Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Feb 16;32(2):e70718. doi: 10.1111/gcb.70718

Climate Change Impacts the Structure and Nitrogen‐Fixing Activities of Subarctic Feather Moss Microbiomes Across a Precipitation Gradient

Danillo O Alvarenga 1,2,✉, Justin T Wynns 1,3, Joseph Nesme 4, Anders Priemé 2,4, Kathrin Rousk 1,2
PMCID: PMC12907784  PMID: 41693497

ABSTRACT

Associations between feather mosses and cyanobacteria are crucial sources of new biologically available nitrogen (N) in arctic and subarctic ecosystems. The physiology of both mosses and cyanobacteria is strongly influenced by environmental factors such as temperature and moisture, which directly affect N2 fixation rates. These associations may be threatened by climate change, since it leads to warmer and drier conditions in polar regions. In this study, we investigated the N2‐fixing microbial communities associated with two common feather mosses across a precipitation gradient in the subarctic tundra, followed by a temperature and moisture experiment. Using acetylene reduction assays, nifH gene sequencing and qPCR, we evaluated how shifts in temperature and moisture influence nitrogenase activity and N2‐fixing community structure. Our results showed that N2 fixation was highest in sites with greater precipitation and increased with both temperature and moisture. Cyanobacteria dominated N2‐fixing communities, but currently unclassified bacteria also seemed to play a significant role, particularly at higher temperatures. The number of cyanobacterial nifH copies tended to remain stable or decrease with temperature, while the relative abundance of unclassified bacteria increased. These findings suggest that the N2‐fixing activity, abundance, and diversity of cyanobacteria associated with feather mosses in the subarctic will decline under warmer and drier conditions, potentially leading to a shift in the composition of feather moss‐associated microbial communities in a warmer Arctic, with potential consequences for N input into the ecosystem.

Keywords: Arctic tundra, Bryophyta, cyanobacteria, moisture, nitrogen fixation, temperature

Associations between mosses and cyanobacteria are crucial sources of new nitrogen in arctic and subarctic ecosystems. The physiology of both mosses and cyanobacteria is strongly influenced by environmental factors such as temperature and moisture, which directly affect N2 fixation rates. These associations may be threatened by climate change, since it leads to warmer and drier conditions in polar regions. In this study, we investigated the N2‐fixing microbial communities associated with two common feather mosses across a precipitation gradient in the subarctic tundra, followed by a temperature and moisture experiment. Our findings suggest that the N2‐fixing activity, abundance, and diversity of cyanobacteria associated with feather mosses in the subarctic will decline under warmer and drier conditions, potentially leading to a shift in the composition of feather moss‐associated microbial communities in a warmer Arctic, with potential consequences for N input into the ecosystem.

graphic file with name GCB-32-e70718-g005.jpg

1. Introduction

Mosses play crucial ecological roles in environments characterized by cold temperatures and low vascular plant abundance, such as tundra ecosystems. In the arctic and subarctic tundra, perennial moss species often dominate ground covers, driving carbon (C) sequestration and hosting nitrogen‐fixing cyanobacteria (Solheim et al. 2004; Douma et al. 2007; Street et al. 2013; Eldridge et al. 2023). Associations between mosses and cyanobacteria introduce significant amounts of new nitrogen (N) into tundra ecosystems despite lower temperatures, higher ultraviolet radiation rates during growing seasons, and fluctuating water availability (Kvíderová et al. 2019). However, the contributions of mosses and associated cyanobacteria to C and N cycles in these environments are increasingly threatened by climate change.

Arctic regions are warming much faster than the rest of the planet due to the polar amplification phenomenon resulting from the loss of sea ice (Dai et al. 2019; Rantanen et al. 2022; Xie et al. 2022), and currently experience their highest temperatures on record (Ballinger et al. 2024). One of the most alarming consequences of the increasing temperatures in these regions is lower humidity and increased frequency of drought events (Finger Higgens et al. 2019). Both moss and cyanobacteria are poikilohydric organisms, and as such they are incapable of regulating their moisture levels, depending entirely on environmental conditions to stay hydrated. Consequently, C sequestration by mosses is highly vulnerable to drought, which decreases their net primary production (Martínez‐García et al. 2024). The efficiency of N2 fixation in moss‐cyanobacteria associations is also significantly impacted by abiotic factors, especially temperature and moisture (Rousk, Jones, and DeLuca 2017; Rousk, Pedersen, et al. 2017; Rousk 2022). Experimental warming was shown to increase nitrogen fixation associated with feather mosses, but this effect was strongly influenced by annual precipitation and moisture availability (Lett et al. 2024). Nevertheless, the effects of climate change on moss‐cyanobacteria associations in arctic and subarctic ecosystems are still unclear.

Some works have shown that climate change directly and indirectly reduces N2 fixation, cyanobacterial colonization, and feather moss ground cover in subarctic ecosystems (Sorensen and Michelsen 2011; Sorensen et al. 2012; Alvarenga and Rousk 2021; Permin et al. 2022). On the other hand, other studies estimated that some of the effects of climate change may actually increase N2 fixation associated with subarctic mosses (Rousk 2022; Lett and Michelsen 2014; Rousk and Michelsen 2017; Lett et al. 2024). In addition, mosses may buffer ecosystems against the consequences of climate change by storing water, C, and N (Slate et al. 2024). These contrasting results indicate that the effects of climate change on the composition and productivity of these N2‐fixing associations are still poorly understood. This information is crucial for accurate predictions of how N‐limited ecosystems will be impacted by changing environmental conditions.

Understanding how climate change will affect the contributions of moss‐cyanobacteria associations to biogeochemical cycles is further complicated by the fact that mosses host distinct microbial communities. Moss microbiomes can vary based on both the identity and evolutionary history of the host species and environmental factors such as light or temperature (Holland‐Moritz et al. 2021). Warming may also shift the composition of N2‐fixing microbial communities (Klarenberg et al. 2022). Furthermore, moss microbiomes contain novel microorganisms with yet unclear ecological functions (Holland‐Moritz et al. 2018), introducing an element of unpredictability into our understanding of how these communities may respond to climate change. This variability in microbial composition poses significant challenges for predicting how N2 fixation will respond to climate change across various moss species and environments.

In this study, we evaluated how climate change‐driven temperature and water stress impacts N2 fixation associated with two dominant feather mosses from subarctic tundra. We sampled two feather moss species growing across a natural precipitation gradient in northern Sweden and analyzed N2 fixation rates, N2‐fixing community structure, and nifH gene abundance under different temperature and moisture levels. We hypothesized that: (1) N2 fixation rates are higher in feather mosses from sites with higher annual precipitation; (2) N2 fixation rates are positively correlated with higher temperatures and moisture levels; and (3) N2 fixation is positively correlated with a larger abundance of cyanobacteria in the feather moss microbiomes.

2. Methods

2.1. Overview of Sampling Sites

This study targeted three different sites in northern Sweden representing a steep precipitation gradient within 40 km: a low precipitation site close to Abisko (hereafter referred to as ABK) (68°20′57.7″ N, 18°49′48.1″ E); a medium precipitation site close to Låktajåkko (LTJ) (68°24′54.6″ N, 18°24′21.9″ E); and a high precipitation site close to Katterjåkk (KTJ) (68°25′10.1″ N, 18°09′51.5″ E). According to the Swedish Meterological and Hydrological Institute (2023) (https://www.smhi.se/), sites ABK, LTJ, and KTJ have mean annual precipitations of 300 mm, 800 mm, and 1200 mm, respectively, and annual mean temperatures are 0.3°C for ABK, −3.4°C for LTJ, and −0.7°C for KTJ. The location of the sampling sites is illustrated in Figure 1.

FIGURE 1.

FIGURE 1

Open in a new tab

Location of sampling sites in northern Sweden. The box in the bottom indicates annual precipitation and temperature averages in each site. ABK, Abisko; KTJ, Katterjåkk; LTJ, Låktajåkko.

