Epsilon Archive for Student Projects

Abstract

This study examined sensor-supported irrigation management and harvest outcomes in a commercial onion field in southeastern Skåne, Sweden, in 2026. The grower determined irrigation timing and supplied the baseline amount, while soil-sensor graphs and field observations informed recommendations for reduced applications. The non-randomised comparison had two plots per nominal refill treatment. The 100% target was an upper soil-moisture reference; the 80% and 60% targets lay 80% and 60% of the way from a lower irrigation trigger towards that reference, not percentages of applied irrigation. Soil-water-content probes and soil-water-tension sensors recorded local wetting and drying. Irrigation and environmental records supported interpretation. Harvest sampling used four 0.50-m sections per plot of the row of onions formed after lifting.

Across nine events, recorded irrigation totalled 156, 147 and 138 mm for the nominal 100%, 80% and 60% treatments, respectively. The 60% treatment received 18 mm less than the 100% treatment (138 versus 156 mm), an 11.5% reduction. Mean sampled onion mass was 7.7, 7.6 and 7.9 kg per 0.50-m section, including remaining attached foliage and any residual soil. The 60% treatment had more sampled bulbs, lower mean mass per bulb and the smallest mass share above 80 mm. On complete paired days, higher soil-water content was generally associated with lower soil-water tension. Clay content ranged from 4.9% to 15.0% among sampled plots.

Similar sampled masses may reflect adequate water supply and differences in how much water the soils could store. With few plots, different soils and no random allocation of treatments, irrigation effects could not be separated from other field differences. The study shows how soil-sensor information can help guide irrigation within a grower's schedule. Soil, weather and harvest records helped interpret the results. Identifying contrasting soils, monitoring areas prone to water shortage and checking responses across the field could help assess the potential of variable-rate irrigation. Further trials are needed to find suitable irrigation amounts and test whether reductions can be repeated without lowering yield.