Back

Phylogenetic biome niche conservatism underlies the evolution of forest paleoendemic legume trees in tropical Africa

Ojeda Alayon, D.; Gorel, A.-P.; Dauby, G.; Cervantes Arango, S.; Pan, A. D.; Marcussen, T.; Vanden Abeele, S.; Boom, A.; Forest, F.; de la Estrella, M.; Hardy, O. J.

2025-01-13 evolutionary biology
10.1101/2025.01.10.632308 bioRxiv
Show abstract

Niche conservatism is prevalent during the evolution of plant lineages. However, inferring biome niche lability and its impact on tropical tree species diversification is currently limited. To better understand biome niche lability and its effect on diversification rates, we analyzed an endemic lineage of African tropical trees, testing whether biome niche type (forest vs savanna) and biome lability are non-randomly distributed. We reconstructed a time-calibrated phylogeny of the Berlinia clade (16 genera, ca 201 species) using 140 nuclear genes, 75% of its extant species, and eight fossil calibrations. We also analyzed the phylogenetic signal of biome shifts across the group and its effect on diversification and extinction rates. We found the forest biome as the ancestral condition and a directional shift from forest to savanna (no reversals). The forest biome type is conserved in the Berlinia clade, but the lability to shift biomes is randomly distributed in the group. We did not find evidence that biome shift was associated with higher diversification or lower extinction rates. Our analyses identified five paleoendemic genera that have persisted solely in the forest biome since the Eocene. These paleoendemic and forest-restricted lineages are potentially more susceptible to habitat alterations of human related activities and future climate change, due to perhaps an innate limitation to adapt to new habitat types.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.