Given their proximity, most environmental factors are similar between the sites, such as bedrock, age, and soil pH, which was around 6.0 (6.0 ± 0.2 at ABK, 6.1 ± 0.1 at LTJ, and 5.9 ± 0.2 at KTJ). The three sites are also similar in vegetation, being often dominated by two feather moss species, Hylocomium splendens (Hedw.) Schimp. and Pleurozium schreberi (Brid.) Mitt., as well as the shrubs Empetrum hermaphroditum Hagerup, Vaccinium uliginosum L., and Betula pubescens Ehrh. However, Abisko lies in the rain shadow of the Scandinavian mountains on the eastern side, which partially blocks precipitation. In contrast, Låktajåkko and Katterjåkk are located on the western slopes toward the Norwegian border, where they are more exposed to heavier rainfall; therefore, creating the aforementioned gradient.

2.2. Sample Collection and Incubation

Sampling was carried out in October 2016, just before the first snowfall in northern Sweden. Six 18 × 18 × 18 cm mesocosms from each site containing the two dominant feather moss species, H. splendens and P. schreberi , were collected and placed in white plastic boxes. H. splendens was not found at LTJ; therefore, only P. schreberi was sampled. Mesocosm boxes were shipped to the University of Copenhagen for downstream experimental work.

Randomly selected 3 g portions of moss shoots from each mesocosm were placed in sterile 50 mL plastic tubes and kept at different temperature regimes (5°C, 15°C, or 25°C). As a full‐factorial setup, for each temperature, the samples were also exposed to the moisture levels of 50 (consisting of partially hydrated shoots), 100 (fully hydrated shoots), or 150% (fully hydrated shoots with an extra volume of water corresponding to half of its moisture) in relation to the dry weight of each sample. The lower moisture levels used in the experiment represent what is usually found in a summer day in the sites; the medium level simulates what moss carpets experience after rain events (usually in the autumn); and the high moisture level represents water saturated conditions after snow melt.

Six replicates were used in the different treatments, resulting in 270 samples. The samples were kept in growth chambers under a photoperiod of 12 h of light at 200 ± 25 μmol photons·m−2·s−1 and 12 h of darkness for 8 weeks, being monitored for water loss once every week. The 12 h photoperiod was chosen to simulate an average diurnal rhythm as much as possible and to eliminate the potential influence of light on the samples.

2.3. Acetylene Reduction Assays

We performed acetylene reduction assays (ARA) as an indirect measure of N2 fixation in the samples (Hardy and Knight 1967). Each 50 mL tube prepared in the previous step was sealed with a rubber Suba Seal septum (Sigma‐Aldrich, Saint Louis, USA) and 10% of the headspace (5 mL) was replaced with acetylene gas. The vials were incubated under the previous conditions for durations that varied with incubation temperatures: samples at 5°C were incubated for 18.5 h; samples at 15°C for 7.5 h; and samples at 25°C for 3 h. Six mL of the vial headspaces were transferred into pre‐evacuated, air‐tight 6 mL Exetainer vials (Labco, Lampeter, UK) and analyzed for ethylene concentrations with an SRI 310C FID gas chromatograph (SRI Instruments, Torrance, USA). Gas chromatography was performed after the samples were incubated for 1 day, 1 week, 2 weeks, 4 weeks, and 8 weeks.

2.4. DNA Isolation

After 8 weeks of incubation, the samples were freeze‐dried and weighed, averaging 30.2 ± 0.1 mg. The greenest parts of the feather moss shoots were separated from older portions and discolored tips, if present, and any organic material not belonging to the target moss species was removed. Five of the six ARA replicates in each treatment (225 samples) were selected for qPCR and DNA sequencing. These samples were cut into fine fragments with sterile scissors and transferred to PowerBead Pro tubes (Qiagen, Hilden, Germany). The cells were broken in the FastPrep‐24 benchtop homogenizer (MP Bio, Solon, USA) at 5.5 m·s−1 for 40 s and DNA isolation was carried out with the DNeasy PowerSoil Pro Kit (Qiagen). Isolated DNA was screened on 1% agarose gels and stored at −20°C.

2.5. nifH Gene Amplification and Sequencing

Amplification of the nifH gene was performed using an inosine‐free modification of the primers 19F (5′‐GCN WTY TAY GGN AAR GGN GG‐3′) and 407R (5′‐AAN CCR CCR CAN ACN ACR TC‐3′) (Ueda et al. 1995). PCR was performed with the PCRBIO HiFi Polymerase (PCR Biosystems, London, UK) under the following conditions: initial denaturation at 95°C for 1 min; 30 cycles of 95°C for 15 s, 51°C for 15 s, and 72°C for 30 s; and final extension at 72°C for 6 min. These PCR products were eluted in nuclease‐free water and used for a second PCR with dual indices and primers modified with the Illumina overhand adapter sequences (5′‐TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG‐3′ and 5′‐GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA G‐3′ to primers 19F and 407R, respectively) (Illumina 2013) using the Nextera XT Index Kit (Illumina, San Diego, USA). The products of each PCR reaction were purified with the HighPrep PCR Kit (MagBio Genomics, Gaithersburg, USA). Samples were normalized using the SequalPrep Normalization Plate Kit (Applied Biosystems, Waltham, USA) and 5 μL of each was pooled together. The sample pool was concentrated using the DNA Clean & Concentrator‐5 kit (Zymo Research, Irvine, USA) and adjusted to 4 nM. The libraries were sequenced in the MiSeq platform using the MiSeq Reagent Kit v3 (Illumina).

2.6. nifH Amplicon Sequence Analysis

The nifH sequences obtained in the previous step were analyzed with the QIIME 2 pipeline version 2023.5 (Bolyen et al. 2019). Primers and adapter sequences were trimmed from raw reads with the q2‐cutadapt plugin. The results were denoised and representative amplicon sequence variants (ASVs) were selected with q2‐dada2. The nitrogenase gene nifH shares a common evolutionary origin and conserved sequence signatures with the chlorophyll and bacteriochlorophyll genes chlL and bchL, thus amplification of these homologous photosynthetic genes is common for PCRs targeting nifH (Gupta 2012; Tahon et al. 2016). Therefore, the inferred ASVs were compared against the nifH, chlL, and bchX database included with NifMAP v. 1.2 (Angel et al. 2018) using FrameBot 1.2.0 (Cole et al. 2014), and those that were closer to nifH than the other genes were retrieved with a custom script. A reference nifH database was prepared from the nifHdada2 2.0.5 dataset (Heller et al. 2014; Moynihan and Furbo Reeder 2025), and the ASVs were taxonomically identified using the q2‐feature‐classifier plugin. ASVs identified potentially belonging to plastids or mitochondria as well as nifH homologs (e.g., bchX and chlL) were filtered from the dataset. The ASVs were aligned with MAFFT (Katoh and Standley 2013) and a maximum likelihood phylogenetic tree was reconstructed with FastTree (Price et al. 2010) via the q2‐phylogeny plugin. Alpha and beta diversity analyses were performed with the q2‐diversity plugin.

2.7. Quantitative PCR of the nifH Gene

Standards were produced by amplifying the nifH gene from a plasmid standard derived from the bacterium Ensifer meliloti (Dangeard) Young (Pedersen 2017) using the modified 19F and 407R primers under the conditions described above. The PCR products were visualized, pooled, and then cleaned using the QIAquick PCR Purification Kit (Qiagen). qPCR was performed with the Brilliant III Ultra‐Fast SYBR Green qPCR Master Mix (Agilent Technologies, Cedar Creek, USA) using inosine‐free versions of the 19F and 407 primers. All reactions were performed in three technical replicates on 96‐well plates, which also included serial dilutions of the standard in two technical replicates, with concentrations decreasing exponentially from 1.35 × 109 and 1.35 × 103 copies·μL−1 using the LightCycler 96 Real‐Time PCR System (Roche Diagnostics, Mannheim, Germany). Reaction conditions were as follows: initial denaturing at 95°C for 10 min; 40 cycles of 95°C for 30 s, 51°C for 30 s, and 72°C for 30 s; and a final cycle of 95°C for 1 min, 55°C for 15 s, and 95°C for 15 s. For each sample, the procedure was repeated until a quantification cycle variance of less than 1 was achieved between technical replicates (Bustin et al. 2009), after which the average of the three corresponding concentrations was calculated.

