Spatially-balanced sampling versus unbalanced stratified sampling for assessing forest change: evidences in favour of spatial balance