Study of Sampling Techniques in Enumerations in Forests

Investigations on the proper sampling technique and the sampling intensity for estimation of growing stock with a reasonable accuracy are in progress in Bombay State. Certain results are arising out of these investigations on the data of complete enumeration in one compartment of Dangs division, are discussed here. It was found that with strip sampling over larger areas, comprising a large number of compartments, the number of trees of teak per acre might be estimated with 4 per cent standard error by adopting eight per cent sampling intensity. While by two-stage random sampling, where certain number of strips are randomly selected in the first stage after dividing the forest area into strata and a certain number of plots are randomly selected from the selected strips at the second stage, the same precision could be obtained by five per cent sampling. In other words it was seen that random line-plot survey gave more precise estimates than strip survey for the same intensity of sampling. Unfortunately there is no data available to judge the relative efficiencies of the two methods from the point of view of cost.