Sampling procedures for spittlebug adults in pastures of Panicum maximium

Abstract

A sampling study of spittlebug adults in pastures of Panicum maximum cv. Colonião was conducted to find an appropriate number of sweep-net sweeps per sample and sample numbers. Based on the values of relative variation (SE/x) x 100, total number of sweeps, and number of samples, ten ten-sweep samples were considered to be better than five 40-sweep samples, five 20-sweep samples, or 20 ten-sweep samples. The number of samples required for a particular level of relative variation at population densities >17 spittlebug adults/10 sweeps remained the same, whereas at densities < 17, the number increased inversely. A regression model to convert relative estimates obtained by sweep-net method to the absolute estimate was presented. The sampling efficiency of sweep-net and cage method was equal; however, the former was considered superior because of the shorter sampling time. The distribution pattern of numbers of spittlebug adults in samples of ten sweep or the numbers in 1 m² of the pasture, generally fitted the Poisson series. Suggestions on choosing sampling sites in a sampling area were given.