Contrasting responses of biological nitrogen fixation in soil and moss to ecosystem warming in an alpine permafrost ecosystem

Due to persistent low temperatures, permafrost ecosystems experience constraints in nitrogen (N) turnover, resulting in long-term ecosystem nitrogen limitation. Climate warming would induce the release of N trapped in permafrost, making it available for plant growth and thereby enhancing ecosystem carbon sequestration. Additionally, increased soil N availability would alleviate nutrient limitations for soil microorganisms, promoting greenhouse gas emissions through enhanced soil organic matter decomposition. Against this background, it is crucial to resolve the response of N cycling in permafrost ecosystems to warming, to accurately understand the feedbacks between permafrost carbon-nitrogen dynamics and climate warming. Given the low N deposition in permafrost zones, biological nitrogen fixation (BNF) serves as the primary N input for ecosystems (~50-80% of the annuals). Nitrogen-fixing microbes in moss and soil play crucial roles in BNF in permafrost ecosystems, through symbiotic and autotrophic pathways. Warming may induce alteration in moss and soil characteristics (e.g. moss and soil drying), which would subsequent affect BNF in moss and soil. However, it remains unclear whether microbial BNF in moss and soil would exhibit contrasting responses to warming, and how active nitrogen-fixing microbes modulates such responses. To address these questions, we performed an interactive experiment involving warming and moss removal (warming vs. ambient × moss removal vs. retention) in response to whole-ecosystem warming at the Simulate Warming at Mountain Permafrost (SWAMP). In the in situ labeling procedure, we took two soil columns (10 cm in diameter and in depth) in the surface of each plot, separated them in the middle, and placed them in two incubation container with a dividers, designing one side for moss removal and the other side for moss retention. The two containers were filled separately with 10% ¹⁴N₂ and ¹⁵N₂, incubated in situ for 7 days to determine the BNF rate. In the in-house experiment, we employed the ¹⁵N-DNA Stable-Isotope Probing to elucidate changes in active microbial groups engaged in BNF, allowing us to uncover their impacts on regulating BNF to warming. Warming resulted in a significant reduction of moss cover by 37.8%. Concurrently, BNF rate significantly increased under warming conditions, especially in the moss-retention treatment. Conversely, warming did not alter BNF rate in the moss-removal treatment. Such finding suggested that warming enhance BNF rate primarily by stimulating a higher microbial BNF rate in moss rather than in soil. The results of microbial functional genes showed that, for moss, although warming didn’t affect the richness of nifH genes, but significantly reduced the Shannon-Wiener index and evenness, leading to an altered functional structure; for soils, warming didn’t change functional structure or any microbial α-diversity indices of nifH genes. These results suggest that the potential for BNF by moss would be further stimulated under warming, resulting in a higher N fixation efficiency. These gains may compensate for the decline in ecosystem-level BNF triggered by the reduction in moss cover. In other words, N supply from BNF in permafrost ecosystems will not decrease due to the trade-off between decreased cover and enhanced BNF ability for moss in a warmer scenario.