peatlands extend to depth, transforming them from carbon sinks into methane sources and severely limit forest recovery. Peat microbiomes influence carbon transformations and forest recovery, yet our understanding of microbiome shifts post-fire is currently limited. Our previous study highlighted altered relationships between the peat surface, water table, aboveground vegetation, and methane flux after fire in a tropical peatland. Here, we link these changes to post-fire shifts in peat microbiome composition and assembly processes across depth. We report kingdom-specific and depth-dependent shifts in alpha diversity post-fire, with large differences at deeper depths. Conversely, we found shifts in microbiome composition across all depths. Compositional shifts extended to functional groups involved in methane turnover, with methanogens enriched and methanotrophs depleted at mid and deeper depths. Finally, we show that community shifts at deeper depths result from homogeneous selection associated with post-fire changes in hydrology and aboveground vegetation. Collectively, our findings provide a biological basis for previously reported methane fluxes after fire and offer new insights into depth-dependent shifts in microbiome assembly processes, which ultimately underlie ecosystem function predictability and ecosystem recovery.

Fig. 1. Post-fire shifts in microbiome diversity and composition.

a Archaea-to-Bacteria ratio increased significantly in deep peat post-fire (Pairwise t test, t: 5.88, p < 0.001), indicating a shift in relative abundances. b No significant difference was observed in archaeal diversity post-fire across any depth. c Bacterial diversity significantly decreased in deep peat post-fire (Pairwise t test, t: -4.05, p = 0.006). a–c Show four independent biological replicates per group. Significance is expressed as ***p < 0.001; **p < 0.01; *p < 0.05. Canonical analysis of principle coordinates of surface, mid, and deep peat show shifts in (d) archaeal and (e) bacterial communities after fire. Symbols in (d, e) represent community centroids for each depth with 95% confidence ellipses (n = 4 for all groups).

Fig. 2. Compositional shifts across peat depth after fire.

Post-fire shifts in the top five abundant classes within (a) archaeal and (b) bacterial communities. These classes were identified based on the harmonic mean of relative abundance and prevalence. Together, these classes represent approximately 68.8–81.64% of bacterial and 98.28–99.53% of archaeal communities, respectively. Mean relative abundances were computed from four independent biological replicates per group.

Fig. 3. Post-fire differences in depth-stratification of functional groups involved in methane turnover.

a Methanogens significantly increased relative to other archaea, and (b) methanotrophs significantly decreased relative to other bacteria at mid and deeper depths. c Methanotrophs-to-methanogens significantly decreased at mid and deeper depths after fire. a–c Represent four biological replicates per group, except for the post-fire surface in (c) which includes only two replicates with nonzero methanogen relative abundances. Significance was assessed using pairwise t-tests and is expressed as ***p < 0.001; **p < 0.01; *p < 0.05. Exact p values with test statistics are provided in the main text. d Changes in relative abundance of predicted methanogens across depth between burnt and intact sites. Each column represents an ASV, and their class labels are displayed above the tree. Mean relative abundance was computed from four biological replicates and square-root transformed for visual clarity. The colour strip at the top represents predicted methanogenesis pathways, and columns are hierarchically clustered based on Spearman-rank correlations.

Fig. 4. Depth-dependent shifts in the relative importance of different assembly processes after fire.

Post-fire changes in dispersal limitation, drift, homogeneous selection, and stochasticity governing archaeal and bacterial community assembly across depth. Stochasticity represents the combined relative importance of drift, dispersal limitation, and homogenizing dispersal. Data (n = 6 comparisons among four biological replicates) are presented as mean value ± SD, with error bars representing standard deviations. Significance was assessed using one-sided bootstrap tests and is indicated as ***p < 0.001; **p < 0.01; *p < 0.05.

Fig. 5. Post-fire shifts in the relative importance of assembly processes governing different archaeal groups at deeper depth.

Shaded bars represent the dominant assembly process governing a particular phylogenetic bin. This was identified qualitatively as the process with the highest relative importance. Data (n = 6 comparisons among four biological replicates) are presented as mean values.