Probability Sampling Techniques for Estimating Soil Erosion

Publisher Website
Google Scholar
Abstract
Data on the lateral variability of soil loss from Submediterranean rangelands are presented. Modified Gerlach troughs were used to measure both rill and interrill erosion. Random, systematic, stratified random, and cluster sampling methods were tested for various numbers of plots and plot widths in order to find the method that gives the most accurate estimate of soil loss. If both the rill and interrill subpopulations are sampled, random sampling and stratified random sampling give almost identical standard errors. Cluster sampling yields larger errors and is therefore less effective. Systematic sampling, however, allows one to choose more accurate samples of an economical size, but its efficacy depends largely on the number of rills in the study area. The smallest standard errors are obtained by measuring the total rill erosion and by sampling only the interrill subpopulation. When this procedure is applied, the four sampling methods turn out to be equally accurate.

This publication has 0 references indexed in Scilit: