Results of a combined model of root system growth and soil water uptake: evaluating the significance of root system architecture to plant water uptake

Root system hydraulic architecture is a key determinant of plants' ability to withdraw water from the soil, satisfying transpirational demand. Presently, the representation of this component of the hydrological cycle in large-scale models is generally very simplistic, even though transpiration accounts for much of the terrestrial heat and water surface fluxes, and exercises control over photosynthetic uptake of CO2. In order to address this gap, we have developed a modelling approach that relies on several components. The first is RootGrow, original MATLAB code that simulates the stochastic growth of a root system as a function of an intrinsic set of parameters as well as its environment. We ran RootGrow coupled to the second component, a finite-element 3D simulation of the physics of water transport in the soil and root system using COMSOL, resulting in a combined model of root system development and water uptake. Model results show that root system architecture can affect water uptake by two separate mechanisms: (a) root system geometry determines the distribution of absorbing surface area throughout the soil domain, and (b) root system topology affects the water potential at the absorbing surfaces. In this study we sample the model's parameter space to demonstrate over what ranges of physically meaningful parameters (including hydraulic conductivity of plant tissues, soil type, and soil moisture level) these mechanisms significantly affect root systems' water withdrawal rate. The two mechanisms identified and our quantitative results will form the basis of a third component in this approach: developing simple analytical relationships characterising water uptake as a function of root system architecture that can be used in Ecosystem Demography Model v2.1 (ED2), a large-scale Dynamic Vegetation Model, based on a method of upscaling individual-based models of plant ecology.