Richness estimation with species identity error

Abstract

Richness estimation of an interesting area is always a challenge statistical work due to small sample size or species identity error. In the literatures, most richness estimators were only proposed to tackle the underestimation of the size-limited sample. However, species identity error almost occurs in each species survey and seriously reduces the accuracy of observed, singleton, and doubleton richness in turns to influence the behavior of richness estimator. Therefore, to estimate the true richness, the biased collected data due to species identity error should be modified before processing the richness estimation work. In the manuscript, we propose a new approach to correct the bias of richness estimation due to species identity error. First, a species list inventory from a subplot obtained by the investigator was used to estimate the species identity error rate. Then, we can correct the biased observed, singleton, and doubleton richness of the raw sampling data from the interesting area. Finally, the richness estimators proposed in the literature could be supplied to get the more correct estimates based on adjusted observed data. To investigate the behavior of the proposed method, we performed simulations by generating data sets from various species models with different species identity error rates. For the purpose of illustration, the real data was supplied to demonstrate our proposed approach. A presence/absence weeds species was surveyed in the organic farmland located at Soft Bridge County in the North of Taiwan.