there were no glaciers at low elevations in the tropics -- but the climate nevertheless changed dramatically during the Pleistocene

in tropical regions, dry periods coincided with glaciation in North America (a cooler atmosphere resulted in less evaporation from tropical oceans and hence less rainfall on nearby continents) -- lowland forests in Amazon basin were restricted to areas with highest rainfall -- populations of forest birds survived in these forest refuges -- during interglacials, as rainfall increased, forests spread across the entire Amazon basin -- populations that had differentiated while isolated came into secondary contact

in the Neotropics (New World tropics), closely related species of toucans provide one example (among many) of secondary contact between populations that spread from forest refuges during dry periods -- sometimes populations hybridize where they meet, sometimes not (hybrids indicate that they have not evolved reproductive isolation) -- sometimes populations overlap where they meet, sometimes not (failure to overlap indicates that they do not differ enough ecologically to coexist)

so in tropical regions, just as in temperate regions, changing climatic conditions have produced shifting connections between "islands" of habitat and thus opportunities for allopatric speciation

molecular differences between populations of birds in the tropics are even greater than those in North America -- so there is an even greater discrepancy between the molecular-clock hypothesis and the Pleistocene-refuge hypothesis in the tropics than in the temperate zone!

once again, there's always more to learn!