Influence of water stress level on determination of soil moisture sensor position under variable rate irrigation

Abstract: Determining positions to represent mean soil water content based on soil clay contents is an alternative method for positioning soil water content sensors in wireless sensor networks for a variable rate irrigation system. The field was divided into 4 management zones according to available soil water holding capacity (AWC). Two of the 4 management zones were selected to arrange the rain-fed treatment and 3 irrigation treatments representing different water stress levels to assess the effect of the levels on the placement of soil water content sensors under variable rate irrigation system. In zone 1, sand fraction largely increased with depth with AWC within 1-m soil profile ranging from 152 to 161 mm. In zone 2, a relatively uniform profile was observed along the profile with AWC within 1-m soil profile ranging from 161 to 171 mm. Based on the temporal stability of soil water content, the effects of soil water status and soil properties on the similarity of soil water spatial pattern and the positions directly representing the plot-mean soil water content were studied. Results showed that both of soil texture and water stress had effect on the structure similarity of soil water content distribution. In zone 1, the average Spearman's rank correlation coefficient of 0-0.6 m was significant at the probability level of 0.05 only in the rain-fed treatment in the 2016 season. In zone 2, the Spearman's rank correlation coefficient was significant at the probability level of 0.05 in all treatments in the 2016 season and in the medium and low water stress treatments in the 2017 season. The percentages of positions directly representing the mean soil water content were almost the same in zones 1 and 2. Affected by soil water status, the percentages increased as the level of soil water stress decreased in zone 1. While in zone 2, as the severity of water stress decreased, the percentages decreased and then had a slight increase. In general, significant linear regressions (P<0.05) between the mean clay content and the clay content representing the mean soil water content sites were found in layers 0-0.2, 0.2-0.4, and 0.4-0.6 m for all the treatments in 2016 and 2017, except for that in the severe water stress treatment in 2016. The fitted equation coefficients ranged from 0.66 to 1.03 in the 2 seasons, demonstrating a clearly increasing trend as the severity of water stress increased in 2017. When the mean clay content was used for a priori identification for positioning soil water content sensors in the management zones under variable rate irrigation system in a field with sandy loam soil, the strategies of water stress management should be considered in determining a fitted equation coefficients.