Please use this identifier to cite or link to this item: http://arks.princeton.edu/ark:/88435/dsp01d217qr83q
 Title: Optimization of Root Depth Profile to Improve Representation of Dynamic Rooting Decisions of Two Tree Species at Free-air CO$$_{2}$$ Enrichment Sites Authors: Best, Kelsea Boehner Advisors: Medvigy, David M. Department: Chemical and Biological Engineering Class Year: 2015 Abstract: Forest ecosystems play an important role in global carbon (C) cycles, and terrestrial biosphere models (TBMs) are powerful tools for predicting how forest ecosystems will respond over long periods of time to different environmental conditions. In this work, I seek to improve root representation in TBMs by introducing a rooting depth profile optimization into the existing TBM, Ecosystem Demography model version 2 (ED2). Loblolly pines, the primary tree species at Duke Forest, and sweetgums, the species located at Oak Ridge National Laboratory (ORNL) were chosen as case studies for simulations in order to improve the model’s ability to replicate key observations in root allocation at free-air CO$$_{2}$$ enrichment (FACE) sites. The rooting depth profile optimization is modeled to maximize the limitation factor (LF), which takes water and nitrogen (N) stress into consideration. Tests with a single layer and two-layer nutrient profile at ambient as well as elevated carbon dioxide (eCO$$_{2}$$) levels show that the optimization, though dependent on initial conditions and N availability, is able to improve growth for plants in heterogeneous soil conditions by enabling roots to move towards pools of available resources. These results also suggest that plants might experience a tradeoff in N and water uptake when allocating roots. In the process, I was able to successfully replicate trends in root allocation at FACE sites. This work also begins to assess the role of dynamic root profiles as an advantage when grown in competition with other plants, and begins to assess the differences in optimal rooting strategies between the plant types. Overall, this research indicates that optimization modeling, with the appropriate target function, could be a promising technique to further improve the representation of below-ground phenomena and overall accuracy of TBMs. Extent: 54 pages URI: http://arks.princeton.edu/ark:/88435/dsp01d217qr83q Type of Material: Princeton University Senior Theses Language: en_US Appears in Collections: Chemical and Biological Engineering, 1931-2016