Contrasting genetic signal of recolonization after rainforest fragmentation in African trees with different dispersal abilities

Abstract

Although today the forest cover is continuous in Central Africa, this may have not always been the case, as the scarce fossil record in this region suggests that arid conditions might have significantly reduced tree density during the ice ages. Our aim was to investigate whether the dry ice age periods left a genetic signature on tree species that can be used to infer the date of the past fragmentation of the rainforest. We sequenced reduced representation libraries of 182 samples representing five widespread legume trees and seven outgroups. Phylogenetic analyses identified an early divergent lineage for all species in West Africa (Upper Guinea) and two clades in Central Africa: Lower Guinea-North and Lower Guinea-South. As the structure separating the Northern and Southern clades—congruent across species—cannot be explained by geographic barriers, we tested other hypotheses with demographic model testing using δαδι. The best estimates indicate that the two clades split between the Upper Pliocene and the Pleistocene, a date compatible with forest fragmentation driven by ice age climatic oscillations. Furthermore, we found remarkably older split dates for the shade-tolerant tree species with nonassisted seed dispersal than for light-demanding species with long-distance wind-dispersed seeds. Different recolonization abilities after recurrent cycles of forest fragmentation seem to explain why species with long-distance dispersal show more recent genetic admixture between the two clades than species with limited seed dispersal. Despite their old history, our results depict the African rainforests as a dynamic biome where tree species have expanded relatively recently after the last glaciation.

Spatial distribution of Annonaceae across biomes and anthromes: Knowledge gaps in spatial and ecological data

“…Based on this data set Annonaceae, only avoid human impacts in areas with the least potential for agriculture (such as tropical areas with low fertility soils). We have virtually no information on what this means, for instance, for the genetic structure of Annonaceae for which only a few studies exist (Collevatti et al, 2014; Helmstetter et al, 2020; Migliore et al, 2019; Piñeiro et al, 2021). Furthermore, Annonaceae only contain a few economically important species (e.g., some in the genus Annona ; Larranaga & Hormaza, 2015 and Cananga odorata Benini et al, 2012), and therefore, their decline could easily go unnoticed (although locally they can be quite important and this impact might be felt much earlier).…”

Evidence for ecotone speciation across an African rainforest‐savanna gradient

“…The cyclical dynamics of rainforest cover are not in question. Strong evidence for periodic fragmentation of rainforest into refugia is found in longitudinal palaeoecological studies (Anhuf et al, 2006; Maley, 1996; Maley & Brenac, 1998), and detailed plant genetic investigations (Helmstetter et al, 2020; Piñeiro et al, 2021). The latter jointly uncover population structure among putative refugial areas, reductions in genetic diversity with distance from them, and expected post‐glacial population expansions out of them during warmer, wetter periods.…”

The origin of an extreme case of sister‐species sympatry in a palm‐pollinator mutualistic system

“…The DG is currently dominated by savannas, but it has undergone recurrent contractions and expansions during the Pleistocene due to the expansion of tropical rain forests during interglacials (e.g., Dupont & Weinelt, 1996). In forest species these climatic and environmental oscillations induced recurrent range shifts and fragmentations of populations, leaving strong genetic signatures (Piñeiro et al, 2021). Though the African oil palm is not a forest species, it was hypothesized that the climatic oscillations during the Pleistocene account for its current genetic structure (Cochard et al, 2009).…”