Within the past million years, the Earth’s climate and biosphere have been strongly influenced by changes associated with the Pleistocene glaciations. In the context of the current concerns about extinction in the moist tropics, special attention should be paid to the effects on tropical diversity of the recent climatic fluctuations.

Many geologists and biogeographers have argued that the tropical rain forests of South America and Africa were largely replaced by dry savannas during the glacial advances. It has been postulated that the rain forests were reduced to a few small refugia, and the locations and extents of these remnant patches have been mapped in both South America and Africa (see Beven et al., 1984; Mayr and O’Hara, 1986; Simberloff, 1986).

If the refugium maps are accurate, they have profound implications for the effects of changes in tropical habitats. From theory, one would expect that total number of species would be reduced due to the greatly decreased habitable area for rain forest species and because of the elimination of the habitat of many geographically restricted species. The current estimates of present reductions in diversity caused by habitat destruction in the tropics are comparable to reductions estimated in the refugium model for the glacial intervals.

Also, if the refugium maps are accepted as reliable, it is difficult to explain the recovery of tropical diversity to present levels in the extremely short time since the last glacial advance. If present insect diversities are as great as recent estimates suggest, how did all the local endemics develop by speciation in such a short time?

There are major problems in applying the refugium model to the Pleistocene history of the tropics. The geological evidence for the climatic change comes mainly from scattered and generally inadequate data on fossil pollen. The biogeographical evidence is inferred from present-day distributions: the argument is that the refugia of the past are reflected now in concordant ranges of living species. That is, the near-coincident geographical ranges of species delineate the refuge patches from which diversification and geographical spreading occurred since the return of warm, moist conditions. There has been much argument in the recent biogeographical literature both for and against the refuge reconstructions. For both South America and Africa, strong cases have been made for opposing conclusions.

Another major problem with the refugium model is the extreme difficulty of documenting Pleistocene extinctions in the affected areas. The fossil record in present rain forest areas is notoriously poor because of the paucity of good rock exposures from which collections can be made. Furthermore, the organisms of most interest in this context—land animals, plants, and insects—have very low fossilization potentials and thus there are poor geological records, even under good circumstances. It is therefore difficult to determine from existing data whether or not the Pleistocene glaciations were accompanied by mass extinctions in the tropics. On a global scale, the Pleistocene epoch was not a time of mass extinction, but it is certainly possible that there were extensive species kills in rain forest areas.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement