Login / Signup

Vanishing refuge? Testing the forest refuge hypothesis in coastal East Africa using genome-wide sequence data for seven amphibians.

Christopher D BarrattBeryl A BwongRobert JehleH Christoph LiedtkePeter NagelRenske E OnsteinDaniel M PortikJeffrey W StreicherSimon P Loader
Published in: Molecular ecology (2018)
High-throughput sequencing data have greatly improved our ability to understand the processes that contribute to current biodiversity patterns. The "vanishing refuge" diversification model is speculated for the coastal forests of eastern Africa, whereby some taxa have persisted and diversified between forest refugia, while others have switched to becoming generalists also present in non-forest habitats. Complex arrangements of geographical barriers (hydrology and topography) and ecological gradients between forest and non-forest habitats may have further influenced the region's biodiversity, but elucidation of general diversification processes has been limited by lack of suitable data. Here, we explicitly test alternative diversification modes in the coastal forests using genome-wide single nucleotide polymorphisms, mtDNA, spatial and environmental data for three forest (Arthroleptis xenodactyloides, Leptopelis flavomaculatus and Afrixalus sylvaticus) and four generalist (Afrixalus fornasini, A. delicatus, Leptopelis concolor and Leptopelis argenteus) amphibians. Multiple analyses provide insight about divergence times, spatial population structure, dispersal barriers, environmental stability and demographic history. We reveal highly congruent intra-specific diversity and population structure across taxa, with most divergences occurring during the late Pliocene and Pleistocene. Although stability models support the existence of some forest refugia, dispersal barriers and demographic models point towards idiosyncratic diversification modes across taxa. We identify a consistent role for riverine barriers in the diversification of generalist taxa, but mechanisms of diversification are more complex for forest taxa and potentially include topographical barriers, forest refugia and ecological gradients. Our work demonstrates the complexity of diversification processes in this region, which vary between forest and generalist taxa, but also for ecologically similar species with shared population boundaries.
Keyphrases
  • climate change
  • human health
  • genome wide
  • electronic health record
  • dna methylation
  • big data
  • south africa
  • machine learning
  • risk assessment
  • copy number
  • mitochondrial dna
  • drug induced
  • data analysis