The process description of plant transpiration and soil water uptake in macroscopic root water uptake models is often based on simplifying assumptions that no longer reflect, or even contradict, the current status of knowledge in plant biology. The sink term in the Richards equation for root water uptake generally comprises four terms: (i) a root resistance function, (ii) a soil resistance function, (iii) a stress function, and (iv) a compensation function. Here we propose to use a detailed three-dimensional model, which integrates current knowledge of soil and root water flow equations, to deduct a one-dimensional effective behavior at the plant scale and to propose improvements for the four functions used in the macroscopic sink term. We show that (i) root hydraulic resistance may be well defined by the root length density but only for homogeneous lateral conductances and no limiting xylem conductance—in other cases a new function depending on the root hydraulic architecture should be used; (ii) soil resistance cannot be neglected, in particular in the rhizosphere where specific processes may occur that alter the soil hydraulic properties and therefore affect uptake; (iii) stress and compensation are two different processes, which should not be linked explicitly; (iv) there is a need for a clear definition of compensatory root water uptake independent of water stress; (v) stress functions should be defined as a maximal actual transpiration in function of an integrated root–soil interface water head rather than in terms of local bulk water heads; and (vi) nonlinearity in the stress function is expected to arise if root hydraulic resistances depend on soil matric head or when it is defined as a function of the bulk soil water head.

You do not currently have access to this article.