A COMPARISON BETWEEN PLOT AND POINT SAMPLING USING A COMPUTER BASED TREE POPULATION OF Pinus caribaea SRI LANKA

Abstract

Research was conducted in the University Forest at Yagirala Forest Reserve, KalutaraDistrict, Sri Lanka, to compare the efficiency of point and plot sampling in Pinus caribaeaplantation using computer simulated sampling on a population of trees using data collectedin the field. In all, 3294 trees on 5.4 hectares constitute the population. The data base hasbeen filed with tree number, dbh, X and Y co-ordinates. Edge effect bias was minimizedusing the reflection method.

Random sampling was used in all cases for sample sizes n = 10, 20, 30, 60 and 100. Inorder to make meaningful comparisons, the concept of equivalent plot was used whichaimed at obtaining equal tallies per sampling unit for point and plot sampling. Basal areafactors applied were 2, 4, 9 and 16. Efficiency for given point - plot equivalents werebased on standard error % and cost values (where cost was based on time). In 19 of the 20point-plot combinations studied, point sampling was found to be the more efficient. Themost suitable BAF is BAF2 and it is recommended that point sampling be applied inplantation forest inventories in Sri Lanka where trained staff are available.