Mycorrhizal associations influence soil nitrogen dynamics via effects on soil acid-base chemistry and N cycling microbial communities

Abstract:

Global environmental changes are shifting the distribution and abundances of species in ecosystems, yet the consequences of such changes are poorly understood. Here we present an analytical framework for understanding the changes in nitrogen (N) cycling that may arise owing to plant community change in forests. The Mycorrhizal-Associated Nutrient Economy (MANE) framework predicts that species that associate with different types of mycorrhizal fungi possess an integrated suite of nutrient-use traits that lead to the maintenance of biogeochemical syndromes in forests. We hypothesized that the commonly-observed pattern of accelerated N cycling in forests dominated by arbuscular mycorrhizal (AM) trees relative to forests dominated by ectomycorrhizal (ECM) trees can be best explained by tree effects on soil acid-base chemistry. Further, we hypothesized that tree-induced alterations of soil pH would affect soil N dynamics by altering the N-cycling soil microbial community.

We tested the first hypothesis using plant and soil data from a landscape-scale study of 230 subplots varying in mycorrhizal dominance, a regional-scale study of AM- and ECM-dominated plots in 10 temperate forests in the eastern USA, and a global meta-analysis of 105 sites where AM and ECM forest stands co-occurred. We tested the second hypothesis by sequencing soils from 54 plots that vary in mycorrhizal dominance across six forests in the eastern USA. We examined microbial communities from upper surface soils, and investigated the microbial taxa most involved in soil N transformations using metagenomic sequencing.

In support of the first hypothesis, we found that AM-dominated plots had higher soil pH and greater exchangeable base cations than ECM-dominated plots (P < 0.05), and soil acid-base status was a consistently strong predictor of soil N cycling rates. Further, our analysis revealed that the amelioration of soil acidity beneath AM trees may be caused by their base cation-rich and lignin-poor leaf litters. These patterns occurred across spatial scales, suggesting the scale-invariance of these dynamics. In support of the second hypothesis, AM- and ECM-dominated plots were associated with distinct microbial communities and activities (P < 0.05), especially those related to N cycling. AM-dominated plots contained an order of magnitude more copies of key N cycle genes, determined from metagenomic sequencing.

Collectively, our results indicate feedbacks between plant nutrient use strategies and soil properties have the potential to impact forest community assembly and ecosystem processes, particularly in the wake of global environmental change. As such, future shifts in the relative abundance of AM- and ECM-associated trees may have profound yet predictable implications for how forests cycle nutrients.