Rates of extinction are very difficult to estimate, because we don't even know within an order of magnitude how many species there are. Nonetheless, we can be quite sure that a large proportion of taxa are threatened with extinction. The Nature Conservancy  assessed the status of 20,892 species in groups that are taxonomically well-characterized in the United States (freshwater mussels, crayfishes, vertebrates, vascular plants, tiger beetles, butterflies/butterflies, and dragonflies/damselflies; Table 3).
These statistics don't tell us how many extinctions are happening. And depending on whether you're an optimist or a pessimist you can take this as good news or as bad news. Two out of three species in the United States are apparently secure. That's the good news. One out of three species in the United States is vulnerable or imperiled. That's the bad news.
But how do these numbers compare with what might have characterized the biota of the U.S. prior to the arrival of humans? Do they, in fact, tell us that extinction rates are elevated, or are they simply characteristic of what we would have found at any time in the geological past? Are a large fraction of species on the edge of extinction during most of geological history? Two different approaches have been employed to try to compare current rates of extinction with those inferred from the fossil record.
Statistics on the rates of extinction have mostly been inferred from a species-area relationship and projections of habitat destruction.
S = CAz
If we're only interested in the proportion of species remaining after some portion of the habitat has been destroyed
is generally between 0.15 and 0.35. If current rates of tropical deforestation continue for another 30 years half of the remaining rain forest will be gone. If we then prevent all further deforestation, between
0.50.35 = 0.78
0.50.15 = 0.90
of the species presently there will remain, i.e., 10%-20% will go extinct. More sophisticated approaches use an exponential decay model, recognizing that the predictions of the above approach apply only after the remaining forest fragments have reach an immigration-extinction equilibrium. Notice, however, that these calculations only reflect the effects of lost habitat, not increased exposure to disease, competition with invasive exotics, overxploitation, or habitat degradation.
It may come as a surprise to you to learn that we can actually get reasonable estimates of current extinction rates from examining documented extinctions in groups that are reasonably well-studied. In the United States alone, for example, 45 vertebrates (over half of which are birds), 347 invertebrates, and 147 plants are either presumed or possibly extinct .3 By calculating the fraction of known species that have gone extinct in historical times, we get a direct estimate of extinction rates. The calculations that follow are based on the data and the approach described in [7,12].
The rates calculated by these very different approaches are reasonably comparable, given the great uncertainties involved. They suggest that contemporary rates of extinction are 100 to 1,000 (and possibley 10,000) times higher than at any time in the last 65 million years. The figures in Table 4 show how many documented extinctions have occurred in the last 400 years, along with an estimate of the corresponding median lifetime of taxa in each group (derived from ).
Critics of estimates based on species-area relationships have pointed out that 95% of the eastern forests in North America were cut while only 4 bird species went extinct4. How can we account for this apparent discrepancy ?
- The region was not simultaneously deforested. As agriculture moved westward, forest reclaimed abandoned fields. Forests always covered 50% of the land area.
0.50.25 = 0.84
16% or 25-26 out of 160 forest species should have gone extinct.
- May not have been enough time for extinction. Species-area relationship is an equilibrium expectation.
Committed to extinction versus actually extinct
- Only 28 of the 160 species are restricted to eastern forests. The rest are widely distributed elsewhere and would have persisted there. They could have later moved back into eastern forests.
(0.50.25) (28) = 23.5 ,
i.e., 23-24 of the 28 species should have persisted. Only 4-5 should have gone extinct, which is about what we saw.
- We lost few bird species from eastern North American forests because we had few endemics to lose.