Abrupt permafrost thaw triggers activity of copiotrophs and microbiome predators

Abstract

Permafrost soils store a substantial part of the global soil carbon and nitrogen. However, global warming causes abrupt erosion and gradual thaw, which make these stocks vulnerable to microbial decomposition into greenhouse gases. Here, we investigated the microbial response to abrupt in situ permafrost thaw. We sequenced the total RNA of a 1 m deep soil core consisting of up to 26 500-year-old permafrost material from an active abrupt erosion site. We analysed the microbial community in the active layer soil, the recently thawed, and the intact permafrost, and found maximum RNA:DNA ratios in recently thawed permafrost indicating a high microbial activity. In thawed permafrost, potentially copiotrophic Burkholderiales and Sphingobacteriales, but also microbiome predators dominated the community. Overall, both thaw-dependent and long-term soil properties significantly correlated with changes in community composition, as did microbiome predator abundance. Bacterial predators were dominated in shallower depths by Myxococcota, while protozoa, especially Cercozoa and Ciliophora, almost tripled in relative abundance in thawed layers. Our findings highlight the ecological importance of a diverse interkingdom and active microbial community highly abundant in abruptly thawing permafrost, as well as predation as potential biological control mechanism.

Keywords: abrupt erosion; copiotrophic; permafrost; protozoa; transcriptomics.

Figure 1.

(A) Sampling site Zackenberg in Northeast Greenland, credit: Google Earth. (B) Sampling site in 2018 after initial permafrost collapse. (C) Soil profile from the surface until still frozen depth at 90 cm during sampling in 2020. (D) Scheme of abrupt permafrost thaw (indicated with blue arrows), depicting the soil profile until the permafrost (PF) layer at 90‒100 cm depth from the moment of collapse in 2018 with visible ice lens below long-term active layer (AL) to the formation of transition zones (TZ = thawed permafrost).

Figure 2.

Venn diagram illustrating variance partitioning, which was performed on thaw-related (layer, H₂O), long-term soil (SOM, pH, age) as well as biotic parameters (bacterial and protozoan predator abundance) and their correlation with variance in the overall community composition. Values are given for R² values, based on PERMANOVA results (999 permutations) on Bray–Curtis dissimilarities.

Figure 3.

All figures are given for subsets of the total community with prokaryotes, bacterial predators, eukaryotes, and protozoan rRNA. Bar plots indicating the relative mean abundance per depth for varying taxonomic orders; * indicates undefined taxonomic levels/incertae sedis. Shannon alpha diversity is given as boxplots with whiskers indicating standard deviation across triplicates. Nonmetric multidimensional scaling (NMDS, middle column) ordination plots performed on rRNA contig abundances per sample. Colours indicate different layers and shape different age horizons. Environmental parameters here are given as layer (active layer: AL, transition zone 1: TZ1 and 2: TZ2, and permafrost: PF) as well as age with young soils (AY), 2634‒3770-year-old soil of medium age (AM), and old material (AO) of up to 22 100‒26 500 years ago.