Using maize to evaluate the Mohammadi–Khataar (M–K) model as a salinity weighting function (

Abstract

Context Soil water availability, as characterised by the integral water capacity, uses weighting functions based on models not yet evaluated using plants, especially in the context of saline soils. Without plant evaluation such weighting functions remain theoretical at best. Aims We aimed to use maize plants to evaluate Mohammadi and Khataar’s (2018) conceptual model for a salinity weighting function, against those used in Hydrus 1D. Methods We conducted glasshouse experiments with large columns of two sandy loams planted with maize irrigated using different salinities, and repeated without plants. Soil matric suction ranged between saturation and 100 cm, and we measured or predicted plant height, transpiration, evaporation, drainage, storage, and solute concentration over time. The soil water retention curve was measured and the weighted mean hydraulic conductivity was obtained using the van Genuchten model. Key results We found a correlation between our salinity weighting function and the relative transpiration rate of maize (grown in two different soils using irrigation water of three different salinities), particularly in the first few days of growth but not thereafter; errors were related to uncertainties in predicting drainage, salt concentration, and soil water storage in planted columns. Conclusions The deviation of transpiration rate from that predicted by our salinity weighting function at higher salinities may relate to the linear nature of the Maas–Hoffman salinity weighting function plus heterogeneity of soil water and solute distributions. Implications Improving the estimates of drainage and soil water storage in future would make our physical model more useful in larger scale hydrological predictions.