An empirical model for root water uptake

Soil water availability estimation is critical for assessing crop development and performance. During periods of soil water deficits, the capability of crop roots to extract soil water depends on the distribution and depth of its root system. Most water uptake models assume a relationship between root water extraction and root length density (RLD). However, models using RLD are difficult to test and several researchers have questioned the various proposed relationships between RLD and water uptake. A simplified water uptake model that does not use RLD was developed, but as an alternative, uses generalizations from measured soil water content changes to predict root water uptake. The daily incrementing model estimates a maximum water uptake rate by roots limited by soil water content that declines exponentially with the soil water content above the lower limit (LL) i.e., the remaining available soil water. The model assumes that: (i) the roots at a given layer have reached a minimum threshold of root density to extract water at a maximum rate; (ii) the transpiration demand is greater than the total root water uptake; and (iii) the water content at LL can be accurately measured or estimated. A critical constant (K) in the exponential model, representing the fraction of extractable water in a soil layer that can be taken up in I day, was found to be 0.096 for several species (cotton, maize, pearl millet, grain sorghum, soybean, sunflower and wheat), and different soil conditions. Values of K smaller than 0.096 were likely caused by root clumping in highly structured (cracking) or compacted soils, where root density was low in deeper soil layers when further downward root growth practically ceased, or by peanut whose K values was 0.064. This new empirical model should help to overcome several of the limitations of current models that rely on the use of measured or predicted RLD. (C) 2003 Elsevier B.V. All rights reserved.