Despite decades of research linking above-ground plant traits to soil processes, relationships between root traits and soil biogeochemistry have been largely ignored. An exception is research relating tree mycorrhizal associations to “rhizosphere effects” (i.e., the stimulation of carbon and nutrient cycling in rhizosphere soils relative to bulk soils), which has shown that roots of ectomycorrhizal (ECM)-associated trees generally have greater rhizosphere effects than roots of arbuscular mycorrhizal (AM)-associated trees. However, the functional variability among species within these groups, the inability of mycorrhizal association to capture the full spectrum of nutrient foraging strategies, and the limited phylogenetic reach of ECM associations call into question whether this ECM-AM dichotomy can capture fine-scale differences in soil biogeochemistry. Furthermore, examinations of rhizosphere effects are commonly conducted in natural forests where root effects may be mediated by leaf litter effects on soil organic matter quality and nutrient availability. To overcome these issues, we evaluated the extent to which a phylogenetic approach could capture variation in rhizosphere effects on nitrogen cycling and extracellular enzyme activities among 25 tree species spanning the angiosperm phylogeny at the Morton Arboretum and West Virginia University’s Core Arboretum where leaf litter impacts are negligible due to removal.
Results/Conclusions
Both evolutionary history and mycorrhizal association explained variation in rhizosphere enzyme activity levels. When analyzed based on mycorrhizal association alone, ECM trees exhibit greater rhizosphere effects on N-degrading enzyme activities (N-acetyl-β-D-glucosaminidase) than AM trees (P=0.015) while AM trees exhibited greater rhizosphere effects on oxidative enzyme activities (polyphenol oxidase) than ECM trees (P<0.001). However, a multivariate model that assumed enzyme activity levels were independent of evolutionary history had a higher AIC (-6.19) than a model that incorporated branch lengths (-8.11), indicating that among-species variation in rhizosphere effects are driven, in part, by evolutionary relatedness. Overall, our results show that rhizosphere effects manifest independently from leaf litter inputs, and the degree of tree evolutionary relatedness is a more useful predictor of tree effects on soil biogeochemistry than mycorrhizal association alone. Our observed phylogenetic variation in rhizosphere effects may be driven by phylogenetic conservatism of other nutrient foraging strategies including low specific root length, high branching intensity, and high root exudate production; these strategies do not always co-vary with mycorrhizal association. As such, phylogenetic relatedness among species may be a useful framework for predicting tree effects on soil biogeochemistry from tree to landscape scales.