The root system is the plant’s organ that has the important function of taking up water and plant nutrients from the soil. Root system functioning depends on its multiscale structure: the size and structure of cells in the conductive tissues, the development and spatial organization of tissues, and the network of root segments that make up the root system. Breeders select for these structural properties of roots and root systems to develop crop varieties with high water and nutrient efficiencies that give a high or stable yield under non-controllable adverse environmental conditions(droughts, nutrient poor soils, soil salinity, …). But, a direct relation between the functioning of a root system and its structure is not at hand. Functional-structural root system models that simulate the relations between root architectural and hydraulic properties, and the spatio-temporal patterns of water and solute fluxes in the root zone can therefore play an important role in this selection process. However, it must be first demonstrated that these models accurately reproduce the processes they intend to describe but this is a challenging task for two reasons. First, the root system is the so-called ‘hidden halve’ of the plant so that its structure and the flow and transport processes in the soil towards individual roots are hard to observe. Secondly, the properties that control water flow in root segments and that are required as input parameters in structural-functional root models vary with age and root order and quantitative information about these relations is scarce. In this paper, we demonstrate how this deadlock could be broken by combining co-registered root structure and tracer distributions obtained from magnetic resonance imaging, a functional-structural root system model, and inverse modeling. The main features in the tracer patterns were well reproduced by the model using root hydraulic parameters that were obtained using inverse modeling and that were found to correspond with available information about these parameters in the literature. The simulation results further demonstrated that water uptake location and intensity cannot be directly derived from neither observations of tracer accumulation nor water depletion. This proves that functionalstructural root system model simulations, combined with observations, are required to translate the observed variables (tracer accumulation and water depletion) into information about local processes and root system properties.
Vanderborght, J., Koch, A., Meunier, F., Garré, S., Pohlmeier, A., & Javaux, M. (2019). Validation of functional-structural root system models using MRI-monitored tracer experiments. Book of Abstracts, p. 627. https://hdl.handle.net/2078.5/222784