Hidden Refugia: How West African Rainforests Survived Millions of Years of Climate Chaos

West African rainforests rank among the planet’s premier biodiversity hotspots, yet they have spent millions of years at the mercy of a wildly oscillating climate. As glacial cycles cooled and dried the tropics, vast stretches of forest gave way to savannah, leaving behind scattered islands of green where species clung to survival. A new review published in Frontiers of Biogeography synthesizes decades of evidence from fossil pollen, comparative biogeography, phylogeography, demographic modelling and species distribution modelling to build the most comprehensive picture yet of where these evolutionary refugia lie, how old they are, and what role they played in generating the extraordinary biodiversity of the Guinean Forests of West Africa.

The review defines rainforest refugia as regions of climatic stability where rainforests and their associated biota persisted through cold and arid periods, and from which species could later expand once conditions improved. These areas act as evolutionary lifeboats: they buffer populations against extinction, preserve deep lineages, and can seed the recolonization of landscapes after the climate relents. Understanding where refugia were located and how they functioned is not merely an academic exercise. Because refugia can mitigate the adverse impacts of contemporary climate change, they are increasingly recognized as priority targets for conservation planning.

The evidence for dramatic vegetation swings in West Africa comes most directly from fossil pollen cores extracted from marine and lake sediments. Deep-sea cores from the Gulf of Guinea, including one near the Nigerian coastline that traces vegetation change over the last ten million years, reveal repeated shifts from closed rainforest to more open vegetation as aridity intensified. Lake cores offer finer, more localized records; the longest continuous archive, from Lake Bosumtwi in southern Ghana, spans roughly 1.08 million years and documents the repeated local disappearance of rainforests during Pleistocene glaciations. Notably, no pollen records yet exist from West African highland regions, leaving a conspicuous gap in the palaeoecological record precisely where some hypothesized refugia may have sheltered forests that captured moisture from stratiform clouds and orographic rainfall.

When pollen data are combined with comparative biogeographic analyses of plants, mammals and birds, a consistent set of large-scale refugia emerges. Regions such as Mount Nimba, Cape Palmas, Cape Three Points and the Niger Delta have been repeatedly identified as areas of long-term climatic stability, and independent diversity surveys confirm elevated species richness and endemism in exactly these places. By contrast, the Dahomey Gap, a dry corridor receiving only around 700 millimetres of annual rainfall between southern-central Ghana and Benin, and the V-Baoulé savannah intrusion in central Côte d’Ivoire were never proposed as refugia, marking them as historically unstable zones that species likely abandoned during arid phases.

Phylogeography, which examines the spatial arrangement of genetic lineages within species, has largely corroborated these large-scale refugia while simultaneously revealing a hidden layer of structure. Across plants, reptiles, amphibians, mammals, birds and fish, recurrent phylogeographic breaks align with major rivers such as the Bandama, Volta, Niger and Cross, as well as with climatically unstable regions. These breaks partition West African rainforests into four well-established sub-regions: Western Upper Guinea, Eastern Upper Guinea, the Dahomey Gap and Western Lower Guinea. Crucially, fine-scale comparative studies of leaf-litter frogs and forest-dwelling rodents uncovered smaller isolated pockets, termed microrefugia, nested within and beyond the classic macrorefugia. Some of these microrefugia, notably in the Fouta Djallon highlands of Guinea and in Ghana, harbour lineages stretching back to the Miocene, suggesting stability over millions of years, while others appear to have persisted only through the more recent Pleistocene aridification cycles.

The timing of these evolutionary splits presents an intriguing puzzle. While phylogeographic patterns are broadly congruent across species, the divergence dates themselves are frequently asynchronous, even among co-distributed species analysed with the same genetic markers. Splits fall predominantly within the Pleistocene but vary widely, and the use of different markers with varying confidence intervals complicates direct comparison. This asynchrony implies that identical spatial patterns may arise from distinct underlying causes, whether climate-driven fragmentation, river barriers or ecological gradients, and underscores the need for standardized analytical frameworks before broad generalizations can be drawn.

To move beyond mapping refugia toward quantifying their evolutionary role, researchers have turned to demographic modelling and species distribution modelling. Demographic approaches compare observed genetic data against simulations run under competing models of diversification: the refugia model, in which populations diverge in isolation and later expand and meet again; the landscape barrier model, in which populations split in strict isolation on opposite sides of a physical obstacle; and the ecotone model, in which divergence occurs alongside continuous gene flow. Studies on West and Central African reptiles and amphibians predominantly support the refugia model, though it has been rejected in favour of barrier or ecotone scenarios for species with broader ecological tolerances, such as arboreal snakes that also inhabit dry forest-savannah mosaics.

Species distribution modelling offers an independent, mechanistic lens on the same question. By correlating species occurrences with environmental variables and projecting those niches back to the Last Interglacial period roughly 125,000 years ago, researchers have reconstructed how suitable habitat expanded and contracted across glacial cycles. The models consistently reveal significant range fluctuations for nearly all examined taxa, from frogs and chimpanzees to forest trees, with the notable exception of the ecologically flexible arboreal snakes. Climatic stability maps derived from these projections tend to incorporate the known refugia of Cape Palmas, Cape Three Points, Mount Nimba and the Niger Delta, while Western Upper Guinea and Western Lower Guinea emerge as extensively stable regions, in contrast to the more fragmented stability hotspots of Eastern Upper Guinea.

The review’s authors emphasize that refugia are not the sole engine of diversification in West Africa, but they played a key role in driving the evolution of rainforest-dependent species in combination with other factors such as river barriers. Ecological traits matter enormously: closely related tree species in the Dahomey Gap responded to the same aridification cycles with different timing of population splits and colonization events, depending on subtle differences in habitat preference and dispersal biology. This species-specific variability means that predicting how rainforest biota will respond to ongoing climate change requires understanding not just where refugia were, but how individual species interacted with them across time.

Looking forward, the authors call for research that spans a broader range of taxa, explores lesser-known regions such as the Fouta Djallon highlands and the Niger Delta, and adopts landscape demography approaches that account for the diverse forest habitats of West Africa, from evergreen lowland to montane forest. Advances in high-throughput sequencing, long-read technologies and integrative bioinformatic tools that combine genomic data with distribution models promise to sharpen the resolution of refugia dynamics across both space and time. As West African rainforests, which have already lost more than 80 percent of their former cover, face intensifying drought and rising temperatures, the lessons encoded in their refugial past may prove the most valuable guide to securing their future.

Subject of Research: Evolutionary refugia in West African rainforests and their role in biodiversity

Article Title: Towards a comprehensive view on evolutionary refugia in West African rainforests

Article References: Ernst, M., Rödel, M.-O., & Blom, M. P. K. (2025). Towards a comprehensive view on evolutionary refugia in West African rainforests. Frontiers of Biogeography, 18, Article e139537. https://doi.org/10.21425/fob.18.139537

Image Credits: AI Generated

DOI: 10.21425/fob.18.139537

Keywords: West Africa, rainforests, refugia, phylogeography, climate change, biodiversity, palaeoecology, species distribution modelling, demographic modelling, Guinean Forests, microrefugia, Pleistocene