2.8. Adjustment of nifH Quantifications

The sequencing results showed that the 19F and 407R primers were not highly specific to nifH, also capturing the paralogous chlL/bchL, especially from moss chloroplasts. Although it is possible to remove the nontarget genes from the dataset when sequences are available, this is not the case with qPCR data, which can result in overestimating nifH expression. However, using the relative abundance data obtained in the previous step, it was possible to determine which percentage of the qPCR dataset resulted from the amplification of the nontarget genes and subtract them from it. Therefore, the nifH qPCR average concentration values were corrected by multiplying absolute values by the fraction of nifH reads recovered from each sample as estimated in the sequence analyses (Angel et al. 2018). Finally, the number of nifH copies for specific taxa was estimated by multiplying the copy number determined by qPCR results for each sample by the corresponding relative abundances obtained with the nifH amplicon analyses.

2.9. Statistical Analyses and Visualization

The data from the acetylene reduction assays, nifH sequencing, and qPCR were statistically analyzed with R 4.2.2 (R Core Team 2022) in Rstudio 2022.12.0 (Posit Team 2022). The normality of the datasets was visualized with Q‐Qplots and evaluated with Shapiro–Wilk's test (Shapiro and Wilk 1965), and the data were transformed with the Yeo‐Johnson method (Yeo and Johnson 2000) using the package car 3.1.3 (Fox and Weisberg 2019).

Differences between N2 fixation rates, ASV numbers, or nifH copies and different hosts, sites, or incubation regimes were evaluated with one‐way and two‐way analyses of variance (ANOVA) (Fisher 1918), and significant variation between factors was evaluated by the Tukey's Honestly Significant Difference post hoc test (Tukey 1949). Linear regressions (Galton 1886) were used to investigate the potential relationships between N2 fixation, temperature, moisture, and nifH copies. The relationship between temperatures and acetylene reduction rates was further investigated using the Arrhenius equation by comparing the natural logarithm of ethylene production against the inverse of absolute temperatures (Davidson and Janssens 2006). The apparent activation energy (Ea) was estimated from the slope of the resulting linear regression, which was used to calculate the temperature coefficient (Q 10), for which confidence intervals were derived by propagating the standard error of the regression slope.

Since differences between acetylene reduction analyses performed at different time points were negligible, the data obtained in different weeks were analyzed together. The results were visualized with the R packages ggplot2 3.5.1 (Wickham 2016), cowplot 1.1.1 (Wilke 2021), and patchwork 1.2.0 (Pedersen 2021). Additional plot editing was performed with the packages ggiraphExtra 0.3.0 (Moon 2021), ggeffects 2.2.1 (Lüdecke 2018), multcompView 0.1‐10 (Selzer 2021), and scales 1.4.0 (Wickham et al. 2021). The packages ggspatial 1.1.9 (Dunnington 2021) and maps 3.4.3 (Deckmyn 2021) were also used for the drawing of the sampling map.

The data and scripts used in these analyses were deposited in the Electronic Research Data Archive database (Alvarenga et al. 2026).

3. Results

3.1. Influence of Hosts and Environmental Factors on N2 Fixation

Overall, samples collected in ABK, where feather mosses experience lower precipitation, produced an average of 7.9 (with a standard error of ±0.7) nmol ethylene·g−1·h−1 in acetylene reduction assays. The average N2 fixation rate in ABK samples was less than half of those under higher annual precipitation in the LTJ and KTJ sites, which presented averages of 16.4 (±1.5) and 18.1 (±1.3) nmol ethylene·g−1·h−1, respectively (one‐way ANOVA, p < 0.001, F 2,1346 = 53.08) (Figure 2A). Considering all samples irrespective of origin, a significant, positive relationship was observed between N2 fixation and temperature (linear regression, p < 0.001, r 2  = 0.16) as well as N2 fixation and moisture, albeit with a very low explanatory power (p < 0.001, r 2  = 0.02) (Figure 2B,C). No interaction between host species and site was found, but a significant interaction between temperature and moisture was found when all samples were considered (two‐way ANOVA, p < 0.001, F 1,1345 = 29.16), thus suggesting the highest N2 fixation rates at the highest temperature (25°C) and moisture levels (150%).

FIGURE 2.

FIGURE 2

Open in a new tab

Nitrogenase activity associated with feather moss samples collected across a natural precipitation gradient in northern Sweden as estimated with acetylene reduction assays. (A) Moss‐associated nitrogenase activity in each site, from driest to wettest. Bars indicate averages, while vertical lines illustrate standard errors. Different letters above bars indicate significant differences in one‐way ANOVAs according to Tukey's Honest Significant Difference post hoc test; n = 540. (B) Relationship between increasing temperatures and the nitrogenase activities associated with feather mosses; n = 450. (C) Relationship between increasing moisture levels and nitrogenase activities; n = 450. (D) Multiple linear regression analyses of the relationship between nitrogenase activities associated with the feather moss P. schreberi with temperatures under different moisture levels; n = 270. (E) Multiple linear regression comparing the nitrogenase activities in association with H. splendens with temperatures under different moisture levels; n = 180.

When considering the host species individually, we found a trend for both feather moss species from all sites to increase nitrogenase activity with increasing temperature (Figure S1). Nevertheless, based on fits to the Arrhenius equation, the N2‐fixing communities associated with P. schreberi responded more strongly to temperature than those associated with H. splendens . The P. schreberi communities had an Ea of 77.05 kJ·mol−1 and a Q 10 of 2.8 (95% confidence interval: 2.5–3.2), while H. splendens communities had an Ea of 56.45 kJ·mol−1 and a Q 10 of 2.1 (1.9–2.4) (Figure S2). Further, unlike communities associated with P. schreberi , the N2‐fixing activities of microorganisms in association with H. splendens were not affected by increasing moisture levels (Figure S3). Nevertheless, a significant interaction between temperature and moisture was observed in multiple linear regressions for both host species (Figure 2D,E). H. splendens had higher N2 fixation rates than P. schreberi at 5°C (one‐way ANOVA, p < 0.001, F 1,448 = 102.73) and 15°C (p < 0.001, F 1,448 = 41.11), while the opposite was observed at 25°C (Figure 3A–C). Similarly, H. splendens presented higher nitrogenase activity at 50% (p < 0.001, F 1,448 = 87.58) and 100% moisture (p < 0.001, F 1,447 = 18.97), while P. schreberi had higher N2 fixation rates at 150% moisture (Figure 3D–F).

FIGURE 3.

FIGURE 3

Open in a new tab

Nitrogenase activity associated with the feather moss species P. schreberi and H. splendens after 8 weeks of incubation under temperatures of 5°C (A), 15°C (B), or 25°C (C) or under 50% (D), 100% (E), or 150% moisture (F). Averages and standard errors are represented as bars and vertical lines, respectively, while significant differences in one‐way ANOVA tests according to Tukey's HSD tests are indicated by different letters; n = 180 and 270 for H. splendens and P. schreberi , respectively.

Site of origin was also a factor in how P. schreberi responded to temperature: samples from the site with the lowest precipitation, in ABK, showed only a slight increase in N2 fixation activity with increasing temperatures (linear regression, p = 0.001, r 2  = 0.04), while the samples from the other sites and from H. splendens showed sharper increases (Figure S1). Moisture also had lower influence on N2 fixation associated with P. schreberi samples collected from ABK than on those from the other sites, and no influence on H. splendens samples from any site (Figure S3).

