Many factors determine root water uptake. The interaction of all these factors makes it difficult to model water uptake. In this study, different aspects of root water uptake were investigated and several models were tested.
Roots have some special features to adequately supply the above ground parts of the plant with water and nutrients. The root hair zone close to the root tip has very high uptake rates. Older root also absorb water, but at much lower rates due to suberization and secondary growth of cork. In mature root systems with large suberized parts, the suberized roots may however absorb most of the water required by the plant. The rate of absorption of water by the roots depends on the hydraulic conductivity of soil and root and on the magnitude of the pressure gradient between root and root surface and root and leave.
Soil moisture distribution influences water uptake in different ways. In the first place, soil moisture content influences water uptake on a specific spot along a root. At low (or very high) soil moisture pressure heads, plants cannot extract water at the maximum possible rate. Secondly, soil moisture distribution affects uptake of a soil system as a whole. Hampered uptake in one part of the root system may be compensated by enhanced uptake in wetter parts. Long term adverse conditions might result in adaptation of the root system, whether by growing into more favorable soil layers or by changing morphology (e.g. aerenchyma formation or enhanced suberization).
Besides soil moisture status, many other factors also play a role in determining water uptake rates. For example evaporative demand of the atmosphere, soil salinity, nutritional status of soil and plant, pests and diseases, developmental stage and root age. These factors are crop-specific and their influence may change due to other environmental conditions.
Water flow in unsaturated or partly saturated soil may be described with Richards’ equation. This equation is a combination of Darcy’s law and the continuity equation. In order to solve Richards’ equation the relationships between hydraulic properties of the unsaturated medium have to be known. These relationships, the retention curve and the hydraulic conductivity curve, may be provided in tabular form or described by empirical functions. Unfortunately, these relationships are affected by hysteresis and therefore not unique.
Many models exist that describe root water uptake. One approach is to regard the total root system as a diffuse sink. This macroscopic approach describes water uptake by an empirical function. Several models are mentioned that calculate water uptake from potential transpiration, root length density and some kind of reduction factor to account for reduced uptake due to water stress. Some of the mentioned models also account for the phenomenon that water stress in one part of the root zone can be compensated by enhanced water uptake from other moister parts. Besides the discussed models, many models with additional features exist, e.g. models that account for salinity stress or incorporate age or root growth.
Another approach to model root water uptake is the microscopic approach. This approach considers water flow into single roots. A disadvantage of this kind of models is the large number of parameters that have to be determined. These comprehensive models are however important to increase knowledge on the behavior of plant roots. A few microscopic models are applicable without a large amount of input parameters. These models are however rather simplified forms of the original mathematically derived equation.
Three macroscopic models and one microscopic model were tested. One model used potential transpiration, root length densities and a reduction factor to account for uptake reduction due to water stress. Two other tested models also took into account water stress compensation. Also a very simple microscopic model was tested. Data were available from an experiment with wheat plants in containers. One container received plenty of water, while another did not receive any water. The models were added to a computer program for the simulation of unsaturated soil moisture flow written in the computer language FST.
It was not possible to choose which model simulated the measured soil moisture profiles of the dry container best. All gave acceptable results, with only small differences between the models. One explanation for this similarity might be the quite favorable conditions. Another explanation might be redistribution of soil moisture. Changed uptake patterns are to a certain extent cancelled out by changed moisture flow patterns. The soil moisture profiles of the irrigated container could not be simulated well, probably because of hysteresis in the soil hydraulic relationships.