Spatial variation of soil water repellency in a commercial orchard irrigated with treated wastewater

The recognition of treated wastewater (TWW) as an alternative water resource is expanding in areas with a shortage of freshwater (FW). While many studies have been devoted to the effects of long-term irrigation with TWW on soil wettability and spatial flow variations in the soil profile, much less attention has been given to the spatial distribution of soil water repellency in the soil surface layer. This is the objective of the current study. Undisturbed soil samples (5 cm thick) were taken at 15-cm intervals parallel to a drip lateral in two adjacent plots of a commercial citrus orchard in central Israel. Each soil sample was sectioned into five consecutive 1-cm layers for which soil water repellency was determined by water drop penetration time method, and soil organic matter by loss-on-ignition method. Geostatistics and multivariate empirical mode decomposition were used to investigate the overall and scale-specific spatial variation of soil water repellency and its dependence on dripper intervals along the lateral. A high degree of soil water repellency with strong spatial variation was found in the surface soil after 4-6 years of TWW irrigation. Weak to moderate spatial dependence of soil water repellency with maximum autocorrelation distance of around 30 cm was discovered by geostatistical analysis. The spatial distribution of soil water repellency was considered to be greatly affected by the location of the drippers, being higher between adjacent drippers and lower underneath them. This soil water repellency distribution is presumed to result from ongoing lateral displacement of the amphiphilic substances in the TWW toward the outer edge of the wetted plume periphery. Multivariate empirical mode decomposition of the overall spatial variation of soil water repellency yielded three scale-specific variations with corresponding characteristic scales of 30 cm, 110 cm and 200 cm. Most of the soil water repellency variation was separated into the 30 cm and 110 cm spatial scales, which were correlated to processes related to the drippers and trees. Replacing TWW with FW for the reclamation of water-repellent soils partially alleviated the intensity of TWW irrigation-induced soil water repellency. Moreover, an inconsistency between the hot spots of water-repellency development between adjacent drippers and the areas that are effectively ameliorated by FW irrigation below the drippers could be developed and affect the spatial distribution of flow pattern in an a priori unpredictable way.