STRATIFIED DOUBLE SAMPLING OF PATCHY VEGETATION TO ESTIMATE BIOMASS

A double sampling scheme is proposed, and is shown to be cost-efficien t, for the estimation of vegetation biomass components in regions that are divided by plant species into relatively homogeneous subregions. Formulae for the mean and variance, and conditions required for the ap proximate multivariate normality of the estimators, are given. These a re applied to the interval estimation of above-ground biomass componen ts for a site consisting of shrub and grass, in such a way as to be su itable for input into fire-behaviour models.