High but variable CH4 emissions post-fire is associated with stochastic recruitment and assembly of methanogens

Abstract

Frequent fires in tropical Peat Swamp Forests (PSFs) that have been repurposed for forestry and agriculture result in substantial emissions of locked carbon. Fire-affected PSFs emit double the amount of CH4 compared to intact ones over extended periods of time. CH4 production is largely driven by communities of microorganisms (microbiomes – archaea in particular), thus, our ability to reduce emissions hinges on (i) identifying those with the capacity to generate CH4, (ii) ecological processes that shape their composition and functioning, and (iii) environmental variables which drive them. Ecological processes are particularly important as they determine our ability to predict the trajectory of communities and their functioning. Trajectories of communities shaped by deterministic processes can be predicted based on environmental variables as opposed to those shaped by stochastic processes whose trajectories are difficult to predict. We fill this knowledge gap by sequencing and comparing the ecological processes shaping peat microbiomes from a fire-impacted PSF to an intact PSF that occur within the same peat dome in Brunei. The composition of archaeal communities were significantly different between the fire-impacted and intact PSFs and strongly stratified by depth. The largest difference was observed between communities from the surface (0-5 cm) and those from below the water table (95-100 cm). In the fire-impacted PSF, archaea doubled in abundance in the anoxic zone compared to the surface, while, no such change was detected in the intact PSF. Archaeal communities occurring in the anoxic layers of the fire-impacted PSF were dominated by methanogens from the class Methanomicrobia and Bathyarchaeia, both of which occur in high-methane flux habitats. We determined ecological processes shaping the assembly of these methanogenic populations using bin-based phylogenetic null models. This showed that methanogenic populations in the fire-impacted PSF were largely shaped by stochastic processes, whereas, similar populations, albeit at lower abundances, were shaped by deterministic processes in the intact PSF. Changes in pH and dissolved oxygen correlated strongly with differences in assembly processes. Our work shows that changes in the environment resulting from fires can set methanogenic communities on unpredictable trajectories, which in turn correlate strongly with both increased and non-homogeneous CH4 emissions. Altering these key environmental correlates could form the basis for developing nature-based solutions for reducing emissions from fire-impacted peatlands.