Many attempts have been made to determine and make projections of species extinction rates in rainforests, but this is extremely difficult. First of all, we know very little about how many species exist in tropical rainforests. Only about 500,000 rainforest species have been described, although these forests hold at least two-thirds of the world’s species (Raven, 1988). The number of insect species alone is surely one million and probably much higher.

a. The species-area relationship: The most-utilized method of estimating species loss rates is based on the “species-area” relationship. The species-area relationship establishes a correlation between the number of species in a given area and the size of that area, and is based on studies done on islands. Generally, the number of species increases in proportion to the size of the island by between the third and the fifth root of the area. Under natural conditions, the number of species on an island is fairly constant because there is continual natural immigration, which more-or-less balances the number of extinctions. However, if the area is reduced in size, species diversity will be reduced, and a new equilibrium will be attained. This method assumes that extinctions are in some way proportional to the loss of area for any particular system. Generally speaking, the smaller the area, the greater the extinction rate, because populations will also be small and have low genetic diversity. Under these circumstances, changes in the environment can easily lead to extinction, particularly of those species which have low adaptational ability. For example, for islands of 1 – 25 km2 area, the extinction rate has been estimated as between 10% and 50% over time. As a very general rule of thumb, when 90% of an ecosystem has been disrupted in an isolated area, 50% of its species will ultimately be lost (Raven, 1988). (This relationship can be described by the equation: S = CAz, where S = number of species, C = a constant, A = area, and z = a biologically-related constant which depends on the group of organisms involved and other factors. The value of z ranges from 0.15 to 0.35. To obtain the relationship mentioned above, z = 0.3) (May and Stumpf, 2000).

b. Estimates of species loss: When dealing with rainforests, the further assumption is made that when they are cut into fragments (by logging and conversion to agriculture or ranching), the fragments can be treated like islands, since they are separated from other forested areas. The same relationship is thought to exist between species number and area in forest fragments as is found in islands. Since tropical forests are being destroyed at rates between 0.8% and 2.0% per year, one could estimate that rainforest species are being lost at a rate which corresponds to the relationship described by the equation in the previous section. Wilson (1989; see also Ehrlich and Wilson, 1991) estimates that between 0.2% and 0.3% of rainforest species are being lost annually (4000 – 6000 species), if one assumes that 1% of the area of rainforests is being cut per year. These estimates are necessarily inexact since, as noted above, we have so little idea of how many species exist in tropical forests, so little idea of their ranges, and so little idea of what size forest fragments need to be to support these species. Thus far humans have eliminated approximately 50% of previously-existing rainforest, and about 15% of rainforest species may have already become extinct. Since many others are being greatly reduced in number by habitat destruction, they will eventually also become extinct due to loss of habitat or inability to reproduce. At current rates of forest destruction, extinction rates are expected to accelerate until about the middle of the 21st century and then decline, because so few species will remain. Of the tropical forest “biodiversity hot spots” (containing endemic species), only 12% remain intact. Even should all of these be completely protected, an estimated 18% of their species would become extinct. But, if current rates of destruction continue, these hotspots will lose approximately 40% of their species within the next century (or 100%, should the forest be completely removed, of course). In the Atlantic forest of Brazil, only 5% of the original forest remains intact, mostly as small fragments surrounded by agricultural fields. According to Cardoso and Tabarelli (2000), 34% of the tree species remaining in this area will become extinct in the near future.

Certain ecologists, however, find fault with this scenario. They note a lack of scientific data on extinctions (and a reliance on anecdotal evidence). They point out that deforestation is not necessarily equivalent to habitat loss, since some organisms can migrate elsewhere and secondary forest may replace the original primary forest. Thus, removal of a forest may not inevitably doom all of its species to extinction. Puerto Rico was deforested around the turn of the twentieth century and, although there were some species losses, most did not vanish, and the island now has a considerable cover of secondary forest.

However, statistical analyses of species extinctions have long suggested a relationship between biodiversity (number of species) and area. If this is generally true, then we must concentrate our efforts not only to save habitats, but to retain as large an area under rainforest as possible. Rosenzweig (1999) suggested that, because of human usurpation of most of the terrestrial parts of the globe, mass extinctions will follow, but in three phases. The first will be immediate, in which species endemic to particular habitats entirely co-opted by humans will vanish. The second phase will be extinctions on the “island” model, during which species will disappear as they are increasingly isolated into “island-like” areas where they cannot maintain their population size (what the author calls “sink” populations). This will occur over centuries, but will result in the loss of approximately half of present-day biodiversity. The third phase of extinction will take hundreds or thousands of years. If we continue to wrest territory from nature, diversity will eventually decline to a low steady-state level, perhaps to 5% of current biodiversity, if, as the author calculates, humans utilize and exclude nature from 95% of the global terrestrial area.

Whatever extinctions are occurring and will occur in the future, they are not random. Certain types of organisms will be more susceptible to extinction under current conditions than others. Highly specialized or localized species, or those requiring undisturbed primary forest or large areas for forage, are the most vulnerable. Because of this, entire groups of such species will become extinct, skewing the distribution of organisms and reducing the ability of ecosystems to recover.