thatif the sampling effort is increased to 100 plots, the expected num-ber of recorded species would be 56 species (lower and upper 95%confidence interval are 46 and 67 species), 64 species (95% con-fidence interval are 50 and 78 species) for 140 plots (twice oursampling effort), and 72 species (95% confidence interval are 51 and93 species) for 210 plots (three times our sample and the largestextrapolation recommended by Colwell et al., 2012) (Table 3). Usingthis approach, between 160 and 180 plots would need to be sam-pled in order to sample the asymptotic number of species in thearea, as estimated by the Jackknife 1 (68 species – the smallest esti-mated richness). Even if more than 210 plots had been sampled, thelargest estimated richness still would not have been achieved (Chao2 – 85 species) (Table 3; Fig. 4). The extrapolation curve suggestedthat to record the same number of species in the phytosociologicalsurvey as was recorded in the floristic list (i.e. 65 species), it wouldbe necessary to double our sampling effort (i.e. 140 m2surveyed).