Dataset DOI: 10.5061/dryad.8931zcs30

Description of the data and file structure

The data archive comprises complementary components, including

• Curated GenBank accession tables organized by clade and gene region;
• Joint biogeographic and diversification analyses (geographic‐range & biome delimitations, RevBayes scripts);
• Cold-habitat connectivity assessments (paleoclimate-informed dispersal matrices and global connectivity metrics).

Files and variables

File: Biogeographic_analysis.zip

Description: This folder contains the input data (delimitations of geographical range and biome, time-calibrated phylogenies), RevBayes scripts for the biogeographic model that jointly estimates the range and biome evolution, outputs from ancestral biogeographic reconstructions and bootstrap replicates.

The directory tree structure is visible bellow:

├── events-bootstrap1000.csv.gz
├── MaxRange3
│   ├── 2_parse-tensor-json.py
│   ├── 3_bootstrap_parse.py
│   ├── area-adjacency.csv
│   ├── bitstates.py
│   ├── clades
│   │   ├── Alchemilla
│   │   │   ├── Alchemilla.smap.log
│   │   │   └── inputs
│   │   │       ├── 01_make-revscript-tensor.py
│   │   │       ├── Alchemilla.nex
│   │   │       ├── Alchemilla_classe_tensor.rev
│   │   │       ├── area-adjacency.csv
│   │   │       ├── biorange.csv
│   │   │       ├── bitstates.py
│   │   │       ├── classe-tensor-template.rev
│   │   │       └── model4.py
│   │   ├── Allium
│   │   │   ├── Allium.smap.log
│   │   │   └── inputs
│   │   │       ├── 01_make-revscript-tensor.py
│   │   │       ├── Allium.nex
│   │   │       ├── Allium_classe_tensor.rev
│   │   │       ├── area-adjacency.csv
│   │   │       ├── biorange.csv
│   │   │       ├── bitstates.py
│   │   │       ├── classe-tensor-template.rev
│   │   │       └── model4.py
│   │   ├── ...
├── MaxRange4
│   ├── 2_parse-tensor-json.py
│   ├── 3_bootstrap_parse.py
│   ├── area-adjacency.csv
│   ├── bitstates.py
│   ├── clades
│   │   ├── Artemisia
│   │   │   ├── Artemisia.smap.log
│   │   │   └── inputs
│   │   │       ├── 01_make-revscript-tensor.py
│   │   │       ├── Artemisia.nex
│   │   │       ├── Artemisia_classe_tensor.rev
│   │   │       ├── area-adjacency.csv
│   │   │       ├── biorange.csv
│   │   │       ├── bitstates.py
│   │   │       ├── classe-tensor-template.rev
│   │   │       └── model4.py
│   │   └── Stipa
│   │       ├── Stipa.smap.log
│   │       └── inputs
│   │           ├── 01_make-revscript-tensor.py
│   │           ├── Stipa.nex
│   │           ├── Stipa_classe_tensor.rev
│   │           ├── area-adjacency.csv
│   │           ├── biorange.csv
│   │           ├── bitstates.py
│   │           ├── classe-tensor-template.rev
│   │           └── model4.py
│   └── model4.py
├── p20
│   ├── events-bootstrap1000_p20.csv.gz
│   └── selected_clades
│       ├── Campanula
│       │   ├── 01_make-revscript-tensor.py
│       │   ├── Campanula.nex
│       │   ├── Campanula_classe_tensor.rev
│       │   ├── area-adjacency.csv
│       │   ├── biorange.csv
│       │   ├── bitstates.py
│       │   ├── classe-tensor-template.rev
│       │   └── model4.py
│       └── Saxifraga
│           ├── 01_make-revscript-tensor.py
│           ├── Saxifraga.nex
│           ├── Saxifraga_classe_tensor.rev
│           ├── area-adjacency.csv
│           ├── biorange.csv
│           ├── bitstates.py
│           ├── classe-tensor-template.rev
│           └── model4.py
└── p50
├── events-bootstrap1000_p50.csv.gz
└── selected_clades
├── Campanula
│   ├── 01_make-revscript-tensor.py
│   ├── Campanula.nex
│   ├── Campanula_classe_tensor.rev
│   ├── area-adjacency.csv
│   ├── biorange.csv
│   ├── bitstates.py
│   ├── classe-tensor-template.rev
│   └── model4.py
└── Saxifraga
├── 01_make-revscript-tensor.py
├── Saxifraga.nex
├── Saxifraga_classe_tensor.rev
├── area-adjacency.csv
├── biorange.csv
├── bitstates.py
├── classe-tensor-template.rev
└── model4.py

events-bootstrap1000.csv.gz: Compressed table contains all reconstructed biogeographic events from 1,000 stochastic mappings across 34 plant clades, including time, range states, event type, and region-level binary encodings needed to reproduce all event-based analyses in the manuscript.