3.2. Diversity of N2 ‐Fixing Communities Per Host and Site

Sequencing of the nifH gene in the feather moss microbiome samples resulted in a total number of 23,528,140 read pairs. Overall, 701 different representative ASVs were identified after filtering, with a mean length of 330 bp (Table S1). Alpha diversity analyses based on the nifH sequences showed that H. splendens samples had a significantly higher diversity for N2‐fixing microorganisms than P. schreberi samples, based on Shannon entropy index (Kruskal–Wallis, p < 0.001, H = 32.38), observed ASVs (p < 0.001, H = 45.19), Pielou's evenness (p < 0.001, H = 13.97), and Faith's phylogenetic diversity (p = 0.042, H = 4.14). Samples from the site with the peak annual precipitation rate, KTJ, had significantly higher Shannon diversity index (p < 0.001, H = 21.65) and Pielou's evenness (p < 0.001, H = 33.27) than those collected in both ABK and LTJ. Statistics on general alpha diversity measures are summarized in Table 1, as well as illustrated in Figures S4–S7. Statistics for pairwise comparisons between treatments are listed in Table S2.

TABLE 1.

Statistical differences in the alpha and beta diversity of microbial communities associated with feather mosses along a precipitation gradient in northern Sweden after incubation under different temperature or moisture conditions. Comparisons were performed using the Kruskal‐Wallis H test across all treatments. Significant differences (p < 0.05) are highlighted in bold. Values are rounded up.

Alpha diversity Shannon entropy Observed features Pielou's evenness Faith's phylogenetic diversity
Factor Group p H p H p H p H
Host species Hylocomium splendens Pleurozium schreberi — 0.001 32.38 0.001 45.19 0.001 13.97 0.042 4.14
Annual precipitation (mm) 300 800 1200 0.001 21.65 0.001 13.99 0.001 33.27 0.035 6.76
Temperature (°C) 5 15 25 0.169 3.57 0.220 3.04 0.349 10.39 0.175 3.49
Moisture (%) 50 100 150 0.161 3.66 0.215 3.08 0.017 8.17 0.003 11.14
Beta diversity Bray–Curtis dissimilarity Jaccard distance Weighted UniFrac Unweighted UniFrac
Factor Group p Pseudo‐F p Pseudo‐F p Pseudo‐F p Pseudo‐F
Host species Hylocomium splendens Pleurozium schreberi — 0.001 11.99 0.001 14.41 0.004 7.04 0.001 7.99
Annual precipitation (mm) 300 800 1200 0.001 19.92 0.001 9.93 0.002 7.13 0.001 4.86
Temperature (°C) 5 15 25 0.001 9.76 0.001 4.12 0.001 17.83 0.001 5.34
Moisture (%) 50 100 150 0.394 1.02 0.018 1.48 0.201 1.56 0.001 3.56
Open in a new tab

Beta diversity analyses also showed that the N2‐fixing communities differed between the feather moss species (Figure 4) in a quantitative sense, as per Bray–Curtis dissimilarity (PERMANOVA, p = 0.001, pseudo‐F = 11.99) and weighted UniFrac (p = 0.004, pseudo‐F = 7.04) distances, as well as a qualitative sense, as per the Jaccard (p = 0.001, pseudo‐F = 14.41) and unweighted UniFrac (p = 0.001, pseudo‐F = 7.99) distance tests. In addition, significant differences between each of the three communities in the ABK, LTJ, and KTJ sites were observed in Bray–Curtis dissimilarity (p = 0.001, pseudo‐F = 19.92), Jaccard distance (p = 0.001, pseudo‐F = 9.93), unweighted UniFrac (p = 0.001, pseudo‐F = 4.86), and weighted UniFrac (p = 0.002, pseudo‐F = 7.13) distance tests. General results for beta diversity PERMANOVA analyses are summarized in Table 1, while statistics for pairwise comparisons between the different treatments are listed in Table S3.

FIGURE 4.

FIGURE 4

Open in a new tab

Principal component analyses of beta diversity analyses of the N2‐fixing communities associated with feather mosses across a precipitation gradient in northern Sweden. (A) Bray–Curtis dissimilarities illustrating the dissimilarities in ASV abundance between samples; n = 74. (B) Jaccard distances depicting how community composition differs between samples, based on ASV presence or absence; n = 74.

3.3. Differences in Diversity Among Incubation Regimes

Based on the nifH analyses, incubations under different temperatures did not lead to significant changes in the alpha diversity of the moss‐associated N2‐fixing communities (Table 1). However, dissimilarities between the communities in all temperature treatments were observed in beta diversity analyses in both qualitative (p = 0.001, pseudo‐F = 4.12 and p = 0.001, pseudo‐F = 5.34 for Jaccard distance and unweighted UniFrac distances, respectively) and quantitative (p < 0.001, pseudo‐F = 9.76 and p = 0.001, pseudo‐F = 17.83 for Bray–Curtis dissimilarity and weighted UniFrac distances) metrics (Table 1, Table S3). Similarly, different moisture treatments did not result in significant differences in Shannon entropy and observed ASVs, although evenness and phylogenetic diversity did significantly decrease in the 150% moisture samples (p = 0.017, H = 8.17 and p = 0.003, H = 11.14, respectively) (Table 1). Dissimilarities in N2‐fixing communities exposed to different moisture levels were revealed by beta diversity analyses in qualitative terms only, as indicated by Jaccard distance (p = 0.001, pseudo‐F = 2.47 and p = 0.006, pseudo‐F = 1.66, respectively) and unweighted UniFrac (p = 0.001, pseudo‐F = 4.82 and p = 0.002, pseudo‐F = 4.60) distances (Table 1).

3.4. Differences in N2 ‐Fixing Community Composition

Sequencing of the nifH gene showed that, overall, cyanobacteria were the dominant members of the N2‐fixing communities evaluated, with an average relative abundance of 62.2 (±1.7) % across all samples (Figure 5). The second most abundant group of N2 fixers in the samples was unclassified bacteria (those not presenting any significant similarity with bacterial references in the nifH database used), representing an average of 27.9 (±1.6) % of the overall community (Figure 5).

FIGURE 5.

FIGURE 5

Open in a new tab

Relative abundances of ASVs for N2‐fixing bacteria and archaea associated with subarctic feather mosses in relation to different temperatures (A), moisture levels (B), or hosts (C) based on nifH gene sequence analyses; n = 74. ASVs for which no significant similarity at the taxonomic level of order or above was found are noted as “unclassified” at their lowest level identified, while ASVs which could not be determined to represent bacteria or archaea are noted as “unidentified.” Sequences related to eukaryotic organisms were removed before plotting.

The relative abundance of cyanobacterial amplicon sequence variants (ASVs) differed significantly between H. splendens , with an average of 69.1 (±2.7) %, and P. schreberi , with 57.7 (±2.9) % (one‐way ANOVA, p < 0.001, F 1,220 = 11.48). These differences were not observed between the different sites across the precipitation gradient, nor between incubations under different moisture levels. However, incubations at 25°C did lead to a significant decrease in the relative abundance of cyanobacteria (47.0% ± 2.4%) in comparison with incubations at 15 (66.4% ± 2.8%) and 5 (73.2% ± 2.8%)°C (one‐way ANOVA, p < 0.001, F 2,219 = 27.77). On the other hand, the incubation at 25°C significantly increased the relative abundance of unclassified bacteria (38.6% ± 2.6%) in comparison with samples incubated at 15°C (25.9% ± 2.7%) and 5°C (19.2% ± 2.5%) (p < 0.001, F 1,219 = 14.88). While there was no difference between the relative abundance of unclassified bacteria in the different hosts, the site with the lowest annual precipitation rates (ABK) also had a significantly lower relative abundance of these bacteria (22.2% ± 2.6% compared to 33.2% ± 2.5% and 30.8% ± 2.4% in KTJ and LTJ, respectively) (p = 0.004, F 1,219 = 5.61).

