Title: Using a generalized area-based truncated model to resolve Fishers paradox when extrapolating biodiversity
Authors: Youhua Chen - Chengdu Institute of Biology, Chinese Academy of Sciences (China) [presenting]
Abstract: Why are so many tropical tree species hyper-rare? Do they really have only one or two individuals on Earth? This question, the so-called Fishers paradox, was put forward by S.P. Hubbell when applying Fishers logseries to estimate tropical tree diversity. Herein, we developed an area-based truncated Fishers logseries model to partially, if not completely, resolve Fishers paradox by assuming that the occurrence of too-rare species is impermissible globally while being possible at local scales due to limited sampling efforts. An empirical test showed that alternative truncated models were indistinguishable at the local forest-plot scale, but they could be told apart at the regional scale. By comparison, a protracted speciation neutral model had similar behaviors. However, the exceptional merit of the truncated model is that by using a small truncation threshold, the prediction of regional species richness was similar to the value predicted by the original Fishers logseries, while completely excluding the possibility of the occurrence of too-rare species. Given the issue of the inability to distinguish among alternative models at the local scale, the truncation threshold might be pre-set by referring to real-world population sizes of trees. Alternatively, the threshold can be estimated if sufficient local biodiversity data are provided.