Global mycorrhizal status drives leaf δ15N patterns

Nitrogen (N) availability, which can be represented by the natural abundance of the stable N isotope δ15N, is crucial to understanding ecosystem-level N dynamics. Specific ecosystems are dominated by different types of mycorrhizae, which can relate to biogeochemistry and affect ecosystem functioning. However, few studies have addressed the impacts of different mycorrhizal associations on variations in foliar δ15N due to climatic and soil physicochemical factors; prior instances of foliar δ15N modeling have not included mycorrhizal types. Here, we used machine learning to produce a global map of foliar δ15N based on climatic, edaphic, vegetation, and dominant mycorrhizal factors. The predicted global average foliar δ15N value was 0.69‰. Plants in tropical areas were predicted to have significantly larger foliar δ15N values than plants from subtropical, temperate, and boreal areas. The mean annual temperature was identified as the primary driver of spatial foliar δ15N patterns. These results provide isotopic evidence of greater N limitations in temperate and boreal regions than tropical or subtropical regions. Furthermore, non-mycorrhizal plant species had the highest foliar δ15N values, followed by plants associated with arbuscular mycorrhizae, orchid mycorrhizae, ectomycorrhiza, then ericoid mycorrhizae.  Overall, changes in foliar δ15N were predicted to be closely associated with the type of mycorrhizal association. This study highlights the importance of incorporating mycorrhizal data to accurately assess patterns of foliar δ15N on a global scale. Ultimately, our findings contribute to a greater understanding of N cycling dynamics across plant types and global ecosystems.

Methods
Foliar δ15N values were obtained from a recent version of the global dataset described by Craine et al. (2018) that was updated with newly published data for Meta-analyses.  Multi-year average MAT, MAP, and PET maps with a spatial resolution of 4 km × 4 km for 1982 through 2018 were extracted from the TerraClimate dataset (Abatzoglou et al., 2018). AI values (defined as the ratio of precipitation to PET) were calculated from MAP and PET values. A digital elevation model (DEM) map with a spatial resolution of 1 km × 1 km was extracted from the Global Land One km Base Elevation (GLOBE) Project (https://www.ngdc.noaa.gov/mgg/topo/globe.html). A slope map was generated from the DEM map. Soil clay, silt, sand, soil organic carbon (SOC), and TN contents with a spatial resolution of 250 m × 250 m were obtained from the SoilGrids dataset (Hengl et al., 2017). Multi-year (1982–2018) GPP values were calculated using data from the Global Land Surface Satellite (GLASS) project (Liang et al., 2021). Multi-year (1982–2015) average normalized difference vegetation index (NDVI) values were calculated from the GIMMS3g dataset (Tucker et al., 2005). The mycorrhizal plant type map (showing the distribution of AM, ECM, ERM, and NM plants) was generated from maps showing the proportional aboveground plant biomass of AM, ECM, ERM, and NM plants (Soudzilovskaia et al., 2019) for Random Forest.