Assessing the efficiency of non-parametric estimators of species richness for marine microplankton

14 pages, 4 tables, 4 figuresNon-parametric asymptotic estimators rely on the assumption that rare species are indicative of the degree of undersampling. We evaluate the performance of 11 non-parametric asymptotic species richness estimators and an individual-based rarefaction and extrapolation (R/E) method using marine microplankton data. These species richness estimators were evaluated by sequentially increasing sampling effort. Their bias was diminished as sample (size) completeness increased. Jackknife estimators were more accurate than others such as Chao’s estimators. Underestimates were larger than 10% of species when applied to more than 30% of the individuals present in the community. The magnitude of bias varied as a function of community structure. R/E curves sorted samples by species richness but with a significant sampling error associated. We find that the use of these analytical methods lacks validity for the comparison of marine microplankton species richness estimates from sample collections spanning large spatial and temporal scalesThe Spanish Ministry of Economy, Industry and Competitiveness through a research grant CTM2011-25035, a Ramón y Cajal contract to P.C., and a Research Training Program FPI to M.B.Peer reviewe