Phylogenomic relationships and historical biogeography in the South American vegetable ivory palms (Phytelepheae)

Published: 14 September 2021| Version 2 | DOI: 10.17632/jsd8j7wj97.2
Contributors:
Sebastián Escobar,
,
,
,

Description

Data for phylogenomic analyses in the vegetable ivory palms (Phytelepheae). Please read the "read me" file and check the published preprint for detailed methods https://doi.org/10.1101/2020.09.03.280941. ABSTRACT: The palm tribe Phytelepheae form a clade of three genera and eight species whose phylogenetic relationships and historical biogeography are not fully understood. Based on morphological similarities and phylogenetic relatedness, it has been suggested that Phytelephas seemannii and Phytelephas schottii are synonyms of Phytelephas macrocarpa, implying the existence of only six species within the Phytelepheae. In addition, uncertainty in their phylogenetic relationships in turn results in blurred biogeographic history. We inferred the phylogenomic relationships in the Phytelepheae by target-capturing 176 nuclear genes and estimated divergence times by using four fossils for time calibration. We lastly explored the biogeographic history of the tribe by inferring its ancestral range evolution. Our phylogenomic trees showed that P. seemannii and P. schottii are not closely related with P. macrocarpa, and therefore, support the existence of eight species in the Phytelepheae. The ancestor of the tribe was widely-distributed in the Chocó, Magdalena, and Amazonia during the Miocene at 19.25 Ma. Early diversification in Phytelephas at 5.27 Ma could have occurred by trans-Andean vicariance after the western Andes uplifted rapidly at ~10 Ma. Our results show the utility of phylogenomic approaches to shed light on species relationships and their biogeographic history.

Files

Steps to reproduce

Data and results of phylogenomic analyses in the vegetable ivory palm (tribe Phytelepheae) and related species. One hundred and seventy six genes were sequenced in 35 individuals. DNA alignments where at least 75% of the gene length was reconstructed in at least 75% of individuals, were retained (75_75). In addition, genes with signs of paralogy were discarded. The rest of alignments were used for phylogenomic analyses using Bayesian inference in BEAST and multi-species coalescence inference in ASTRAL-III. BEAST 75_75 alignments: Here are stored the 133 DNA alignments where at least 75% of the gene length was reconstructed in at least 75% of individuals and were cleaned using the software Gblocks. Sixteen of these genes showed signs of paralogy and are within the folder paralogs, leaving 117 usable DNA alignments. Nexus alignments: Here are stored the alignments of the 32 most clock-like genes , which were transformed to nexus format for Bayesian phylogenomic analyses in BEAST. BEAST: The file beast_outgroup3_5cal_UNIFORM_300M was used in the software BEAST, using the 32 Nexus alignments and five calibration points. The results from three independent runs in BEAST are stored in the folders beast_1, beast_2, and beast_3. The file tree_combiner_10% is the combination of the BEAST results with 10% of burn-in in the software LogCombiner. The file tree_10% is the resulting phylogenomic tree obtained using the software TreeAnnotator. ASTRAL_III 75_75 alignments: Here are stored the 133 DNA alignments where at least 75% of the gene length was reconstructed in at least 75% of individuals and were cleaned using the software Gblocks and Trimal. Sixteen of these genes showed signs of paralogy and are within the folder paralogs, leaving 117 usable DNA alignments. Trees: 117 Phylogenetic trees constructed from the 117 DNA alignments using the software RAxML. ASTRAL-III: The 117 RAxML trees were placed in the single file all. The branches from these trees with posterior support < 10 were collapsed, and these new trees are in the file all_bs10. This file was run in ASTRAL-III annotating the posterior probability, resulting in the final coalescence tree astral_all_bs10_PP.