Table 3. Fisher’s paradox.
Hyper-rare species [defined as species with fewer than 1000 individuals (24, 25)] and hyperdominant species (the most abundant species, accounting for ≈ 50 % of the total number of individuals) percentages were predicted in the whole area of each tropical forest obtained by applying both the NB and LS methods. We found that by using our NB method, the number of hyper-rare species in most of the forests was drastically reduced with respect to the LS method, thus suggesting that the extremely high value of hyper-rare species predicted in previous studies (24, 25) is an artifact of the LS method. Nevertheless, we found that the hyper-rarity phenomenon is a genuine emergent pattern in tropical forests.
| Forest | Hyper-rare (%) | Hyperdominant (%) | ||
| NB method | LS method | NB method | LS method | |
| Amazonia | 33 | 37 | 2.2 | 2.0 |
| Barro Colorado Nature Monument | 47 | 60 | 5.5 | 4.8 |
| Bukit Barisan | 22 | 46 | 7.9 | 1.9 |
| Bwindi Impenetrable Forest | 15 | 48 | 7.4 | 3.5 |
| Caxiuana | 6 | 49 | 10.3 | 3.2 |
| Cocha Cashu Manu National Park | 7 | 41 | 8.4 | 2.5 |
| Korup National Park | 9 | 51 | 9.3 | 3.1 |
| Manaus | 6 | 59 | 14.5 | 2.8 |
| Nouabalé-Ndoki | 4 | 43 | 11.2 | 2.4 |
| Pasoh Forest Reserve | 34 | 55 | 6.5 | 3.1 |
| Ranomafana | 12 | 49 | 7.5 | 2.7 |
| Udzungwa Mountain National Park | 12 | 48 | 6.3 | 3.0 |
| Volcan Barva | 8 | 52 | 10.5 | 2.5 |
| Yanachaga Chemillén National Park | 54 | 56 | 3.0 | 2.7 |
| Yasuni National Park | 39 | 74 | 11.6 | 4.4 |