species is too large, to have a general rule of thumb as to when to stop sampling. In the present study, we applied a framework base don species accumulation curves, richness estimators and sample based extrapolation (Col well et al., 2012) to allow aquatic botanists to evaluate the optimum sampling efforts at their study sites.Our results suggest that in the aquatic ecosystems occurring in the semiarid regions of Brazil, a reasonable sampling effort would be between 100 and 200 plots (1 m 2 each). We believe this is an achievable number, as it could be accomplished in a reasonable amount of time, and it would sample approximately 80% of the local plant community according to the adopted procedures.In the species accumulation and richness estimation curves(based in 70 plots – Fig. 3), there was a clear positive relation-ship between the number of sampling units and the estimated species richness (i.e. the larger the sampling, the larger the richnessestimated). This suggests that estimators are being conservative(underestimation bias – Brose et al., 2003), and that a largersampling size would probably result in a larger predicted richness.