Comparison of species richness estimators.

<p><b>A–D</b> The Chao1bc (blue), ACE (grey), Bootstrap (green), Good-Turing (black), and negative-exponential estimators (orange) are applied to <i>in silico</i> random subsamples of observed data. Examples for HTLV-1, microbial, and TCR data are shown. Estimates systematically increase with sample size in datasets where rarefaction curves do not plateau (e.g. in <b>I</b>, <b>J</b>, <b>K</b>). Where rarefaction curves do plateau (e.g. in <b>L</b>), estimates are consistent. <b>E–H </b><i>DivE</i> (red) is applied to same subsamples as the other estimators. Performance of <i>DivE</i> was evaluated by comparing the error of estimates (<i>Ŝ_obs</i>), to the (known) number of species <i>S_obs</i> in the full observed data (purple line), i.e. error = |<i>S_obs</i> - <i>Ŝ_obs</i>| /<i>S_obs</i>. In all datasets, <i>DivE</i> accurately estimates the species richness of the full observed data from subsamples of that data. <b>I–L</b> Corresponding HTLV-1, microbial and TCR rarefaction curves: arrows denote the size of the subsample to which each estimator was applied.