Because there are so many species in a rainforest, very few individuals of any single tree species will survive to adulthood. Thus, in the case of woody trees, there are few individuals per species (typically, one or fewer per hectare), but many species are represented per unit area. (Palm tree forests are a notable exception to this.) Gentry (1988) found 58 different species in a sample of 66 trees in a tiny 0.1 hectare plot of land in the Peruvian Amazon; of the first 50 trees sampled on a one-hectare plot of land, only two were of the same species! At most, one species may comprise 15% of the individuals in a given area. This is vastly different from temperate forests, which may consist of great tracts containing only one or a few species of trees. The scarcity of individuals of a single species can be advantageous, although it exacerbates the difficulty of cross-pollination, and since it reduces the probability of disease, pest and predator transmission. However, smaller trees of the same species and rare species tend to be more clumped than would be expected if distribution were random, whereas larger-diameter trees are less aggregated (Condit, et al., 2000). Non-woody plants such as herbs, climbers and rattans are usually more densely distributed throughout the forest than are trees.
A rainforest is a cohesive ecological unit, but a dynamic one, in which the proportion of different species will vary over time and space. Indeed, the composition of a rainforest can change dramatically within a very short distance, because of differences in altitude, soil type, water distribution and so on. Tuomisto, Ruokolainen and Yli-Halla (2003) attribute the distribution characteristics of two species of plants in Amazonian forests mainly to variations in environmental factors, such as soils, and, to a lesser degree, to limitations on dispersal mechanisms. They suggest that their results may be valid for many other species and other tropical forests, as well.