Caitlin Hicks Pries, Dartmouth College (Principal Investigator)

Nina Wurzburger, University of Georgia (Co-Investigator)

Benjamin Sulman, Oak Ridge National Laboratory (Co-Investigator)

Richard Lankau, University of Wisconsin (Co-Investigator)

Symbiotic mycorrhizal fungi are major drivers of forest soil biogeochemical processes. Mycorrhizal fungi provide trees with nutrients and in return, the trees provide the fungi with organic carbon. There are two main types of mycorrhizal associations—arbuscular mycorrhizae and ectomycorhizae—which differ in how they interact with plant roots and acquire nutrients. Most tree species associate with one of these mycorrhizal types. Generally, trees that associate with arbuscular mycorrhizae have faster-decomposing leaves than trees that associate with ectomycorrhizae. Tree mycorrhizal associations are thus a promising framework by which to study the effect of aboveground tree communities on belowground soil processes. However, most research on mycorrhizal associations has occurred across natural gradients and has not involved manipulative experiments that can address mechanisms and causation. There are two competing hypotheses to explain how mycorrhizal association affects soil carbon and nitrogen that have yet to be resolved: mycorrhizal nutrient acquisition strategies versus inherent differences in litter decomposability. Due to the widespread changes in the tree species composition of temperate forests due to climate-driven range shifts, oak decline, and invasive pests, a mechanistic understanding of mycorrhizal associations is key to accurately predicting the consequences of changing forest composition in biogeochemical models.

Our main objective is to understand the degree to which the observed differences in soil carbon and nitrogen dynamics in ectomycorrhizal (EcM) and arbuscular mycorrhizal (AM) dominated forests are driven by litter decomposability versus mycorrhizal fungal function using targeted experiments, and to incorporate this knowledge into a process-based soil organic matter model.  Previous model simulations based on litter decomposability suggest there is a larger proportion of mineral-associated organic matter in AM forest stands relative to EcM forest stands. However, preliminary data suggest that mycorrhizal fungal abundance is a stronger driver of soil organic matter patterns than litter. To achieve our objective, we will test the mechanisms by which soil carbon and nitrogen are affected by mycorrhizal dominance via two novel experiments: 1) a growth chamber experiment across four EcM and five AM tree species using a ¹³C-labeled atmosphere to trace seedling-derived carbon into hyphae, the rhizosphere, and various soil fractions; and 2) an in situ decomposition experiment of six different ¹³C and ¹⁵N labeled litters that range in decomposability incubated across a gradient of EcM dominance at three sites that capture important variation in climate, soils, and forest species composition. Our experiments are motivated by uncertainties in model structural representation of plant-mycorrhizae-soil interactions, and the results of our experiments will inform a mycorrhizal community-explicit version of the soil organic matter model named FUN-CORPSE. This new version of FUN-CORPSE will be developed simultaneously with the experiments, tested or parameterized using the experimental data, and used to project the consequences of several global change scenarios.

Key outcomes of this work will include (1) a mechanistic understanding of how mycorrhizal associations affect soil carbon and nitrogen cycling; (2) a parameterized and tested soil organic matter model incorporating mechanistic differences between major mycorrhizal types using a model framework that has already been successfully coupled into global land surface models; (3) experimental constraints on mycorrhizal model parameters that have been previously identified as key sources of predictive uncertainty; (4) empirical tests of whether incorporating mycorrhizal function into models leads to more accurate estimates of forest soil carbon and nitrogen cycling. The improved model will enhance our ability to predict how soil carbon and nitrogen across the Eastern U.S. will be affected by global change.