Dataset for: Paleoenvironmental models for Australia and the impact of aridification on blindsnake diversification
Data files
Jul 21, 2023 version files 136.35 KB
-
2023_blindsnakebiogeography_JBI.zip
-
README.md
Mar 14, 2024 version files 197.29 KB
-
2023_blindsnakebiogeography_JBI.zip
-
README.md
Abstract
Aim: Shifts in diversification rates of Australian flora and fauna have been associated with aridification, but the relationship between diversification rates and aridity has never been quantified. We employed multiple approaches to accurately reconstruct paleoenvironments of Australia for the first time. We used this information, and phylogenetic-based analyses, to explore how changes in temperature and increasing aridity during the Neogene influenced the diversification of the Australian blindsnakes. We tested whether diversification rates differ between arid-adapted and mesic-adapted lineages.
Taxon: Typhlopidae, Anilios blindsnakes
Location: Australia
Materials and Methods: We estimated the historical biogeography of blindsnakes using BioGeoBEARS. We synthesised multiple approaches to reconstruct paleotemperature and paleoaridity of Australia during the Neogene. We fit several birth-death models and estimate diversification rates under paleoenvironmental conditions using RPANDA. We further compare diversification rates between arid-adapted lineages versus mesic-adapted lineages using ClaDS and GeoHiSSE.
Results: Ancestral area estimation indicated Australian blindsnakes have tropical grassland origins. We found that Australia-specific regional paleotemperature and paleoaridity provided a better explanation for diversification rate variation than global paleotemperature. Specifically, our best-fitting model indicated that speciation rates of blindsnakes decreased with increasing aridity. We found no difference in diversification rates between arid- and mesic-adapted lineages.
Main conclusions: Soon after dispersing to Australia, the common ancestors of Australian blindsnakes diversified rapidly in mesic habitats during the early Miocene. However, as the continent became increasingly arid, diversification rates decreased. We found that shifts in the environment led to the emergence of two major clades: one remaining in primarily mesic habitats and the other adapting to the expanding arid biome. Our results emphasise the importance of both arid and tropical biomes as sources and sinks of diversification.
README: Data set for "Paleoenvironmental models for Australia and the impact of aridification on blindsnake diversification"
These data sets were used to perform analyses included in the research paper "Paleoenvironmental models for Australia and the impact of aridification on blindsnake diversification."
Main aims for the project:
- Estimate the historical biogeography of Australian blindsnakes using 'BioGeoBEARS.'
- Fit birth-death models and estimate diversification rates under different paleoenvironmental conditions using 'RPANDA.'
- Compare diversification rates between arid-adapted and mesic-adapted lineages using 'ClaDS' and 'hisse.'
Data and file structure
/2021_ALA_blindsnake_occurence_data/ -- this folder should be replaced with the 2021_ALA_blindsnake_occurence_data.zip on Zenodo (https://doi.org/10.5281/zenodo.7571419)
- Data from ALA can be downloaded at https://doi.org/10.26197/ala.d92678b1-ad2d-437b-9457-9f52737ba003
/bears_txt/ -- this folder contains .txt files necessary for fitting biogeographical models in' BioGeoBEARS'
- geofile.txt - geographic range for each species
- biome_distance.txt - modify distance for +x analysis
/intermediate_data/bears/ -- files in this folder are generated from code
- blindsnake_b.tre - phylogeny for fitting BioGeoBEARS models.
/intermediate_data/diversification_analyses/ -- file for 'RPANDA' analyses
- blindsnake.trees - contains two versions of Anilios trees.
/intermediate_data/geohisse/
- 2022_species_list_arid_nonarid_widespread.csv - includes information about which species are in the phylogeny to calculate fraction.
- arid_nonarid_both_states.csv - list of lineages and their geographic state. 0 = widespread, 1 = arid, and 2 = mesic. Designation of geographic states were based on distribution data and literature review.
- arid_nonarid_both_states.txt - .txt version of the previous file
/paleo_env -- this folder contains the reconstructed paleoenvironment data from three data sources as described in the paper.
- Australia_climate_data_45Mya_Scotese.csv - reconstructed from Scotese & Wright, (2018)
- Australia_climate_data_45Mya_Straume.csv - reconstructed from Straume et al., (2020)
- Australia_climate_data_45Mya_Valdes.csv - reconstructed from Valdes et al., (2021)
- Australia_climate_data_70Mya_Scotese.csv - reconstructed from Scotese & Wright, (2018)
- Australia_climate_data_65Mya_Straume.csv - reconstructed from Straume et al., (2021)
- climate_data_headers.txt - expalanation of climate data headers
/tree/ -- this folder contains the subset of phylogenies for Anilios.
- anilios_newick_b.tre - tree with A. splendidus
- anilios_newick_st.tre - tree with A. splendidus
- subset_anilios_newick_b.tre - tree without A. splendidus
- subset_anilios_newick_st.tre - tree without A. splendidus
Other information
Zenodo repository for "Data set: Australia's hidden radiation - phylogenomic analysis reveals rapid Miocene radiation of blindsnakes"
Code/Software
All scripts can be run using open source software.
- R is required to run R scripts (.R).
- Julia is required to run Julia scripts (.jl).
/Code
- 00_*.R - scripts are used for preparing data for analyses
- 01_fit_*.R - scripts were used to fit various RPANDA models. Note difference in initial lambda parameters for some models.
- 01_env_data_plots.R - script to plot different paleoenvironmental data
- 02_diversification_plots_b.R - script for plotting results
- 02_diversification_plots_b.R - script for plotting results
- 02_table_fit_env_results_b.R - script to summarise model fit
- 03_BioGeoBEARS_analyses_parallel.R - script for fitting multiple BioGeoBEARS models
- 04_BioGeoBEARS_results_bsm.R - Conduct Biogeographic Stochastic Mapping for the best fitting model
- 04_BioGeoBEARS_results_plots.R - plotting Biogeographic Stochastic Mapping from the best fitting model
- 04_BioGeoBEARS_results_tables.R - Biogeographic Stochastic Mapping result table from the best fitting model
- 05_GeoHiSSE_fit.R - fit GeoSSE and GeoHiSSE using 'hisse' package.
- 06_ClaDS2.jl - Julia script to estimate branch-specific rate under ClaDS
- 06_ClaDS_plot_tips.R - plot results from ClaDS
Contact
Should you have questions about these analysis scripts, please do not hesitate to contact Sarin Tiatragul (contact information can be found in the paper) or on Github (https://github.com/stiatragul)
Usage notes
Check the README.md for details about how to structure your analyses folder.
All scripts can be run using open-source software.
- R is required to run R scripts (.R).
- Julia is required to run Julia scripts (.jl).