Climate change is driving permafrost thaw, releasing previously frozen resources, such as nitrogen, to the soil active layer. In low-nitrogen systems, like boreal peatlands, this novel nitrogen source may benefit plant productivity. However, other resource limitations (e.g., light) may limit plant access to thaw-front nitrogen. We used a stable isotope experiment to explore variations in understory boreal plant species’ ability to take up different forms of newly released nitrogen from permafrost thaw under different canopy covers. This experiment occurred in a peatland in the sporadic discontinuous permafrost zone of the Northwest Territories, Canada. We added 15N labelled ammonium, nitrate, and the amino acid glycine at the thaw front (40 cm depth) at two sites differentiated by high and low canopy cover and determined uptake of 15N in leaves of several common and abundant boreal plant species. At each site we had 6 quadrats that were 2 m by 2 m and applied one of the four treatments (control (no addition), glycine, ammonium, nitrate) in each corner. This data set provides nitrogen permil values, percent nitrogen, and foliage dry weight analyzed for nitrogen. Foliage was collected one year after 15N addition. We found that the probability of plant uptake of thaw-front nitrogen was significantly greater at low canopy cover sites; however, nitrogen form, plant species, and foliar Nmass had no effect. We further found that Rubus chamaemorus had the highest foliar Nmass followed by Rhododendron groenlandicum, Chamaedaphne calyculata, and Vaccinium vitis-idea. Our results demonstrate that access to nitrogen released from permafrost thaw by boreal plants may be mediated by light availability. Understanding the variation in site response to permafrost thaw contributes to our understanding of how boreal peatlands will change with ongoing climate change.

This is the final dataset for the publication <a href="https://doi.org/10.1007/s10021-023-00899-1">"Response of boreal plant communities and forest floor carbon fluxes to experimental nutrient additions"</a> published in the peer-reviewed journal <i>Ecosystems</i> in 2024. This dataset includes three primary files:<ol><p><li>the plant community composition data</li><li>the plant functional trait data</li><li>the forest floor carbon flux and associated environmental variables data.</li></ol>Combined, the plant community composition data and plant functional trait data were used to calculate the community-weighted mean traits presented in the manuscript. These data explored changes in community-weighted mean traits of both moss and understory vascular plant communities as well as forest floor carbon flux with nutrient fertilization emulating nutrient increases following permafrost thaw in a boreal peatland (Scotty Creek Research Station, NT: 61°18′ N, 121°,18′ W). We had two sites in this boreal peatland: one with high canopy cover, and one with low canopy cover. <p><p>We conducted an experiment whereby we set up 25, 0.5 m^2 quadrats at two sites within a boreal peatland (50 total quadrats). We had five different nutrient treatments with five replicates each: an unmanipulated control (no fertilizer or mechanical disturbance), a manipulated control (no fertilizer, with mechanical disturbance similar to the fertilized treatments), shallow fertilization (nutrient added at 20cm soil depth emulating increasing microbial mineralization with warming soil), deep fertilization (nutrient added at 40 cm soil depth emulating release of previously frozen nutrient-rich soil to the active layer), and shallow + deep fertilization (emulating both the shallow and deep treatments simultaneously). 12-4-8 NPK MiracleGro slow-release fertilizer was applied at a dosage of 6 g N m-2. This value is roughly four times that of the predicted maximum rate of increase in N from rising soil temperature and permafrost thaw (e.g., 1.3 g N m-2 yr-1; Keuper and others 2012, <i>Global Change Biology</i> 18: 1998) but is similar to Keuper and others (2017, <i>Global Change Biology</i> 23: 4257). Furthermore, being slow-release fertilizer, we expected slow nutrient release over the two-year experiment, mimicking natural conditions. The two sites whereby we set up the experiment differed in canopy cover and therefore allowed us to test if light availability effected the response of carbon fluxes and plant communities to nutrient increases. See (link to paper) for more detailed methodology.

This is the final dataset associated with the publication, “Greater variation in boreal plant community composition and community-level traits at local- than regional-scale”, published in the <i>Journal of Vegetation Science</i> (2023). This dataset includes three primary files:<p><ul><li>the plant community composition dataset</li><li>the associated environmental variables dataset</li><li>the plant functional trait dataset</li></ul><p> A fourth file containing geospatial reference data for the four study sites is also included.<p>Combined, the community composition and trait datasets were used to calculate the community-weighted means presented in the manuscript. These data explored how plant community composition and traits change across the latitudinal extent of the boreal biome of western North America, from central Saskatchewan to the northern treeline near Inuvik, NT. <p>In this study, we selected four boreal peatlands sites and within each site, sampled across dominant topoedaphic and/or hydrologic gradients to explore variation both within sites (local-scale) and among sites (regional-scale). At each site, we stratified sample collection by dominant landscape features across each gradient, and we randomly placed two or more 1m by 1m quadrats to assess community composition of vascular plants via stem counts and measure canopy cover, active and organic layer thickness. At three sites (Southern Old Black Spruce, Key Pile Camp, and Havipak Creek) we placed a total of 20 quadrats and, due to an intensive study at Scotty Creek Research Station (see <a href="https://doi.org/10.5683/SP2/R4FTPW">Standen and Baltzer 2021</a>), we placed 80 quadrats. Within each dominant landscape feature at each site, we collected functional trait data on at least 3 replicates of each abundant species. We explored patterns of community weighted mean functional traits pertaining to carbon exchange and community composition of both vascular plants and bryophyte communities and determined whether interspecific or intraspecific trait variation were driving changes within and across these sites. See <a href="https://doi.org/10.1111/jvs.13206">Standen and Baltzer (2023)</a> for more detailed methods.

This is the final dataset associated with the publication, “Permafrost condition determines plant community composition and community-level foliar functional traits in a boreal peatland”, submitted to Ecology and Evolution in April 2021. Included are three data files: the plant community composition dataset, the associated environmental variables dataset, and the plant functional trait dataset. Combined, the community composition and trait datasets were used to calculate the community-weighted means presented in the manuscript. These data explored how plant community composition and traits change across the Scotty Creek Forest Dynamics plot in response to environmental variation, including active layer thickness, organic layer thickness, and forest structure (i.e., canopy cover and tree basal area). To do this, we used a random stratified design to select ten 20 m by 20 m grid cells belonging to each of four aboveground tree biomass categories for a total of 40 grid cells across the Scotty Creek Forest Dynamics Plot. Within each grid cell, we randomly placed two 1 m by 1m quadrats to assess community composition of vascular plants via stem counts and measure canopy cover, active and organic layer thickness. These data were averaged to provide an estimate at the grid cell level. Basal area was calculated at the level of the grid cell. Plant functional traits were collected from 3 replicate individuals per species within 2-3 grid cells per aboveground tree biomass category. See Standen and Baltzer (2021) for more detailed methods.