3.5. nifH Gene Quantification

As expected, there was a positive relationship between the number of copies of the nifH gene in the different samples as measured by qPCR and their overall N2 fixation rates observed in acetylene reduction assays, although the analysis had low explanatory power (linear regression, p < 0.001, r 2  = 0.06) (Figure 6A). The number of nifH gene copies in the samples originating in different sites along the precipitation gradient also followed a similar pattern as the one observed for N2 fixation, but only in association with P. schreberi shoots (Figure 6B,C). P. schreberi samples from ABK presented an average number of copies (3.2 × 104) significantly lower than the ones from the LTJ (1.2 × 105) and KTJ (1.8 × 105) sites, which did not statistically differ from one another (one‐way ANOVA, p < 0.001, F 2,130 = 8.43) (Figure 7). The number of nifH copies identified as cyanobacteria and unclassified bacteria also increased along the precipitation gradient for P. schreberi samples, while nifH copies from unclassified bacteria decreased significantly in H. splendens communities from the site with the highest annual precipitation, KTJ (Figure S8). On the other hand, incubation at 25°C tended to decrease the number of cyanobacterial nifH copies (Figure 7A,C), while incubation at 150% moisture tended to increase their number (Figure 7B,D) in both host species. The different temperature and moisture treatments did not significantly affect the number of nifH gene copies for unclassified bacteria.

FIGURE 6.

FIGURE 6

Open in a new tab

Copies of the nitrogenase gene nifH in microbial communities associated with feather mosses from Swedish subarctic tundra as quantified by qPCR. (A) Relationship between nifH copies and nitrogenase activity based on a linear regression analysis; n = 225. (B) Number of nifH copies expressed in microbial communities associated with P. schreberi from three sites forming a precipitation gradient; n = 45. (C) Differences in the number of copies of the nifH gene expressed in communities associated with H. splendens samples collected in sites with different annual precipitation rates; n = 45.

FIGURE 7.

FIGURE 7

Open in a new tab

Number of nifH gene copies for N2‐fixing cyanobacteria associated with feather mosses from northern Sweden. (A) Effects of different temperatures on the expression of the nifH gene by cyanobacteria associated with P. schreberi . (B) Number of cyanobacteria nifH gene copies expressed in association with P. schreberi under different moisture levels. (C) Number of cyanobacterial nifH gene copies expressed in association with H. splendens incubated under different temperatures. (D) Variation in the number of copies of the nifH gene from N2‐fixing cyanobacteria associated with H. splendens under different moisture regimes. Bars and vertical lines represent averages and standard errors, respectively. Treatments receiving different letters were significantly different in one‐way ANOVA tests according to Tukey's HSD post hoc analyses; n = 45.

4. Discussion

Our first hypothesis, which anticipated that mosses collected from sites with higher annual precipitation in the Swedish subarctic would exhibit higher N2 fixation rates, was supported by our results (Figure 2A). Since moisture was normalized between samples before the experiments, the observed effect was unlikely to be due to a carryover of water content from field conditions and its physiological effects in the feather mosses. More likely, it was a direct result of the lasting influence of their environments of origin on the microbial communities associated with the feather moss samples.

Furthermore, nifH gene copy numbers and nitrogenase activity in P. schreberi samples from the driest site, ABK, showed only slight increases in response to higher temperature and moisture, despite being more sensitive to temperature than H. splendens samples (Figures S1–S3). This suggests that the legacy of extended exposure to field factors continues to influence the N2‐fixing P. schreberi communities even when conditions change. In contrast, the H. splendens ‐associated communities from different sites showed consistently higher nitrogenase activity with greater relative abundance of cyanobacterial ASVs and little variation in nifH gene copies, suggesting more stable communities requiring a lower level of moisture for their activities, not benefiting from extra water. In addition, alpha diversity was highest in association with mosses from the site with the highest annual precipitation (KTJ). This suggests that, overall, precipitation favors the diversity and N2‐fixing activities of feather moss microbiomes in the subarctic.

Environmental conditions were often shown to directly impact the microbiomes of vascular plants, creating long‐lasting effects on the composition and function of their microbial communities (Trivedi et al. 2022). Our results indicate that environmental factors also have profound effects on microbial communities associated with non‐vascular plants. Considering that mosses display highly plastic phenotypes that respond strongly to environmental fluctuations (Mohanasundaram and Pandey 2022) and that plants and their microbiomes function as a holobiont (Vandenkoornhuyse et al. 2015), it is not surprising that moss microbiomes are also significantly impacted by environmental factors. Furthermore, while vascular plants have a more complex internal anatomy composed of conducting tissues and highly compartmentalized organs, non‐vascular plants like mosses lack these internal structures. Consequently, mosses provide fewer internal colonization sites, leading most of their associated microorganisms to live on the external surfaces of their hosts (Alvarenga and Rousk 2022). This may result in moss‐associated microbial communities being more exposed to environmental pressures. Our findings also suggest that as climate change alters temperature and precipitation patterns in the subarctic, it will probably leave imprints on moss microbiomes, influencing their growth and ecological functions.

Overall, our second hypothesis, which stated that the N2 fixation rates associated with the feather moss samples would be positively influenced by temperature and moisture, was overall confirmed (Figure 2). However, there were noticeable differences between the communities associated with different hosts. Sites with higher moisture did seem to support greater activity by N2 fixers in P. schreberi , especially cyanobacteria and unclassified bacteria (Figure 6). Nevertheless, when H. splendens samples were evaluated separately, increasing moisture levels did not lead to higher nitrogenase activity or variation in the structure of its associated communities (Figure S2D,E). This apparent contradiction is unexpected, but could be partially explained by the differences in the microbial communities between the moss species (Figure 4). One possibility is that H. splendens hosts more resilient N2‐fixing microorganisms than P. schreberi , whose N2‐fixing communities have a much stronger reaction to temperature and moisture. It is also possible that the hosts themselves react differently to changes in temperature and moisture and, as a consequence, influence their microbiomes in diverging ways.

Not much is known about the microbial diversity associated with mosses across different ecosystems. Unfortunately, moss microbiomes are still overlooked in ecological studies, and therefore the identity of a significant percentage of the members of these communities remains unknown. The unclassified bacteria detected using the nifH marker gene in this study are potentially novel. Future work should focus on identifying these N2 fixers with analyses based on whole‐metagenome shotgun sequencing, since metagenome‐assembled genomes could anchor nifH gene sequences together with molecular markers that do enable their correct taxonomic placement. Nevertheless, considerable differences between the composition of bacterial communities in the microbiomes of different hosts have been observed, possibly resulting from co‐diversification or ecological filtering driven by moss traits (Klarenberg et al. 2022; Holland‐Moritz et al. 2018). Plant species and genotypes are directly related to the different effects microbiomes and their hosts have on each other (van Rensburg et al. 2024), potentially influencing microbial responses to the environment in unique ways. Therefore, despite occupying similar niches in the subarctic tundra and boreal forests that sometimes even lead the two moss species to physically grow together, these mosses have significantly distinct microbial communities that react differently to their environments.

Finally, as predicted by our third hypothesis, we found a significant relationship between temperature and moisture levels in relation to cyanobacterial ASVs (Figure 6). In general, the effects of temperature on the N2‐fixing activities of cyanobacteria seem to be dependent on their environment and habitat. For example, N2‐fixing cyanobacteria from aquatic environments tend to grow and perform better in temperatures of 25°C or a bit higher (Thomas and Litchman 2016), while the optimal temperature for the N2‐fixing activities of cyanobacteria in boreal forests may vary between 16°C and 27°C (Rousk, Jones, and DeLuca 2017; Rousk, Pedersen, et al. 2017). Some studies have found that arctic and subarctic cyanobacteria in various habitats have a comparable temperature optimum for N2 fixation around 25°C and 32°C (Zielke et al. 2002; Salazar et al. 2022), while another study observed higher N2‐fixing activities in temperatures under 14°C (Rousk et al. 2018). Our results provide additional evidence that some cyanobacteria inhabiting tundra ecosystems may fix more N2 under colder temperatures.

