Effect of spatial variability on the estimation of the soluble salt content in a drip-irrigated saline loam soil

Abstract

The distribution and total mass of soluble salt in a drip-irrigated vineyard was investigated. Eighty-four positions in a 0.52-ha area were sampled at five depths each, resulting in a total sample number of 420. The salt content increased exponentially with distance from the emitter. At equal distances from the emitter, significantly higher values were observed outside, compared to within, the vineyard row. Outside the row the salt content decreased significantly with depth, but within the row it was constant down to 1 m. Depth, distance from emitter, and position relative to the emitter could account for 52% of the observed variation in salt content. The total salt mass within the study area to a depth of 1 m was estimated to be ca. 22.5 ton/ha. Calculation of the required sample size, combining the central limit theorem with the statistics of the present study, showed that at certain spatial positions relative to an emitter, the type II error of erroneously accepting the null hypothesis and the first estimate of the salt content could be as high as 41%. The initial sampling scheme could be improved by taking account of the observed spatial variation.