IMPACTS OF MIOCENE CLIMATE CHANGE ON PALM TRAIT EVOLUTION IN MADAGASCAR AND MAINLAND AFRICA

Throughout the late Miocene (~10 Ma), climate change facilitated grassland expansion in the Afrotropics, greatly influencing patterns of modern-day plant biodiversity. Previous work has documented that a Miocene radiation shaped present-day palm species richness in continental Africa; however, little is known about how such climatic shifts have shaped the trait evolution and overall diversification dynamics of palms in Madagascar. Comparing traits of palms between these two regions, which continued experiencing biotic interchanges after separating ~170 Ma, can illuminate the processes that produce disparate evolutionary patterns.

In this study, we compiled seed length information from existing literature for palm species in Madagascar and mainland Africa and compared the fit of seven models of seed size evolution using the largest available species-level palm phylogeny. The best-fit model was an Ornstein-Uhlenbeck (OU) model, implying trait evolution toward an optimum, where each region had its own ideal seed size. The modeled seed size optimum for Madagascar is smaller than for Africa, suggesting that Malagasy palms have evolved toward smaller seed sizes over time. Ancestral trait reconstructions using the best-fit OU model suggest that divergence in trait evolution in the two regions may have occurred sometime in the past 20 million years, potentially coinciding with late Miocene grassland expansion. The species that radiated and evolved in each area in response to Miocene climatic change continue to shape modern differences in regional seed sizes.

We conclude that palms in Africa and Madagascar evolved during the Miocene in fundamentally different ways: in Africa, open-habitat palms pre-adapted to semi-arid grassland environments thrived as this biome became established. In contrast, in Madagascar, rainforest palms evolved smaller seeds to adapt to increasingly seasonal environments and speciated in forest fragments that were broken up by encroaching grasslands. We suggest that Madagascar’s greater topographic relief and smaller spatial scale could have played a key role in influencing the different evolutionary patterns observed. Our results emphasize the importance of considering the role of spatial heterogeneity and spatial scale in shaping evolutionary processes and outcomes.