Virtually all the N fixed in moss‐cyanobacteria symbioses can be retained within host tissues in the short term, but precipitation can cause it to significantly leach to the ground, depending on the host species (Song et al. 2024). This process is known to introduce a significant amount of N into arctic and subarctic soils that can be taken up by other organisms, allowing them to bypass the N limitations that are common in these ecosystems (Elser et al. 2007). Based on the results we obtained, however, climate change could decrease the abundance of cyanobacterial ASVs and nifH copies in these communities (Figure 5) as it leads to not only enhanced shrub growth outcompeting mosses (Elmendorf et al. 2012) but also to a warmer and more drought‐prone Arctic.

This may seem paradoxical, as we observed that, overall, higher temperatures tend to lead to increased nitrogenase activity (Figures 2, 3, 4, Figure S8). This result was apparently driven by the rise in the abundance of ASVs belonging to N2‐fixing bacteria from currently unclassified taxa (Figure 5), which was also observed by Holland‐Moritz et al. (2018). Similarly, Klarenberg et al. (2022) have observed the partial replacement of N2‐fixing cyanobacteria with other diazotrophs in the microbiome of the subarctic moss Racomitrium lanuginosum , although followed by a decrease in the number of overall nifH gene copies in the community. We observed stable nifH copy numbers and nitrogenase activity from unclassified bacteria across incubation temperatures despite the change in taxa composition and a drop in cyanobacterial nifH gene copies, suggesting that unclassified bacterial taxa may take over the N2‐fixing role from cyanobacteria without compromising overall nitrogenase activity, thus allowing these associations to remain important players in the biogeochemical cycle of N in a changing Arctic. Driven by increased temperature, this replacement rate may accelerate in the near future due to an increase in the frequency and intensity of hot weather events, as witnessed by the 38°C recently measured in the Arctic region (World Meteorological Organization 2021), which is well above the optimal temperatures for N2 fixation ever estimated for cyanobacteria from any habitat. On the other hand, this replacement effect seems to be host‐dependent, with H. splendens showing relatively more stable N2‐fixing communities that are nonetheless more sensitive to temperature in comparison to those in association with P. schreberi .

As climate change increases temperatures and the frequency of drought events in the Arctic, the contribution of moss‐associated cyanobacteria to N2 fixation in the tundra will likely decrease due to its susceptibility to low precipitation. Even though different moss species harbor distinct N2‐fixing communities, they are all affected by climate change, although not in the same manner. Our findings suggest that currently unclassified bacteria may take over the role of N2 fixation from cyanobacteria in feather moss microbiomes, potentially compensating for the decline in cyanobacterial activity. While this shift could help maintain or even enhance N2 fixation rates in these communities, the broader ecological consequences of this microbial turnover remain uncertain.

Author Contributions

Danillo O. Alvarenga: data curation, formal analysis, investigation, software, validation, visualization, writing – original draft, writing – review and editing. Justin T. Wynns: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft, writing – review and editing. Joseph Nesme: data curation, formal analysis, software, validation, writing – review and editing. Anders Priemé: data curation, supervision, validation, writing – review and editing. Kathrin Rousk: conceptualization, supervision, data curation, funding acquisition, investigation, methodology, project administration, resources, writing – original draft, writing – review and editing.

Funding

This work was funded by the European Research Council's Horizon 2020 research and innovation program (Grant #947719 to K.R.), with further support from the Independent Research Fund Denmark (IRFD; #6108‐00089 to K.R.). The Danish National Research Foundation supported activities within the Center for Volatile Interactions (VOLT, #DNRF168). The analyses in this work were also funded by the Danish e‐Infrastructure Consortium (Grants #DeiC‐AAU‐N1‐2024087 and #DeiC‐KU‐N3‐2024088 to D.O.A.).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figures S1–S8: gcb70718‐sup‐0001‐FigureS1‐S8.pdf.

GCB-32-e70718-s002.pdf (1.5MB, pdf)

Table S1: gcb70718‐sup‐0002‐TableS1.xls.

GCB-32-e70718-s003.xls (316.5KB, xls)

Table S2: gcb70718‐sup‐0003‐TableS2.xls.

GCB-32-e70718-s004.xls (11.5KB, xls)

Table S3: gcb70718‐sup‐0004‐TableS3.xls.

GCB-32-e70718-s001.xls (11.5KB, xls)

Acknowledgments

We would like to thank Johannes Rousk and Anders Michelsen for helping with the sampling, and Pia A. Pedersen for providing the plasmids used for producing the qPCR standards. We are also grateful to Michael Westbury, Julian Regalado Perez, and Jakob Russel for their valuable help during the early stages of this work.

Alvarenga, D. O. , Wynns J. T., Nesme J., Priemé A., and Rousk K.. 2026. “Climate Change Impacts the Structure and Nitrogen‐Fixing Activities of Subarctic Feather Moss Microbiomes Across a Precipitation Gradient.” Global Change Biology 32, no. 2: e70718. 10.1111/gcb.70718.

Data Availability Statement

The raw sequences obtained in this work are available in National Center for Biotechnology Information's Sequence Read Archive database under the accession number PRJNA1277518. The ASVs were deposited in NCBI's GenBank database under the accession number PX342784–PX343484. The rest of the data and scripts to analyze them were deposited in the Electronic Research Data Archive under the DOI https://doi.org/10.17894/ucph.7700e7d3‐64c9‐43a1‐863d‐44f29176962f.

