Assessing soil water thresholds for irrigation of maize in north-eastern Austria

The consequences of climate change will affect irrigation also in areas that have so far been little affected by water shortages. In the north-eastern part of Austria with predominantly sub-humid conditions (550 mm mean annual rainfall, 11°C mean annual temperature), water demand is expected to double by 2050. With regard to irrigation, this means both increased crop water requirements and reduced water availability. Improving water use efficiency is one way to sustain agricultural productivity and water resources in the long term. This poses a future challenge also for farmers who have had little to do with the requirements of highly efficient irrigation management so far. In this regard, a potential on-farm strategy is to monitor soil water status and control irrigation based on the plant available soil water. On the technical side, sensors and telemetry networks are available that regularly collect and transmit data. However, the appropriate thresholds for irrigation control must also be known. These depend on the crop, the root development, the soil type and the installation depth of the sensors. Last but not least, farmers need to trust the decision support and use the information appropriately. The aim of this study was to determine threshold values ​​for the irrigation of maize at a location in north-eastern Austria (in the country's largest arable farming area, approx. 35 km east of Vienna) and to subject them to a practical test in the field.

The irrigation system used was a hose reel with an irrigation boom with rotating nozzles (BAUER GmbH). HydraProbe (Stevens Water Monitoring Systems Inc.) and Watermark (Irrometer Company Inc.) sensors were used to measure soil water content and matric potential, respectively. The sensors were integrated into a telemetry network (ADCON by OTT HydroMet GmbH); data were available via a web-application. For practical reasons, the installation depth of the sensors was set at 20 cm. Soil was a sandy loam. A target value for the maximum allowed depletion under optimal to slightly stressed conditions was calculated using a water balance model. Based on HYDRUS-1D (PC-Progress s.r.o.) simulations, the corresponding matric potential at 20 cm depth was −100 kPa. This value should serve as the target value for irrigation. The soil water content data served as a control for the simulation using the FAO AquaCrop model. The latter was used to evaluate the irrigation carried out by the farmer.

The soil water data revealed that the specified threshold was not reached in the practical test. It seems that farmers experience was oriented towards a more sufficient water supply. The simulation showed that water use efficiency could have been improved by using less water; however, in this case a reduction in yield of a few percent would have to be expected. Measuring with only one sensor at a depth of 20 cm proved to be a viable procedure; however, data interpretability could be improved by sensors at several depths.