Amplified Arctic warming can induce strong ecosystem changes with adverse climate feedbacks through greenhouse gas (GHG) release. Shifting plant species and traits with permafrost thaw may contribute to the permafrost carbon feedback. How vegetation dynamics in thawing permafrost systems affect GHG release and how this varies with season, plant species, and soil conditions is poorly understood. Here, we assessed GHG emissions, redox potentials, and geochemical signatures as well as the carbon input in the form of root exudation along a vegetation density gradient and a permafrost thaw gradient over a growing season in Stordalen mire, Sweden. Ecosystem respiration and CH₄ emissions increased along the thaw gradient from bog to fen, possibly due to high graminoid root carbon release rates into an anoxic soil, fuelling fast organic matter oxidation and lowering redox potentials to enhance methanogenesis. CH₄ emissions increased seven-fold with increasing graminoid cover compared to non-vascular plant controls in the thawed soil. Plants may mediate CH₄ transport, which was responsible for 80% of the graminoid-induced increase in CH₄ emissions in the bog environment. In the fen environment, graminoid root carbon release stimulated CH₄ formation, which dominated by contributing 70% of the graminoid-induced increase. Overall, photosynthesis-related CO₂ fixation was substantial in the early and peak growing season, but when expressed as CO₂ equivalents, CH₄ release offset this uptake, resulting in net positive radiative forcings from graminoid-vegetated thawed soils throughout the growing season. Graminoids increased the net CO₂-equivalent flux up to 8.9-fold compared to non-vascular plant locations with the strongest forcing toward late season in graminoid-vegetated fens. Our study showcases how fine-scaled, plant-mediated processes differently contribute to GHG emissions across a thawed bog and fen soil and how the time of growing season can overprint these effects to determine whether the system is a net GHG source or sink.

Arctic permafrost faces multiple interactive changes in thaw, drainage, and vegetation shift due to climate warming. Thaw-induced waterlogging can shift vegetation from shrubs to graminoids, altering greenhouse gas emissions. This study aims to quantify and mechanistically explain how vegetation-specific root exudates amplify or dampen greenhouse gas emissions from Arctic permafrost during thaw and drainage transitions. Using field observations and soil incubations, we show that environmentally relevant concentrations of root organic exudates (15% of dissolved organic carbon), obtained from thaw stage-specific plants, altered greenhouse gas fluxes under site-realistic redox conditions. Graminoid exudates were richer in sugars and carboxylates, whereas shrub exudates were richer in amino acids. In thawed, anoxic permafrost soil incubations, graminoid exudates stimulated the emission of 51% more CO₂ and 83% more CH₄ compared to the absence of exudates. In intact, drained permafrost soil, shrub exudates stimulated 14% more CO₂ and negligible CH₄ compared to untreated soils. Geochemical and microbial analyses revealed that soil, including their hydrology, and exudate differences drove exudate and soil organic matter decomposition. These soil incubation findings were supported by field measurements: Bare locations per soil-habitat provided baseline greenhouse gas fluxes in relation to each soil's properties of moisture and geochemistry. Vegetation by graminoids increased greenhouse gas emissions from thawed permafrost soils significantly while shrubs barely affected greenhouse gas emissions from drained permafrost soils. Collectively, this study shows that thaw-specific vegetation shapes greenhouse gas fluxes, indicating that vegetation shifts can intensify radiative forcing beyond the known direct effect of permafrost thaw and associated hydrological transitions. Clarifying the context-dependence and mechanisms underlying these distinct exudate effects may improve projections of Arctic terrestrial climate feedback.

Global warming increases the vegetation cover and leads to shifts in vegetation types in the Arctic. An increase in the vegetation cover might substantially enhance carbon dioxide (CO₂) emissions from northern permafrost soils, since root exudation of labile carbon and nitrogen can stimulate soil organic matter (SOM) decomposition via the rhizosphere priming effect. The current understanding of Arctic rhizosphere priming largely rests on soil incubation studies that simulate root exudation by adding various organic substrates in varying concentrations to soils. How the specific exudates of different plants influence rhizosphere priming is unclear as Arctic plant root exudate release rates and composition are largely unknown. Using targeted and non-targeted liquid chromatography–mass spectrometry, we compared the exudate composition and exudation rates of total organic carbon, 7 organic acids, 14 amino acids and 9 carbohydrates from three abundant and functionally different tundra plants (Betula glandulosa, Alnus viridis and Eriophorum vaginatum). While organic carbon and primary metabolites exudation were similar among the studied plants despite their different nitrogen acquisition strategies, distinct differences between the plant species were found in the overall root exudate composition. Between 80 and 94 % of the root exudate metabolome was not shared among the three plants. Our findings indicate that a change in vegetation types across the Arctic will primarily alter the release of secondary plant metabolites into the soil and thereby could alter soil microbial processes. Our observations further suggest that previous laboratory experiments studying priming frequently oversaturated microorganisms with labile substrates compared to natural conditions; this highlights the need for more realistic priming studies. Our data on root exudation provide critical background information for improving laboratory experiments.

