Jackknifing in estimating species diversity of a logged-over dipterocarp forest in Surigao del Sur

摘要

The impact of the Jackknifing procedure as a tool to correct the bias of the usual estimation procedure for species diversity was studied through simulation. A test population using available data on relative abundance of the different species observed in four forest stands, under the seed tree method in the logged-over dipterocarp forest was used to draw samples through simulation for the comparison of the relative reduction in bias of the jackknifed estimates from sample-based estimates of the three diversity indices, namely, species richness, Shannon's index, and Simpson's index. Jackknifing resulted in the substantial reduction in the bias, which may exceed 50% for n >= 20, with the relative bias < 0.10. On the other hand, Jackknifing has led toan increase in the variance of the estimates. This, however, did not diminish the advantage of the Jackknifing procedure over the usual sample-based procedure, since the mean square error of the sample-based estimates were considerably reduced when jackknifed.