STATISTICAL AND COST EFFICIENCY OF SIMPLE RANDOM SAMPLING AND SYSTEMATIC SAMPLING IN ESTIMATING FOREST STAND PARAMETERS AT YAGIRALA

Abstract

The study was carried out at a 24 years old Pinus rurbaca plantation located in lowlandwet zone of Sri Lanka. Statistical and cost efficiency were compared at samplingintensities or 11.1 %. 1l).4(/rJ and 32.4%. using simple random sampling design andsystematic sampling design. Circular sampling units of 0.0."> ha with slope correction. andboundary adjustment were also applied. Sampling units of simple random sampling werearranged in a hexagonal pattern. and selected (without replacement) after generatingrandom numbers. Diameter at breast height and total height within each sampling unitwere measured 10 estimate hasal area per ha and volume per ha. Walking time andmeasuring lime were recorded 10 calculate the variable costs of the inventory of all speciesat 32.4% sampling intensity.

The pine population shows normal distribution. and simple random sampling represents itmore closely than systematic sampling. Naturally regenerated tree species under pineplantation arc effecting the diameter distrihution.

S irnplc random sampling gives hcucr estimates of basal area per ha and volume per ha forpine than systematic sampling. For all species. systematic sampling gives better estimatesat lower sampling intensities.

III simple random sampling. the most efficient sampling intensity is 32.4%. However.11.1 % can be recommended for 2.">% sampling errol". In all species. relative efficiency ofthe two sampling designs does not show a significant difference. Simple random samplingis more efficient lor all species at 32.4% intensity.

The delayed silvicultural operation is found to effect the pine population and there is aneed for proper management of the forest,