• Targeted removal experiments are a powerful tool to assess the effects of plant species or (functional) groups on ecosystem functions. However, removing plant biomass in itself can bias the observed responses. This bias is commonly addressed by waiting until ecosystem recovery, but this is inherently based on unverified proxies or anecdotal evidence. Statistical control methods are efficient, but restricted in scope by underlying assumptions.
• We propose accounting for such biases within the experimental design, using a gradient of biomass removal controls. We demonstrate the relevance of this design by presenting (1) conceptual examples of suspected biases and (2) how to observe and control for these biases.
• Using data from a mycorrhizal association-based removal experiment, we show that ignoring biomass removal biases (including by assuming ecosystem recovery) can lead to incorrect, or even contrary conclusions (e.g. false positive and false negative). Our gradient design can prevent such incorrect interpretations, regardless of whether aboveground biomass has fully recovered.
• Our approach provides more objective and quantitative insights, independently assessed for each variable, than using a proxy to assume ecosystem recovery. Our approach circumvents the strict statistical assumptions of, for example, ANCOVA and thus offers greater flexibility in data analysis.

Vegetation changes in a warming Arctic may affect plant-associated soil microbial communities with possible consequences for the biogeochemical cycling of carbon (C) and nitrogen (N). In a sub-arctic tundra heath, we factorially removed plant species with ecto- and ericoid mycorrhizal associations. After two years, we explored how mycorrhizal type-specific plant removal influences microbial communities, soil and microbial C and N pools, and extracellular enzymatic activities. Removal of ecto- and ericoid mycorrhizal plants did not change the soil fungal or bacterial community composition or their extracellular enzyme activities. However, ericoid plant removal decreased microbial C:N ratio, suggesting a stoichiometric effect decoupled from microbial community composition. In other words, microbial communities appear to show initial plasticity in response to major changes in tundra vegetation. This highlights the importance of longer-term perspectives when investigating the effects of vegetation changes on biogeochemical processes in Arctic ecosystems.

Rapid Arctic warming increases permafrost thaw and CO₂ production from soil organic matter decomposition, but also enhances CO₂ uptake by plants. Conversely, plants can also stimulate soil organic matter decomposition near their roots, via rhizosphere priming. The recent PrimeSCale model suggests that this can accelerate Arctic soil carbon loss at a globally relevant rate, and points to large potential contributions from carbon-rich permafrost peatlands. At the same time, the high carbon content of peatlands might render them insusceptible to input of easily available organic compounds by plant roots, which is considered a key component of priming. We here investigated the sensitivity of permafrost peat soils to priming by plant compounds under aerobic conditions that resemble the dominant rooting zone, based on a 30-week laboratory incubation of peat soils from five circum-Arctic locations. No significant change in CO₂ production from peat organic matter by organic carbon addition was observed, and an increase of 24% by organic nitrogen addition. Combining our data with a literature meta-analysis of priming studies showed similar, low priming sensitivity in organic layers of mineral soils, and significantly stronger priming in mineral horizons where organic carbon and nitrogen increased decomposition by 32% and 62%, respectively. Low sensitivity of permafrost peat to input of organic compounds was also supported under anaerobic conditions, by incubation of one soil type. In a new PrimeSCale sensitivity analysis, we show that excluding peatlands would reduce estimates of priming-induced carbon loss from the circum-Arctic by up to 40% (up to 18 Pg) until 2100, depending on peat priming sensitivity. While our study suggests a limited effect of plant-released organic compounds on peat decomposition, it does not preclude an effect of vegetation on decomposition under natural conditions, through other mechanisms. The large range of possible priming-induced peat carbon losses, and expected changes in vegetation and drainage, call for a sharpened focus on the combined effect of living plants on soil processes beyond carbon input, including changes in nutrient and water availability, aggregation, and microbial communities.

