Rarefaction, richness estimation and extrapolation methods in the evaluation of unseen plant diversity in aquatic ecosystems

• We present one of the first phytosociological surveys on aquatic plant communities in Brazil.
• We use the studied ecosystem as a model to assess sampling issues in aquatic ecosystems.
• We used rarefaction, extrapolation and richness estimators to evaluate sampling effort.
• We show how this mathematical framework can be employed as a sampling evaluation tool by field botanists.

Defining the adequate sampling effort necessary in tropical ecosystems is a sensitive subject. It is virtually impossible to sample every single species in an ecosystem. Thus, sampling must be a compromise between the desired completeness of the dataset and the amount of time and work that is feasible for researchers to do during fieldwork. Here, we used rarefaction procedures and richness estimators to evaluate sampling issues in a coastal lagoon in Ceará state, Brazil. We used 70 plots (1 m2 each) to perform a phytosociological study in the lagoon and complemented the data with opportunistic floristic observations. We then applied statistical methods to the phytosociological survey to estimate how much of the local plant diversity had been sampled. Extrapolating our sampling accumulation curve to a sampling effort three times larger than the actual sampling done (210 plots), we were able to estimate the cumulative richness expected. After this, we used a different approach and calculated how many plots would be necessary to record a certain proportion of the total plant community. We recorded 49 species using phytosociological plots and 65 species in the floristic compilation. Richness estimators calculated the local alfa diversity to be between 68 and 85 species. We show that sampling 80% of the total local richness would require 100 plots, which is a reasonable sampling effort, but to sample 99% of the species, an extremely large amount of sampling would be needed (580 plots). This is not practical during fieldwork. We show here how rarefaction and extrapolation can be used as a framework to gauge the amount of field effort needed in aquatic ecosystems.