References

  1. Alvarenga, D. O. , and Rousk K.. 2021. “Indirect Effects of Climate Change Inhibit N2 Fixation Associated With the Feathermoss Hylocomium splendens in Subarctic Tundra.” Science of the Total Environment 795: 148676. [DOI] [PubMed] [Google Scholar]
  2. Alvarenga, D. O. , and Rousk K.. 2022. “Unraveling Host‐Microbe Interactions and Ecosystem Functions in Moss‐Bacteria Symbioses.” Journal of Experimental Botany 73: 4473–4486. [DOI] [PubMed] [Google Scholar]
  3. Alvarenga, D. O. , Wynns J. T., Nesme J., Priemé A., and Rousk K.. 2026. “Climate Change Impacts the Structure and Nitrogen‐Fixing Activities of Subarctic Feather Moss Microbiomes Across a Precipitation Gradient—Dataset.” Electronic Research Data Archive: 2026‐01. 10.17894/ucph.7700e7d3-64c9-43a1-863d-44f29176962f. [DOI] [PMC free article] [PubMed]
  4. Angel, R. , Nepel M., Panhölzl C., et al. 2018. “Evaluation of Primers Targeting the Diazotroph Functional Gene and Development of NifMAP – A Bioinformatics Pipeline for Analyziing nifH Amplicon Data.” Frontiers in Microbiology 9: 703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ballinger, T. J. , Bigalke S., Walsh J. E., et al. 2024. “Surface Air Temperature.” In Arctic Report Card, edited by Thoman R. L., Moon T. A., and Druckenmiller M. L.. National Oceanic and Atmospheric Administration. [Google Scholar]
  6. Bolyen, E. , Rideout J. R., Dillon M. R., et al. 2019. “Reproducible, Interactive, Scalable and Extensible Microbiome Data Science Using QIIME 2.” Nature Biotechnology 37: 852–857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bustin, S. A. , Benes V., Garson J. A., et al. 2009. “The MIQE Guidelines: Minimum Information for Publication of Quantitative Real‐Time PCR Experiments.” Clinical Chemistry 55: 611–622. [DOI] [PubMed] [Google Scholar]
  8. Cole, J. R. , Wang Q., Fish J. A., et al. 2014. “Ribosomal Database Project: Data and Tools for Hight Throughput rRNA Analysis.” Nucleic Acids Research 42: D633–D642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dai, A. , Luo D., Song M., and Liu J.. 2019. “Arctic Amplification Is Caused by Sea‐Ice Loss Under Increasing CO2 .” Nature Communications 10: 121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Davidson, E. A. , and Janssens I. A.. 2006. “Temperature Sensitivity of Soil Carbon Decomposition and Feedbacks to Climate Change.” Nature 440: 165–173. [DOI] [PubMed] [Google Scholar]
  11. Deckmyn, A. 2021. “Maps.” Accessed December 14, 2021. https://github.com/adeckmyn/maps.
  12. Douma, J. C. , Van Wijk T., Lang S. I., and Shaver G. R.. 2007. “The Contribution of Mosses to the Carbon and Water Exchange of Arctic Ecosystems: Quantification and Relatioships With System Properties.” Plant, Cell & Environment 30: 1205–1215. [DOI] [PubMed] [Google Scholar]
  13. Dunnington, D. 2021. “ggspatial.” Accessed December 14, 2021. https://paleolimbot.github.io/ggspatial/.
  14. Eldridge, D. J. , Guirado E., Reich P. B., et al. 2023. “The Global Contribution of Soil Mosses to Ecosystem Services.” Nature Geoscience 16: 430–438. [Google Scholar]
  15. Elmendorf, S. C. , Henry G. H. G., Hollister R. D., et al. 2012. “Global Assessment of Experimental Climate Warming on Tundra Vegetation: Heterogeneity Over Space and Time.” Ecology Letters 15: 164–175. [DOI] [PubMed] [Google Scholar]
  16. Elser, J. J. , Bracken M. E. S., Cleland E. E., et al. 2007. “Global Analysis of Nitrogen and Phosphorus Limitation of Primary Producers in Freshwater, Marine and Terrestrial Ecosystems.” Ecology Letters 10: 1135–1142. [DOI] [PubMed] [Google Scholar]
  17. Finger Higgens, R. A. , Chipman J. W., Lutz D. A., Culler L. E., Virginia R. A., and Ogden L. A.. 2019. “Changing Lake Dynamics Indicate a Drier Arctic in Western Greenland.” Journal of Geophysical Research: Biogeosciences 124: 870–883. [Google Scholar]
  18. Fisher, R. A. 1918. “The Correlation Between Relatives on the Supposition of Mendelian Inheritance.” Transactions of the Royal Society of Edinburgh 52: 399–433. [Google Scholar]
  19. Fox, J. , and Weisberg S.. 2019. An R Companion to Applied Regression. Sage. [Google Scholar]
  20. Galton, F. 1886. “Regression Towards Mediocrity in Hereditary Stature.” Journal of the Anthropological Institute of Great Britain and Ireland 15: 246–263. [Google Scholar]
  21. Gupta, R. 2012. “Origin and Spread of Photosynthesis Based Upon Conserved Sequence Features in Key Bacteriochlorophyll Biosynthesis Proteins.” Molecular Biology and Evolution 29: 3397–3412. [DOI] [PubMed] [Google Scholar]
  22. Hardy, R. W. F. , and Knight E.. 1967. “ATP‐Dependent Reduction of Azide and HCN by N2‐Fixing Enzymes of Azotobacter vinelandii and Clostridium pasteurianum .” Biochimica et Biophysica Acta 139: 69–90. [DOI] [PubMed] [Google Scholar]
  23. Heller, P. , Tripp H. J., Turk‐Kubo K., and Zehr J. P.. 2014. “ARBitrator: A Software Pipeline for On‐Demand Retrieval of Auto‐Curate nifH Sequences From GenBank.” Bioinformatics 30: 2883–2890. [DOI] [PubMed] [Google Scholar]
  24. Holland‐Moritz, H. , Stuart J., Lewis L. R., et al. 2018. “Novel Bacterial Lineages Associated With Boreal Moss Species.” Environmental Microbiology 20: 2625–2638. [DOI] [PubMed] [Google Scholar]
  25. Holland‐Moritz, H. , Stuart J. E. M., Lewis L. R., et al. 2021. “The Bacterial Communities of Alaskan Mosses and Their Contributions to N2‐Fixation.” Microbiome 9: 53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Illumina . 2013. 16S Metagenomic Sequencing Library Preparation. Illumina. [Google Scholar]
  27. Katoh, K. , and Standley D. M.. 2013. “MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability.” Molecular Biology and Evolution 30: 772–780. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Klarenberg, I. J. , Keuschnig C., Colmenares A. J. R., et al. 2022. “Long‐Term Warming Effects on the Microbiome and nifH Gene Abundance of a Common Moss Species in Sub‐Arctic Tundra.” New Phytologist 234: 2044–2056. [DOI] [PubMed] [Google Scholar]
  29. Kvíderová, J. , Elster J., and Komárek J.. 2019. “Ecophysiology of Cyanobacteria in the Polar Regions.” In Cyanobacteria: From Basic Science to Applications, edited by Mishra A. K. and Tiwari D. N., 277–302. Academic Press. [Google Scholar]
  30. Lett, S. , Christiansen C. T., Dorrepaal E., et al. 2024. “Moss Species and Precipitation Mediate Experimental Warming Stimulation of Growing Season N2 Fixation in Subarctic Tundra.” Global Change Biology 30: e17401. [DOI] [PubMed] [Google Scholar]
  31. Lett, S. , and Michelsen A.. 2014. “Seasonal Variation in Nitrogen Fixation and Effects of Climate Change in a Subarctic Heath.” Plant and Soil 379: 193–204. [Google Scholar]
  32. Lüdecke, D. 2018. “Ggeffects: Tidy Data Frames of Marginal Effects From Regression Models.” Journal of Open Source Software 3: 772. [Google Scholar]
  33. Martínez‐García, E. , Nilsson M. B., Laudon H., et al. 2024. “Drought Response of the Boreal Forest Carbon Sink Is Driven by Understorey‐Tree Composition.” Nature Geoscience 17: 197–204. [Google Scholar]
  34. Mohanasundaram, B. , and Pandey S.. 2022. “Effects of Environmental Signals on Growth and Development in Mosses.” Journal of Experimental Botany 73: 4514–4527. [DOI] [PubMed] [Google Scholar]
  35. Moon, K. W. 2021. “ggiraphExtra.” Accessed December 14, 2021. https://github.com/cardiomoon/ggiraphExtra/.
  36. Moynihan, M. A. , and Furbo Reeder C.. 2025. “nifHdada2 Repository, v2.0.5.” Accessed 13 June 2025. 10.5281/zenodo.7996213. [DOI]
  37. Pedersen, P. A. 2017. “The Effect of Freeze‐Thaw Cycles on Nitrogen Fixation in Moss‐Associated Cyanobacteria.” Copenhagen University of Copenhagen.
