A paradigm to be discarded: Geological and paleoecological data falsify the HAFFER & PRANCE refuge hypothesis of Amazonian speciation

Abstract

All geological data from Amazonian landforms imply continuous humid weathering throughout late Tertiary and Quaternary times, with all claims for arid land processes shown to be in error. Sand dunes exist only where thick deposits of sand prevent stable vegetative cover. A ground truth survey shows that proposed dune fields in the Pantanal do Mato Grosso do not in fact exist and that dunes in Pantanal Setentrional continue to be active. All available Amazonian pollen data, without exception and including new data, imply biome stability: no pollen data suggest increased coverage of savanna in glacial times, claims to the contrary being demonstrably in error. Amazonian climate is not monolithic, with secular climatic changes across the basin not in phase. New evidence shows that vegetation response to lowered temperatures, lowered CO2 and fluctuating dry seasons produced by MILANKOVITCH forcing resulted only in population changes within plant communities without biome replacements. Diversity between habitats within the forest provides vicariance for alternative evolutionary models. The "aridity with refuges paradigm" now impedes Amazonian research and should be discarded.