Arctic soils play an important role in Earth's climate system, as they store large amounts of carbon that, if released, could strongly increase greenhouse gas levels in our atmosphere. Most research to date has focused on how the turnover of organic matter in these soils is regulated by abiotic factors, and few studies have considered the potential role of biotic regulation. However, arctic soils are currently missing important groups of soil organisms, and here, we highlight recent empirical evidence that soil organisms' presence or absence is key to understanding and predicting future climate feedbacks from arctic soils. We propose that the arrival of soil organisms into arctic soils may introduce novel functions, resulting in increased rates of, for example, nitrification, methanogenesis, litter fragmentation, or bioturbation, and thereby alleviate functional limitations of the current community. This alleviation can greatly enhance decomposition rates, in parity with effects predicted due to increasing temperatures. We base this argument on a series of emerging experimental evidence suggesting that the dispersal of until-then absent micro-, meso-, and macroorganisms (i.e. from bacteria to earthworms) into new regions and newly thawed soil layers can drastically affect soil functioning. These new observations make us question the current view that neglects organism-driven alleviation effects when predicting future feedbacks between arctic ecosystems and our planet's climate. We therefore advocate for an updated framework in which soil biota and the functions by which they influence ecosystem processes become essential when predicting the fate of soil functions in warming arctic ecosystems.

Contrary to most soils, permafrost soils have the atypical feature of being almost entirely deprived of soil fauna. Abiotic constraints on the fate of permafrost carbon after thawing are increasingly understood, but biotic constraints remain scarcely investigated. Incubation studies, essential to estimate effects of permafrost thaw on carbon cycling, typically measure the consequences of permafrost thaw in isolation from the topsoil and thus do not account for the effects of altered biotic interactions because of e.g. colonization by soil fauna. Microarthropods facilitate the dispersal of microorganisms in soil, both on their cuticle (ectozoochory) and through their digestive tract (endozoochory), which may be particularly important in permafrost soils, considering that microbial community composition can strongly constrain permafrost biogeochemical processes.

Here we tested how a model species of microarthropod (the Collembola Folsomia candida) affected aerobic CO₂ production of permafrost soil over a 25 d incubation. By using Collembola stock cultures grown on permafrost soil or on an arctic topsoil, we aimed to assess the potential for endo- and ectozoochory of soil bacteria, while cultures grown on gypsum and sprayed with soil suspensions would allow the observation of only ectozoochory.

The presence of Collembola introduced bacterial amplicon sequence variants (ASVs) absent in the no-Collembola control, regardless of their microbiome manipulation, when considering presence–absence metrics (unweighted UniFrac metrics), which resulted in increased species richness. However, these introduced ASVs did not induce changes in bacterial community composition as a whole (accounting for relative abundances, weighted UniFrac), which might only become detectable in the longer term.

CO₂ production was increased by 25.85 % in the presence of Collembola, about half of which could be attributed to Collembola respiration based on respiration rates measured in the absence of soil. We argue that the rest of the CO₂ being respired can be considered a priming effect of the presence of Collembola, i.e. a stimulation of permafrost CO₂ production in the presence of active microarthropod decomposers. Overall, our findings underline the importance of biotic interactions in permafrost biogeochemical processes and the need to explore the additive or interactive effects of other soil food web groups of which permafrost soils are deprived.

As global temperatures continue to rise, a key uncertainty of climate projections is the microbial decomposition of vast organic carbon stocks in thawing permafrost soils. Decomposition rates can accelerate up to fourfold in the presence of plant roots, and this mechanism-termed the rhizosphere priming effect-may be especially relevant to thawing permafrost soils as rising temperatures also stimulate plant productivity in the Arctic. However, priming is currently not explicitly included in any model projections of future carbon losses from the permafrost area. Here, we combine high-resolution spatial and depth-resolved datasets of key plant and permafrost properties with empirical relationships of priming effects from living plants on microbial respiration. We show that rhizosphere priming amplifies overall soil respiration in permafrost-affected ecosystems by similar to 12%, which translates to a priming-induced absolute loss of similar to 40 Pg soil carbon from the northern permafrost area by 2100. Our findings highlight the need to include fine-scale ecological interactions in order to accurately predict large-scale greenhouse gas emissions, and suggest even tighter restrictions on the estimated 200 Pg anthropogenic carbon emission budget to keep global warming below 1.5 degrees C.