Get access

Gap percolation in rainforests


  • Ricard V. Solé,

  • Frederic Bartumeus,

  • Javier G. P. Gamarra

R. V. Solé and F. Bartumeus, Complex Systems Lab-ICREA, Universitat Pompeu Fabra, Dr. Aiguader 80, ES-08003 Barcelona, Spain ( RVS also at: Santa Fe Institute, 1399 Hyde Park Road, NM 87501, USA. FB also at: Centre d'Estudis Avançats de Blanes CEAB, CSIC, Accés Cala Sant Francesc 14, ES-17300 Blanes, Girona, Spain. – J. G. P. Gamarra, Center for the Environment, Dept of Natural Resources, 103 Rice Hall, Cornell Univ. Ithaca, NY, USA.


Rainforests biodiversity is sustained by the three-dimensional structure of their canopy which provides a wide range of physical microenvironments. Given the dynamic nature of the forest, the recognition of stable vertical layers or strata in the canopy is controversial. The spatial characterisation of potential habitats of understory species is not straightforward due to the complex structure of rainforest canopies and the wide ecological variability to which rainforest species can be adapted. Here we present a new description of potential understory habitats that give rise to a well-defined characteristic vertical scale of forest organization hc≈13 m. Species living in microenvironments occurring at canopy heights below this critical height hc can only experience landscapes with disconnected habitat patches (i.e. fragmented habitat landscapes), while those species capable of living also above hc will experience a fully connected landscape of suitable microenvironmental conditions. The possible implications for plant dispersal and animal colonisation strategies living at the understory or close-to-floor are discussed in relation to rainforest gap-dynamics, habitat loss and habitat fragmentation processes. Long-range and directed dispersal strategies (e.g. plant seed dispersal by animals) are optimal for those species living below hc, providing the best exploration of scarce habitats and a major robustness to habitat changes. On the other hand, dispersal strategies of those species capable to exploit habitats above hc need not to be based on directed long-range mechanisms. Different dispersal strategies may in turn imply different sensitiveness of species to habitat loss and habitat fragmentation processes in the rainforest.