Open Access Research article

Large-scale pattern of genetic differentiation within African rainforest trees: insights on the roles of ecological gradients and past climate changes on the evolution of Erythrophleum spp (Fabaceae)

Jerome Duminil12*, Richard P Brown3, Eben-Ezer BK Ewédjè14, Patrick Mardulyn1, Jean-Louis Doucet5 and Olivier J Hardy1

Author Affiliations

1 Service Evolution Biologique et Ecologie, CP160⁄12, Faculté des Sciences, Université Libre de Bruxelles, 50 Av. F. Roosevelt, 1050 Brussels, Belgium

2 Bioversity International, Forest Genetic Resources Programme, Sub-Regional Office for Central Africa, P.O. Box 2008 Messa, Yaoundé, Cameroon

3 School of Natural Sciences and Psychology, Liverpool John Moores University, Liverpool L3 3AF, UK

4 Faculté des Sciences et Techniques, BP 4521, Université d’Abomey-Calavi, Abomey-Calavi, Benin

5 Laboratoire de Foresterie des Régions Tropicales et Subtropicales, Unité de Gestion des Ressources Forestières et des Milieux Naturels, Gembloux Agro-Bio Tech, Université de Liège, Passage des Déportés, 5030 Gembloux, Belgium

For all author emails, please log on.

BMC Evolutionary Biology 2013, 13:195  doi:10.1186/1471-2148-13-195

Published: 12 September 2013



The evolutionary events that have shaped biodiversity patterns in the African rainforests are still poorly documented. Past forest fragmentation and ecological gradients have been advocated as important drivers of genetic differentiation but their respective roles remain unclear. Using nuclear microsatellites (nSSRs) and chloroplast non-coding sequences (pDNA), we characterised the spatial genetic structure of Erythrophleum (Fabaceae) forest trees in West and Central Africa (Guinea Region, GR). This widespread genus displays a wide ecological amplitude and taxonomists recognize two forest tree species, E. ivorense and E. suaveolens, which are difficult to distinguish in the field and often confused.


Bayesian-clustering applied on nSSRs of a blind sample of 648 specimens identified three major gene pools showing no or very limited introgression. They present parapatric distributions correlated to rainfall gradients and forest types. One gene pool is restricted to coastal evergreen forests and corresponds to E. ivorense; a second one is found in gallery forests from the dry forest zone of West Africa and North-West Cameroon and corresponds to West-African E. suaveolens; the third gene pool occurs in semi-evergreen forests and corresponds to Central African E. suaveolens. These gene pools have mostly unique pDNA haplotypes but they do not form reciprocally monophyletic clades. Nevertheless, pDNA molecular dating indicates that the divergence between E. ivorense and Central African E. suaveolens predates the Pleistocene. Further Bayesian-clustering applied within each major gene pool identified diffuse genetic discontinuities (minor gene pools displaying substantial introgression) at a latitude between 0 and 2°N in Central Africa for both species, and at a longitude between 5° and 8°E for E. ivorense. Moreover, we detected evidence of past population declines which are consistent with historical habitat fragmentation induced by Pleistocene climate changes.


Overall, deep genetic differentiation (major gene pools) follows ecological gradients that may be at the origin of speciation, while diffuse differentiation (minor gene pools) are tentatively interpreted as the signature of past forest fragmentation induced by past climate changes.

Cluster analysis; Fragmentation; Last glacial maximum; Phylogeography; Range expansion; Refugium; Species delimitation; Tropical rainforest