These subdirectories (MaxRange3/, MaxRange4/, p20/, p50/) represent different settings for the maximum biogeographic occupancy, used to ensure the computational feasibility of the tensor-based biogeographic model. Each subdirectory contains the input clades and the associated scripts used for parsing, running, and summarizing the analyses.

Each subdirectory includes:

• clades/ folder containing the output (stochastic mapping *.smap.log file) and the required inputs for each clade:
  • the dated phylogenetic tree (*.nex)
  • the range–biome binary matrix (biorange.csv)
  • the .rev script automatically generated by 01_make-revscript-tensor.py
• Supporting files required to run the RevBayes analyses:
  • template script classe-tensor-template.rev
  • custom modules bitstates.py and model4.py
  • adjacency matrix of geographic regions (area-adjacency.csv)

Scripts for parsing and bootstrap replication

Each subdirectory also includes the scripts used to parse and process tensor outputs:

• 2_parse-tensor-json.py
  Parses the RevBayes tensor JSON outputs into standardized event tables.
• 3_bootstrap_parse.py
  Performs random sampling to obtain 1,000 replicated biogeographic histories, generating bootstrap summaries of events across clades under each range-size constraint.

Sensitivity analysis subdirectories: p20/ and p50/

The p20/ and p50/ folders follow the same structure described above but contain analyses in which Campanula and Saxifraga were subsampled to 50% or 20% of their total species diversity.
These reduced datasets were used for sensitivity analyses to evaluate the robustness of biogeographic inferences to sampling fraction.

Each of these folders includes the corresponding compressed tables of reconstructed biogeographic events from 1,000 stochastic mappings under the reduced sampling schemes.

File: Connectivity_analysis.zip

Description: This directory contains paleoclimate-based connectivity layers generated at 1-million-year intervals from 0 to 35 Ma, used to quantify long-term changes in the structural connectedness of alpine and cold-adapted habitats across the Northern Hemisphere.

Each time slice contains four types of files:

• *_temp.tif — reprojected paleotemperature raster (annual mean temperature)
• *_alpine_biome.tif — binary raster of alpine biome pixels
• *_cost_250-links.* — Graphab output files describing least-cost links between alpine-biome patches (shapefile + CSV adjacency tables)
• *_quickplot.png — quick-look diagnostic map for visual inspection

The final file, connectivity_over_time.rds, stores the probability connectivity (PC) calculated for all 35 million-year intervals.

File: Plots.zip

This directory contains scripts and intermediate processed files used to generate the figures included in the main text. These scripts read the biogeographic event tables, summarize temporal dynamics, and produce rate-through-time curves, cumulative dispersal plots, and event-type visualizations. 

All Python scripts were run using Python 3.9, with required packages listed at the top of each script. The R script was run using R 4.4, with required packages loaded within the script. These package imports provide the information needed to reproduce the analyses; no additional environment files are required.

• Fig1_plot_nestedPieChat.py
  Generates the nested pie-chart visualizations summarizing the relative contributions of different biogeographic event types across mountain systems (Figure 1).
• Fig2B_rate_emigration.py
  Computes emigration (out-of-region) rates through time for all clades and produces the rate-through-time curves used in Figure 2B.
• Fig2B_rate_immigration.py
  Computes immigration (into-region) rates through time and generates the complementary rate-through-time curves for Figure 2B.
• Fig2C_dispersal_cumsum_A.csv
  Pre-processed summary table of dispersal events used for cumulative dispersal calculations in Figure 2C. Range and biome abbreviations shown in column matric, from and to follow the definitions provided in the main text and Supplementary Materials.
• Fig2C_dispersal_cumsum.py
  Calculates cumulative dispersal events over time for each region and generates the csv file Fig2C_dispersal_cumsum_A.csv.
• Fig2C_sumBarchart.R
  R script for summarizing and visualizing net immigration or emigration as bar charts, providing the categorical summaries for Figure 2C.
• Fig3_insitu_rate_all.py
  Estimates in situ speciation rates across mountain regions, generating the rate-through-time curves displayed in Figure 3.

File: GenBank_accession.zip

Description: This directory contains all curated GenBank accession numbers for the molecular sequences used in the phylogenetic reconstruction for this study. For each clade, the most commonly sequenced gene regions (e.g., ITS, matK, rbcL, petD, trnL–trnF) were selected to maximize taxonomic coverage.

Each .csv file corresponds to one major clade or lineage included in the analyses. Columns represent the species names and the accession numbers for each gene region, with missing sequences indicated by “–”.