Mycorrhizal associations of tree species influence soil nitrogen dynamics via effects on soil acid–base chemistry

Abstract Aim Plants and their associated microbes influence nutrient cycling in terrestrial ecosystems, yet we have a limited understanding of how soil acidity mediates the process. Here, we investigate whether reported differences in nitrogen (N) cycling between forests dominated by arbuscular mycorrhizal (AM) trees and ectomycorrhizal (ECM) trees are related to changes in soil acid–base chemistry induced by mycorrhizal associations. Location Global. Time period 1969–2018. Major taxa studied Trees. Methods We measured and synthesized variables of leaf litter quality, soil acid–base chemistry and N cycling from: (1) a landscape‐scale study of 230 subplots varying widely in AM tree dominance in a 25 ha forest plot; (2) a regional‐scale study of 40 AM‐ and 56 ECM‐dominated plots in 10 temperate forests across the eastern USA; (3) a continental‐scale study of > 3,000 forest plots from 10 ecoregions across the contiguous USA; and (4) a global meta‐analysis of 105 study sites with co‐occurring AM and ECM forest stands. Results Across all spatial scales, ECM‐dominated forests were associated with greater soil acidity. In particular, ECM‐dominated soils exhibited lower soil pH and base cations, although the magnitude of mycorrhizal‐associated differences in soil acid–base chemistry depended on the biomes, with differences being more pronounced in temperate than in sub/tropical forests. Higher lignin and lower base cations in ECM tree leaf litter were related to greater soil acidity in ECM‐dominated forests. Moreover, the lower inorganic N concentrations and slower N transformation rates in ECM‐dominated forests were associated with their greater soil acidity. Main conclusions Our results indicate that the scale‐invariant feedbacks between plant nutrient‐use strategies and soil properties have the potential to impact forest community assembly and ecosystem processes, particularly in the context of global change.