Application of a tree-level hydrodynamic model to simulate plot-level transpiration in the upland oak/pine forest in New Jersey

The Finite-Elements Tree-Crown Hydrodynamics model version 2 (FETCH2) simulates water flow through the tree using the porous media analogy. Empirical equations relate water potential within the stem to stomatal conductance at the leaf level. Leaves are connected to the stem at each height. While still simplified, this approach brings realism to the simulation of transpiration compared with models where stomatal conductance is directly linked to soil moisture. The FETCH2 model accounts for plant hydraulic traits such as xylem conductivity, area of hydro-active xylem, vertical distribution of leaf area, and maximal and minimal xylem water content, and their effect on the dynamics of water flow in the tree system. Such a modeling tool enhances our understanding of the role of hydraulic limitations and allows us to incorporate the effects of short-term water stresses on transpiration. Here, we use FETCH2 parameterized and evaluated with a large sap-flow observations data set, collected from 21 trees of two genera (oak/pine) at Silas Little Experimental Forest, NJ. The well-drained deep sandy soil leads to water stress during many days throughout the growing season. We conduct a set of tree-level transpiration simulations, and use the results to evaluate the effects of different hydraulic strategies on daily transpiration and water use efficiency. We define these "hydraulic strategies" through combinations of multiple sets of parameters in the model that describe the root, stem and leaf hydraulics. After evaluating the performance of the model, we use the results to shed light on the future trajectory of the forest in terms of species-specific transpiration responses. Application of the model on the two co-occurring oak species (Quercus prinus L. and Quercus velutina Lam) shows that the applied modeling approach was successfully captures the differences in water-use strategy through optimizing multiple physiological and hydraulic parameters.