A New Proposed Estimator for Reducing Bias Due to Undetected Species

Abstract

The present paper addresses a new approach to reduce bias when there are undetected species in a plot. Partially density matrix plays essential role in this new proposed estimator. The performance of the new proposed estimator (Ĥ0) was compared to bias-corrected MLE (MLEBC), Jackknife (JK) and the proposed estimator of Chao and Shen (Ĥcs) using Principle component analysis (PCA). The result of the first PCA applied to the data including the estimators’ values of the assemblages showed that Ĥ0 is located between JK and Ĥcs and its’ nearest neighbor becomes JK. The second PCA was applied to the data belonging to the relative estimator values between the pairwise assemblages and, it was found that Ĥ0 is still located between JK and Ĥcs but its’ nearest neighbor becomes Ĥcs in this time along the first axis. Those results were evaluated that Ĥ0 is a better estimator than MLEBC. Thus the new proposed estimator (Ĥ0) can also be used as an alternative bias-corrected estimator in addition to the other estimators.