Skip to main content
Dryad

Discordant phylogenetic endemism patterns in a recently diversified Brassicaceae lineage from the Atacama Desert: when choices in phylogenetic and species distribution information matter

Cite this dataset

Toro-Nuñez, Oscar; Lira-Noriega, Andres (2021). Discordant phylogenetic endemism patterns in a recently diversified Brassicaceae lineage from the Atacama Desert: when choices in phylogenetic and species distribution information matter [Dataset]. Dryad. https://doi.org/10.5061/dryad.djh9w0vwp

Abstract

Aim: Phylogenetic endemism (PE) has become a useful estimator of evolutionary history in conservation, given its capacity to estimate diversity hotspots as the combination of phylogenetic diversity (PD) and range size distributions of co-occurring taxa. However, potential limitations could preclude a general assessment of PE, especially in the presence of incongruent phylogenetic signals and the use of different estimates of species distribution at fine spatial scales. Here, we assess the utility of using PE in a recently diversified plant tribe.

Location: Atacama Desert, Argentina, Chile, and Peru.

Taxon: Schizopetalae tribe (Brassicaceae).

Methods: We assessed PD and PE to determine whether incongruent phylogenetic trees derived from nuclear and chloroplast DNA and different estimates of species distribution (based on species distribution modelling-SDM- and minimum convex polygons-MCP) impact the location of PE hotspots.

Results: Results show no influence from discordant phylogenetic signal between gene trees on PE estimates. Instead, the choice of the summary of species’ range (i.e., SDM or MCP) has greater influence on the location of PE hotspots.

Main conclusions: Despite the observed discrepancies, results in PE accumulation are congruent enough to support the use of this index in conservation-related analyses for the Atacama Desert flora. Nonetheless, further analyses are recommended to determine and contextualize the impacts of phylogenetic incongruence and species distributional estimates on PE, especially when other scenarios involving larger spatial and taxonomic sampling scales remain unexplored.

Usage notes

Files

Schizopetalae_Bayes_JBIOG_5_window_size.bps: Biodiverse file (tested on Biodiverse 2.0) which contains the results of all spatial indices presented in the manuscript, at 5 different neighbor window sampler sizes: 10, 20, 30, 40 and 50 km. Indices were obtained from the modules "Endemism" and "Phylogenetic Indices", as available in Biodiverse.

Schizopetalae_Bayes_JBIOG_5_GridCell_Size.bps: Biodiverse file (tested on Biodiverse 2.0) which contains the results of all spatial indices obtained with 3 different grid cell sizes: 25, 50 and 100 km. Indices were obtained from the modules "Endemism" and "Phylogenetic Indices", as available in Biodiverse.

Schizopetalae_Bayes_JBIOG_5_Randomizations.bps: Biodiverse file (tested on Biodiverse 2.0) which contains the results of randomization on all spatial indices obtained. Indices were obtained from the modules "Endemism" and "Phylogenetic Indices",  as available in Biodiverse.

Ratio_plots.xlsx: Excel file with ratios obtained from standarized PD and PE values (plotted in Fig. 4). The table presents a labele in each column header, which includes the index, the dataset and the resolution (e.g., PD_CV_30). Respective ratios are found in a respective third column, defined with a “r” letter (e.g., r_PD_30). The rest of the matrix includes coordinates for plotting the cell grid (Axis_0 and Axis_1) and the final columns present the labeled hotspots for each cell. At the end of each column, minimum, maximum and percentile values are calculated. All data was generated from Biodiverse 1.0 and it can be plotted in any GIS system.