Allocation of grass fields to Hybrid Soil Moisture Deficit model drainage classes using visual indicators

Abstract

The Hybrid Soil Moisture Deficit (SMD) model predicts soil water status relative to field-capacity and can be calculated directly as a soil water mass balance for three drainage classes (well-drained, moderately drained and poorly drained). For the Hybrid SMD model to be deployed, grassland fields must be assigned to the correct drainage class. A series of context and vegetation indicators along with visual soil indicators were selected and evaluated as signs of drainage status for objective, rapid allocation of a field to a drainage class. Two groups of testers applied the methodology in the field. The first group of soil scientists assisted with selecting the best terminology and simplify the indicators to the most parsimonious grouping possible, while the second group (non-experts) was used to evaluate whether a consensus could be achieved with adequate reliability of indicators in the hands of non-experts. The performance of soil indicators for hydrological discrimination was also tested by comparing them with saturated hydraulic conductivity measured at 10 field sites. A quick infiltration test derived from the inversed auger hole method and structure-related soil indicators were found to be significantly correlated with saturated hydraulic conductivity. Although indicators are not absolute, the balance of indicators leaning towards either poorly drained or well-drained resulted in correct allocation of fields to SMD drainage classes by both non-experts and soil scientists. The contribution of each indicator to the allocation of a field into one of the three SMD drainage class was weighted depending on operator consistency and indicator reliability for hydrological discrimination between fields to indicate their relative contribution to the final decision of class allocation. The combination of plant, soil and water indicators can be learned rapidly, easily certified for use by advisors and rapidly deployed to classify fields in the context of the Hybrid SMD model.