Field estimation of water extraction coefficients with APSIM-Slurp for water uptake assessments in perennial forages

The assessment of water uptake in agricultural systems commonly relies on the use of process-based biophysical models. To accurately represent crop-soil relationships, these models require local calibration which is often limited by the availability of site-specific data. This is the case for the water extraction coefficient (kl) in perennial forage cropping systems. The kl is critical to control water uptake by roots influencing soil moisture dynamics at different soil depths. This crop-soil parameter is particularly challenging to assess in species that establish deep root systems during long periods of regrowth, such as perennial forages. Using three years of detailed field data, we test a method to estimate kl for two perennial crops of broad socio-economic significance (lucerne and perennial ryegrass). The method is based on three physically meaningful parameters: Root Front Velocity (RFV, mm/day), surface kl (kl0,/day) and the rate of kl decay with soil depth (λkl, dimensionless). Our analysis showed that soil volumetric water content dynamics was most sensitive to kl0 values. A model fitting procedure showed that the highest accuracy of soil volumetric water content (θ) simulations was obtained with RFV of 10 mm/day and a kl0 of 0.11/day for both forages. In contrast, λkl estimates differed among species, being higher for shallow fibrous ryegrass roots than for the deep lucerne taproots. The analysis also highlighted that multiple parameter set combinations were found to give simulations of acceptable accuracy. The proposed method can be used as a first approximation to parameterise kl when local calibration data is unavailable. However, even for the fitted parameter sets, the accuracy of θ estimates was low at soil depths where there was a substantial transition of soil texture. These insights highlight important aspects to be considered in future development and parameterisation of water uptake models for perennial forages.