  38. Pedersen, T. L. 2021. “patchwork.” Accessed December 14, 2021. https://patchwork.data‐imaginist.com/.
  39. Permin, A. , Michelsen A., and Rousk K.. 2022. “Direct and Indirect Effects of Warming on Moss Abundance and Associated Nitrogen Fixation in Subarctic Ecosystems.” Plant and Soil 471: 343–358. [Google Scholar]
  40. Posit Team . 2022. “RStudio: Integrated Development Environment for R.” Accessed December 14, 2022. http://www.posit.co/.
  41. Price, M. N. , Dehal P. S., and Arkin A. P.. 2010. “FastTree 2—Approximately Maximum‐Likelihood Trees for Large Alignments.” PLoS One 5: e9490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. R Core Team . 2022. “R: A Language and Environment for Statistical Computing.” Accessed October 31, 2022. https://www.R‐project.org/.
  43. Rantanen, M. , Karpechko A. Y., Lipponen A., et al. 2022. “The Arctic Has Warmed Nearly Four Times Faster Than the Globe Since 1979.” Communications Earth & Environment 3: 168. [Google Scholar]
  44. Rousk, K. 2022. “Biotic and Abiotic Controls of Nitrogen Fixation in Cyanobacteria–Moss Associations.” New Phytologist 235: 1330–1335. [DOI] [PubMed] [Google Scholar]
  45. Rousk, K. , Jones D. L., and DeLuca T. H.. 2017. “Moss‐Cyanobacteria Associations as Biogenic Sources of Nitrogen in Boreal Forest Ecosystems.” Frontiers in Microbiology 4: 150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Rousk, K. , and Michelsen A.. 2017. “Ecosystem Nitrogen Fixation Throughout the Snow‐Free Period in Subarctic Tundra: Effects of Willow and Birch Litter Addition and Warming.” Global Change Biology 23: 1552–1563. [DOI] [PubMed] [Google Scholar]
  47. Rousk, K. , Pedersen P. A., Dyrnum K., and Michelsen A.. 2017. “The Interactive Effects of Temperature and Moisture on Nitrogen Fixation in Two Temperate‐Arctic Mosses.” Theoretical and Experimental Plant Physiology 29: 25–36. [Google Scholar]
  48. Rousk, K. , Sorensen P. L., and Michelsen A.. 2018. “What Drives Biological Nitrogen Fixation in High Arctic Tundra: Moisture or Temperature?” Ecosphere 9: e02117. [Google Scholar]
  49. Salazar, A. , Warshan D., Vasquez‐Mejia C., and Andrésson O. S.. 2022. “Environmental Change Alters Nitrogen Fixation Rates and Microbial Parameters in a Subarctic Biological Soil Crust.” Oikos 11: e09239. [Google Scholar]
  50. Selzer, L. 2021. “multcompView.” Accessed December 14, 2021. https://github.com/lselzer/multcompview.
  51. Shapiro, S. S. , and Wilk M. B.. 1965. “An Analysis of Variance Test for Normality (Complete Samples).” Biometrika 52: 591–611. [Google Scholar]
  52. Slate, M. L. , Antoninka A., Bailey L., et al. 2024. “Impact of Changing Climate on Bryophyte Contributions to Terrestrial Water, Carbon, and Nitrogen Cycles.” New Phytologist 242: 2411–2429. [DOI] [PubMed] [Google Scholar]
  53. Solheim, B. , Wiggen H., Røberg S., and Spaink H. P.. 2004. “Associations Between Arctic Cyanobacteria and Mosses.” Symbiosis 37: 169–187. [Google Scholar]
  54. Song, G. , Clasen L. A., and Rousk K.. 2024. “Short‐Term Fate of Nitrogen Fixed by Moss‐Cyanobacteria Associations Under Different Rainfall Regimes.” Basic and Applied Ecology 79: 9–16. [Google Scholar]
  55. Sorensen, P. L. , Lett S., and Michelsen A.. 2012. “Moss‐Specific Changes in Nitrogen Fixation Following Two Decades of Warming, Shading, and Fertilizer Addition.” Plant Ecology 213: 695–706. [Google Scholar]
  56. Sorensen, P. L. , and Michelsen A.. 2011. “Long‐Term Warming and Litter Addition Affects Nitrogen Fixation in a Subarctic Heath.” Global Change Biology 17: 528–537. [Google Scholar]
  57. Street, L. E. , Subke J. A., Sommerkorn M., et al. 2013. “The Role of Mosses in Carbon Uptake and Partitioning in Arctic Vegetation.” New Phytologist 199: 163–175. [DOI] [PubMed] [Google Scholar]
  58. Swedish Meterological and Hydrological Institute . 2023. “SMHI.” Accessed June 13, 2023. https://www.smhi.se/.
  59. Tahon, G. , Tytgat B., and Willems A.. 2016. “Diversity of Photrophic Genes Suggests Multiple Bacteria May Be Able to Exploit Sunlight in Exposed Soils From the Sør Rondane Mountains, East Antarctica.” Frontiers in Microbiology 7: 2026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Thomas, M. K. , and Litchman E.. 2016. “Effects of Temperature and Nitrogen Availability on the Growth of Invasive and Native Cyanobacteria.” Hydrobiologia 763: 357–369. [Google Scholar]
  61. Trivedi, P. , Batista B. D., Bazany K. E., and Singh B. K.. 2022. “Plant–Microbiome Interactions Under a Changing World: Responses, Consequences and Perspectives.” New Phytologist 234: 1951–1959. [DOI] [PubMed] [Google Scholar]
  62. Tukey, J. W. 1949. “Comparing Individual Means in the Analysis of Variance.” Biometrics 5: 99–114. [PubMed] [Google Scholar]
  63. Ueda, T. , Suga Y., Yahiro N., and Matsuguchi T.. 1995. “Remarkable N2‐Fixing Bacterial Diversity Detected in Rice Roots by Molecular Evolutionary Analysis of nifH Gene Sequences.” Journal of Bacteriology 177: 1414–1417. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. van Rensburg, H. J. , Stengele K., and Schlaeppi K.. 2024. “Understanding Plant Responsiveness to Microbiome Feedbacks.” Current Opinion in Plant Biology 81: 102603. [DOI] [PubMed] [Google Scholar]
  65. Vandenkoornhuyse, P. , Quaiser A., Duhamel M., Le Van A., and Dufresne A.. 2015. “The Importance of the Microbiome of the Plant Holobiont.” New Phytologist 206: 1196–1206. [DOI] [PubMed] [Google Scholar]
  66. Wickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer‐Verlag. [Google Scholar]
  67. Wickham, H. , Pedersen T. L., and Seidel D.. 2021. “Scale Functions for Visualization.” Accessed December 14, 2021. https://scales.r‐lib.org/.
  68. Wilke, C. O. 2021. “cowplot – Streamlined Plot Theme and Plot Annotations for ggplot2.” Accessed December 14, 2021. https://wilkelab.org/cowplot/.
  69. World Meteorological Organization . 2021. “WMO Recognizes New Arctic Temperature Record of 38°C.” Accessed December 14, 2021. https://wmo.int/media/news/wmo‐recognizes‐new‐arctic‐temperature‐record‐of‐380c.
  70. Xie, A. , Zhu J., Kang S., Qin X., Xu B., and Wang Y.. 2022. “Polar Amplification Comparison Among Earth's Three Poles Under Different Socioeconomic Scenarios From CMIP6 Surface Air Temperature.” Scientific Reports 12: 16548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Yeo, I. K. , and Johnson R. A.. 2000. “A New Family of Power Transformations to Improve Normality or Symmetry.” Biometrika 87: 954–959. [Google Scholar]
  72. Zielke, M. , Ekker A. S., Olsen R. A., Spjelkavik S., and Solheim B.. 2002. “The Influence of Abiotic Factors on Biological Nitrogen Fixation in Different Types of Vegetation in the High Arctic, Svalbard.” Arctic, Antarctic, and Alpine Research 34: 293–299. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figures S1–S8: gcb70718‐sup‐0001‐FigureS1‐S8.pdf.

GCB-32-e70718-s002.pdf (1.5MB, pdf)

Table S1: gcb70718‐sup‐0002‐TableS1.xls.

GCB-32-e70718-s003.xls (316.5KB, xls)

Table S2: gcb70718‐sup‐0003‐TableS2.xls.

GCB-32-e70718-s004.xls (11.5KB, xls)

Table S3: gcb70718‐sup‐0004‐TableS3.xls.

GCB-32-e70718-s001.xls (11.5KB, xls)

Data Availability Statement

The raw sequences obtained in this work are available in National Center for Biotechnology Information's Sequence Read Archive database under the accession number PRJNA1277518. The ASVs were deposited in NCBI's GenBank database under the accession number PX342784–PX343484. The rest of the data and scripts to analyze them were deposited in the Electronic Research Data Archive under the DOI https://doi.org/10.17894/ucph.7700e7d3‐64c9‐43a1‐863d‐44f29176962f.

Articles from Global Change Biology are provided here courtesy of Wiley

RESOURCES