Among the hypotheses invoked to explain high species richness in tropical forests, the niche differentiation hypothesis has received observational and experimental support (Ashton 1969, Chesson 2000, Clark et al. 1999, Souza & Martins 2004, Svenning 2001, Terborgh & Mathews 1999). Habitat specialization with regard to edaphic factors and topography has been observed in several plant groups including trees, lianas, shrubs, ferns and palms (Clark et al. 1999, Ibarra-Manriquez & Martinez-Ramos 2002, Svenning 1999, Tuomisto & Ruokolainen 1993). Treefall gaps and light gradients have also been found to be important ecological factors affecting plant distribution and niche differentiation in the understory (Chazdon 1986, Poorter & Arets 2003, Terborgh & Mathews 1999). It is suggested that such habitat heterogeneity may not only maintain biodiversity but also may lead to the origin of new species through the process of parapatric speciation (Gentry 1989, Haffer 1997, Patton & Smith 1992).
The limits to parapatric speciation 3: evolution of strong reproductive isolation in presence of gene flow despite limited ecological differentiation