Geographic range shapes influence species richness in global hotspots
Data files
Jul 14, 2025 version files 14.22 MB
-
AvianRangeShapes.jl.zip
14.22 MB
-
README.md
3.63 KB
Abstract
The extraordinary richness of species in tropical mountain regions is often attributed to aggregations of small-ranged species, allowing tight spatial packing of their ranges. However, the ranges of species in these regions are also distinctly more patchy and elongated than those found in adjacent lowlands. Our global analysis of mainland birds demonstrates that these range shapes augment spatial variation in species richness. Both the linearity and patchiness of species ranges are most pronounced in aseasonal and topographically complex regions, most notably in the tropical Andes, where these range shapes contribute more to the pattern of species richness than broad-scale variations in range size. Consequently, niche-based processes that govern the spatial configuration of habitats contribute to patterns of species richness in global hotspots over and above the processes that affect species richness through range sizes.
Dryad DOI: https://doi.org/10.5061/dryad.zgmsbccrd
The file AvianRangeShapes.jl.zip contains supplementary data and code used for the biogeographical simulation models that simulate species ranges with different assumptions about range linearity and patchiness.
The files are copied from the GitHub repository, July 3, 2025: https://github.com/JesSonne/AvianRangeShapes.jl
Package files
Project.toml: Defines the package’s metadata (name, version, dependencies). Required for managing the package with Julia's package manager.
Manifest.toml: Captures the exact versions of all dependencies used.
LICENSE: Specifies the license under which the package is distributed.
script/null models.jl file
Contains the Julia script for running each of the biogeographical simulations.
data folder
Data files used in the biogeographical simulations.
clim1.tif, clim4.tif, clim12.tif, clim15.tif: Rasters of bioclimatic variables used for determining the climate volume occupied by a focal species
sf1_mainland.tif: Raster file of global mainland regions used as dispersal domain in the study
top_q_proj.tif: Topographical raster used for constraining the geographical domain according to the elevational range limits of a focal species
bird ranges.jld2: Julia file containing the geographical distribution of example species
elevational range limits.jld2: Julia file containing the elevational range limits of the example species
standardized_range_sizes.jld2: Julia file containing the standardized range sizes of the example species
src folder
AvianRangeShapes.jl: Contains the source code, but see functions.jl for the main functionality
functions.jl: contains the code (written in Julia) used for running the simulations.
misc/output folder
Contains raster maps of empirical (empirical_richness) and simulated species richness (null_rich), respectively, at 0.25 (025deg) and 1.0 degree (1deg) spatial resolution.
The null models have different assumptions about the linearity and patchiness of a species' distribution.
Null Model 1 (nm1) has the fewest constraints, as the species can disperse anywhere on a geographic domain from a specified starting position. This null model tends to generate compact range shapes that are restricted only by the boundaries of the continental coastlines.
In contrast, Null Model 2 (nm2) allows species to disperse only within their minimum and maximum elevational range limits. This constraint generates range shapes that are bound to the spatial configuration of the topographic landscape and the contemporary climate conditions in which the species is found. The resulting ranges become geographically structured (usually linear, in mountain cordilleras), following the elevational zones of habitats.
Null Models 3 (nm3) and 4 (nm4) are similar to 1 and 2 but relax the assumption of range coherency, allowing multiple starting positions, one for each individual empirical range patch.
We reran each null model with range sizes within the mainland realms of the world standardised to match the global range size frequency distribution (RsStd).
Example:
"null_rich_nm4_RsStd_1deg.tif" contains a raster map of the simulated species richness by null model 4 using standardized range size frequency distributions at a 1-degree spatial resolution.
"null_rich_nm2_025deg.tif" contains a raster map of the simulated species richness by null model 2 using empirical range size frequency distributions at a 0.25 degrees spatial resolution.