Abstract
Climatic-vegetational fluctuations due to astronomical Milankovitch cycles caused global changes in the
distribution of tropical forests and nonforest vegetation during the Cenozoic (Tertiary-Quaternary) and
before. Forest and nonforest biomes on the continents changed continuously in distribution during the
geological past, breaking up into isolated blocks and again expanding and coalescing under the varying dry
to humid climatic conditions. However, plant and animal communities disrupted and species changed their
distributions individualistically during the various climatic phases. Field data indicating vegetational
changes in Amazonia exist for the Quaternary;
The refuge theory postulates that extensive patches of humid rainforests persisted during dry periods
of the Tertiary and Quaternary, especially near areas of surface relief in peripheral portions of Amazonia,
where many extant species and subspecies of plants and animals probably originated. The humid 'refugia'
may have been separated by various types of savanna and dry forests as well as other intermediate
vegetation types of seasonally dry climates. The number and size of refugia during different dry periods
remain unknown. Biogeographic evidence for the former existence of forest refugia include areas of
endemism and sharply defined contact zones between species and subspecies of Amazonian forest birds
and other animals which represent zones of conspicuous biogeographic discontinuity in a continuous forest
environment.
Alternative models of barrier formation in Amazonia leading to allopatric speciation include the river
hypothesis, river-refuge hypothesis; canopy-density hypothesis, disturbance-vicariance hypothesis, museum
hypothesis and various paleogeography hypotheses, some aspects of which may be applicable to certain
periods in the evolution of the biota.
