A1852
Title: Reliable Shannon entropy estimation from incomplete species inventories
Authors: Tsung-Jen Shen - National Chung Hsing University (Taiwan) [presenting]
Abstract: Shannon's entropy is widely used to quantify species diversity because it captures the overall complexity of community composition while balancing the contributions of rare and abundant species. In ecological surveys, however, many species remain undetected, especially when sampling effort is limited. Such incompleteness can lead to biased and unstable entropy estimates. A new estimator for Shannon's entropy that explicitly accounts for the missing portion of a species inventory is presented. The proposed method combines information from observed species and species frequency counts to improve estimation under incomplete sampling and provides uncertainty assessment via bootstrap resampling. Its performance is evaluated through simulation studies based on several empirical communities, including butterflies, beetles, small mammals, corals, and insects. Compared with eight existing estimators, the proposed estimator generally yields smaller bias and more stable performance, particularly when sample sizes are limited or moderate. These results suggest that the method offers a practical and reliable tool for biodiversity assessment when complete inventories are difficult to obtain.
