To estimate the occurrence of spittle bug spumes by the quadrat method, the definition of the number and the size of quadrat to achieve a determined degree of precision is necessary. In five São Paulo State localities (Brazil) areas of one hectare growíng, Brachiaria decumbens (Stapf) were divided into 25 plots measuring 20m x 20m each one. In each of these plots, quadrats of two sizes (0.25m x 0.25m and 0.50m x 0.50m) were thrown at randon and the occurrence of spittle bug spumes was evaluated. The variance componente method was used to estimate the sample size that allow to obtain the average number of spumes por ares with a bound on the error of estimation of 0.10 y. The most efficient number of quadrat for determining the number of spumes varied with the level of infestation. One hundred quadrats of 0.25m x 0.25m or forty quadrats of 0.50m x 0.50m are reasonable sample sizes to estimate the average number of spumes per one hectare.
variance component estimate; confidente interval for the mean; average number of spumes; damage evaluation; goodness of fit