LAUSR

With extraordinary levels of plant diversity and endemism, the Brazilian campos rupestres across the Espinhaço Range have a species/area ratio 40 times higher than the lowland Amazon. Although diversification drivers in campos rupestres remain a matter of debate, the Pleistocene refugium hypothesis (PRH) is often adopted as the most plausible explanation for their high diversity. The PRH has two main postulates: highland interglacial refugia and a species pump mechanism catalysed by climatic changes. We critically assessed studies on campos rupestres diversification at different evolutionary levels and conclude that most of them are affected by sampling biases, unrealistic assumptions or inaccurate results that do not support the PRH. By modelling the palaeo-range of campos rupestres based on the distribution of 1123 species of vascular plants endemic to the Espinhaço Range and using climate and edaphic variables, we projected a virtually constant suitable area for campos rupestres across the last glacial cycle. We challenge the great importance placed on Pleistocene climatic oscillations in campos rupestres plant diversification and offer an alternative explanation named escape-to-radiate model, which emphasizes niche shifts. Under this biogeographic model of diversification, the long-term fragmentation of campos rupestres combined with recurrent extinctions after genetic drift and sporadic events of adaptive radiation may provide an explanation for the current diversity and endemism in the Espinhaço Range. We conclude that long-term diversification dynamics in campos rupestres are mainly driven by selection, while most endemic diversity is ephemeral, extremely fragile and mainly driven by drift.