Figure 2

Graphical representation of observed statistics and their neutral confidence intervals for each of the analysed genes. (a) Tajima's D statistic; (b) Fu and Li's F* statistic; (c) Fu and Li's D* statistic; (d) Fu's F _S statistic; (e) Nei's genetic diversity H. A, Carapa guianensis population SE; B, Carapa guianensis population NW; C, Eperua falcata, population NW; D, Eperua falcata, population SW; E, Eperua grandiflora; F, Virola sebifera; G, Pachira quinata. For each species and population, genes are the same as in Table 3. For C and D, genes appear in the order: EfaPIP1.1, EfaPIP1.2, EfaPIP2.1. Filled squares: observed values for each test (red, green = significant, P < 0.05; orange = marginally significant, 0.10 <P < 0.05). Dashed lines: neutral confidence intervals. Values exceeding the upper limit of the confidence interval (red) indicate balanced selection or bottleneck; values lower than the lower limit of the confidence interval (green) indicate negative selection or population expansion (in (e), green squares indicate significant values after Bonferroni correction). "No polymorphism": this population did not show any sequence variation for this gene.