From rhizosphere to forest: Scaling shifts in plant-microbe interactions in infected eastern hemlocks to predict changes in
ecosystem carbon and nitrogen cycling
A. Keiser, University of Massachusetts Amherst (Principal Investigator)
K. DeAngelis, University of Massachusetts Amherst (Co-Investigator)
D. Sihi, Emory University (Co-Investigator)
B. Sulman, Oak Ridge National Laboratory (Unfunded Collaborator)
Non-native insects and diseases have proliferated across U.S. forests, resulting in a reduction of carbon (C) stored within infected live trees. However, unmeasured changes belowground could further shift ecosystem C cycling in infected forests. A combination of altered climate plus non-native pests can influence how trees interact with the soil. Chemical interactions at the root-microbial interface, called the rhizosphere, can help trees fight non-native pests by stimulating shifts in the rhizosphere microbial community or soil physicochemical properties. Exudate compounds released by the roots help provide energy for rhizosphere microorganisms to mineralize soil organic matter (SOM); thus a shift in the amount or composition of exudates could alter C and N cycling and storage for an infected forest. Infected eastern hemlocks demonstrate variable rates of decline across space, and this variation could be explained by rhizosphere plant-microbial feedbacks as mediated by local soil chemical or physical properties. Our goal is to understand how rhizosphere plant-microbe feedbacks temper the decline of tree species infected by non-native pests and the subsequent impacts on ecosystem carbon (C) and nitrogen (N) cycling. We will use a model-experiment (ModEx) approach combining field, lab, and modeling experiments to test our central hypothesis that C and N cycling in forests infected by non-native pests is controlled by shifts in exudate quantity and quality driving changes in the composition of the rhizosphere microbial community and SOM mineralization but mediated by soil type and chemical properties.
The hemlock woolly adelgid (Adelges tsugae, or HWA), which infects the eastern hemlock (Tsuga canadensis), is one of 15 non-native insects and diseases responsible for the most significant damage to U.S. forests. We will use the eastern hemlock as a model species for the growing pressure on U.S. forests from non-native pests to test three specific objectives: (1) quantify changes in eastern hemlock root exudates and associated shifts in the rhizosphere microbial community across variable infection levels and soil physicochemical properties; (2) evaluate microbial and soil organic matter (SOM) turnover under altered root exudation and precipitation regimes; and (3) integrate our empirical data into two variations of an ecosystem model, FUN-CORPSE (couples SOM cycling with plant-microbe feedbacks) and myco-CORPSE (explicitly represents mycorrhizal and saprotrophic rhizosphere communities), to refine predictions of ecosystem C and N cycling across altered hemlock forests.
Working across two field gradients with uninfected, low infection, and high infection stands of mature hemlocks (Gradient A) or seedlings (Gradient B), we will collect in situ root exudates to analyze for quantity and composition. We will then quantify changes along the rhizosphere root-microbe-soil continuum, including the composition and size of the microbial community, the size and quality of SOM pools, C and N transformations, and soil physicochemical properties, to advance our understanding of how plant-soil feedbacks moderate the impact of a non-native pest and alter C and N cycles. We will complement field measurements with a laboratory incubation study using model exudates enriched with 13C to isolate the formation and turnover of soil C pools and shifts in microbial community composition across soil types. Our field and lab measurements will support an iterative ModEx approach to parameterize an ecosystem model that couples mechanisms represented by FUN-CORPSE and myco-CORPSE. While FUN-CORPSE combines plant-microbe interactions with SOM turnover, myco-CORPSE incorporates competition between rhizosphere saprotrophs and mycorrhizal fungi. Combining these two models within a single platform will allow us to test whether microbially-explicit parameters belowground (myco-CORPSE) are more important for constraining C following shifts in plant-soil feedbacks (FUN-CORPSE) occurring in the rhizosphere. Together, our ModEx approach will advance our understanding of and ability to accurately predict forest ecosystem C and N cycling when belowground rhizosphere dynamics are altered by aboveground stressors.
