Modelling root growth of grain sorghum using the CERES approach

Abstract

A simple model is described, based on the approach used in the CERES crop growth models, which simulates the depth of rooting and root length density in each soil layer for grain sorghum growing under soil drying. The model has five main components: (1) daily accumulation of root length is proportional to above-ground biomass growth, (2) the root front descends at a constant rate from sowing until early grain-filling, (3) daily accumulation of root length in water non-limiting conditions is partitioned among the occupied soil layers in an exponential pattern with depth, (4) proliferation of root length is restricted in any layer if the extractable soil water in that layer declines below a threshold, and (5) a fixed proportion of existing root length is lost due to senescence each day. The parameter values for the relationships were derived from analysis of measured depth distributions of root length from crops of grain sorghum grown in the sub-humid subtropics of Australia, on oxisol and vertisol soil types. The soils had no physical or chemical restrictions to root growth.

The model was validated using four independent data sets. Overall, the model simulated the root distribution with depth well, but predictions of accumulated root length were less reliable. One of the most sensitive parameters affecting the modelled distribution of root length with depth was the factor used to partition daily accumulation of root length among the occupied layers, and the value of this parameter varied between well-watered and water-limited crops. The study shows that it is possible to model root growth of field crops using only five simple relationships, with inputs that are already used in most crop growth models.