Data from: Biotic velocity identifies potential climate refugia beyond climate velocity estimates
Data files
Dec 01, 2025 version files 8.15 GB
-
Data.zip
1.46 GB
-
README.md
4.84 KB
-
SDM_outputs.zip
6.69 GB
Abstract
We developed a biotic velocity–based framework to identify three functionally distinct types of climate refugia with significant conservation value at the regional scale: (1) in-situ refugia, where species assemblage can persist locally without spatial displacement (i.e., areas with mean forward or backward biotic velocity equal to 0), (2) stepping-stone refugia, where species can track shifting climates with relatively low forward biotic velocity, and (3) ex-situ refugia, which are areas that can be readily colonized by species from other regions due to low backward biotic velocity. This framework, which integrates species distribution models (SDMs), climate exposure metrics, and biotic velocity, was applied to four taxa (mammals, reptiles, birds, and amphibians) under two climate scenarios in the Middle and Lower Reaches of the Yangtze River (MLRYR), China.
Dataset DOI: 10.5061/dryad.v15dv427w
This dataset supports the research titled "Biotic velocity identifies potential climate refugia beyond climate velocity estimates". It contains the data and scripts required to reproduce key analytical steps, including species distribution models (SDMs), climate exposure assessments, and biotic velocity calculations.
Description of the data and file structure
1.Data.zip
- Species List.xlsx
Provides detailed information for the 95 species included in this study, encompassing: scientific name, taxonomy, IUCN Red List threatened categories, number of points generated from the Area of Habitat (AOH - referring to species distribution ranges refined by incorporating land cover and elevation information), number of pre-processed occurrence records obtained from GBIF and iDigBio, number of occurrence points after filtering using environmental subsampling methods, habitat type, and elevation.
- Analysis scripts|
A directory containing the main R scripts:- Biotic velocity calculation.R: Calculates the distance from each cell within a species’ projected current (or future) range to the nearest cell of future (or current) suitable habitat.
- Climate exposure assessment.R: Identify regions where the degree of projected climate change is relatively low.
- Species distribution modelling.R: Projects current and future suitable habitats for species by integrating environmental variables and species occurrence data (using the megaSDM R package).
- Principal component analysis.R: Performs dimensionality reduction on 19 bioclimatic variables to extract independent principal component (PC) axes.
- Parameter tuning results|
SDM parameter tuning results (.csv) from the ENMeval R package for eight representative species. For each species, 30 combinations of five feature classes (FCs: "L", "LQ", "LQH", "LQHP", "LQHPT") and six regularization multipliers (RMs: 0.5, 1, 2, 3, 4, 5) were tested. The results include information such as the corrected Akaike Information Criterion (AICc) and average validation AUC. - Sample|
Contains complete modelling data for the example species Panthera tigris, provided to facilitate replication of the SDM.-
aoh_points|
| - Panthera_tigris.csv
Coordinates of Panthera tigris. -
trainingarea|
10 Environmental variables (.bil) under current (1985s) climate conditions, including BIO3 (Isothermality, %), BIO8 (Mean Temperature of Wettest Quarter, °C), BIO9 (Mean Temperature of Driest Quarter, °C), BIO14 (Precipitation of Driest Month, mm), BIO15 (Precipitation Seasonality, Coefficient of Variation, %), BIO18 (Precipitation of Warmest Quarter, mm), land use (categorical, unitless), aspect (°), slope, (°), and the distance to the nearest water (m). -
predictenv|
| - SSP5-8.5|
| |- 2090|
Projections for the same 10 environmental variables (.bil) under future (2090s, SSP5-8.5) climate conditions.*Note: MegaSDM requires environmental variables in .bil format.
-
2.SDM_outputs.zip
- The SDM outputs of 89 individual species are categorized into four subfolders—Amphibians, Birds, Mammals, and Reptiles—based on their taxonomic groups. Within the subfolder, distribution maps for each species are stored in separate folders named by their Latin names. Taking the species Andrias davidianus as an example, the structure of the folder named 'Andrias davidianus' is as follows:
-
Andrias_davidianus|
[Current range maps] Including GRD files of continuous (0-1) and binary (0/1) habitat suitability maps for the current period (1985s).| - map_pdfs|
Including preview images of continuous (0-1) and binary (0/1) habitat suitability maps for the 1985s (1970-2000) and 2090s (2081-2100, under SSP5-8.5).| - SSP585|
Including GRD files of continuous (0-1) and binary (0/1) habitat suitability maps for the future period (2090s).
-
Sharing/Access information
Data Sources
- Initial species distribution maps were derived from:
- IUCN Red List https://www.iucnredlist.org/en.
- BirdLife International https://www.birdlife.org/.
- 19 Bioclimatic variables were derived from:
- WorldClim version 2.1 https://worldclim.org/.
Code/Software
All scripts were run using R version 4.3.1 and contain comprehensive annotations throughout. All code derived from external sources has been explicitly credited within the corresponding script comments.