Dataset DOI: 10.5061/dryad.f7m0cfz8n

Description of the data and file structure

Dataset used to compare two metrics of species’ range size and two metrics of species’ range structure that are relevant at coarse, continental spatial resolutions. We calculated these metrics for 813 species endemic to Europe, and we evaluated whether drivers of species’ range sizes correspond to drivers of range structure.

Files and variables

File: Input_data.csv

Description: Geographic range size and range structure metrics for 813 species’ calculated from complete occurrence data at 50✕50 km² resolution sourced from the database of Atlas Florae Europaeae.

Variables

• Species: Species names; The taxonomy and nomenclature follow the database of the AFE.
• Model: Occurrence data
• range.size: Geographic range size expressed as Extent of occurrence. Area of all occupied non-occupied habitat patches within a species’ geographic range boundaries.
• total.area: Geographic range size expressed as Area of occupancy. The sum of the area of all occupied habitat patches within a species’ range (as Extent of occurrence, but it includes only occupied grid cells within the species’ range boundaries)
• prop.landscape: Geographic range structure expressed as Geographic range filling. The proportion of a species’ geographic range represented by occupied habitat patches. It equals the area of occupancy divided by the extent of occurrence, multiplied by 100 (to convert to a percentage).
• effective.mesh.size: Geographic range structure expressed as patch size distribution. Measures the subdivision of a species’ geographic range into occupied habitat patches of different sizes relative to the extent of species’ geographic range. Calculated as the occupied patch area squared, summed across all occupied patches, divided by the extent of occurrence.
• nwidth_mcp: Climatic niche width based on the minimum convex polygon. Based on eigth climate variables: tdr.max (mean diurnal temperature range, Bio02, maximum), tar.min (temperature annual range, Bio07, minimum), twetq.min (mean temperature of wettest quarter, Bio08, minimum), tdryq.max (mean temperature of driest quarter, Bio09, maximum), twarmq.max (mean temperature of warmest quarter, Bio10, maximum), p.mean (annual precipitation, Bio 12, mean), ps.max (precipitation seasonality, Bio15, maximum), and pdryq.min (precipitation of driest quarter, Bio17, minimum). Climate variables were downloaded from the CliMond Archive (V1.2; https://www.climond.org/BioclimRegistry.aspx, Kriticos et al. 2012) (Kriticos et al. 2014). To calculate climatic niche breadth, we first reduced the dimensionality of the climate space to two axes using a Principal Component Analysis (PCA) calibrated on all grid cells in Europe. Then, for each species, we summed the standard deviation of values corresponding to each occupied grid cell, which we weighted by the corresponding axis Eigenvalue.
• nwidth_kd: Climatic niche width. 
• sd.elev: Range-wide topographic heterogeneity. Expressed as the standard deviation of the elevation above sea level across occupied grid cells extracted from elevation maps (Jarvis et al. 2008). The elevation layer was downloaded at 30’’ resolution (~1✕1 km²). We calculated mean, minimum and maximum across the 1✕1 km² pixels contained in each 50✕50 km² grid cell.
• sp.dist.equal.mean: Geographic range proximity to sea coastline, calculated as the average distance in decimal degrees of each occupied grid cell to the closest shore in Europe.
• median.long: Species’ geographic range position, calculated as the median longitude of all 50 ✕ 50 km² occupied grids cells within a species’ range.
• median.lat: Species’ geographic range position calculated as the median latitude of all 50 ✕ 50 km² occupied grids cells within a species’ range.

File: tree1

Description: The phylogenetic tree available in Zanne et al. (2014) for 813 species in the dataset.

File: Script_GRS_Size_vs_structure.Rmd

Description: The provided script contains code for exploratory data analyses, Pearson's r calculations, Phylogenetic Generalized Least Squares model (pGLS; Freckleton et al. 2002) and plotting.

Code/software

The pGLS models were fitted with the ‘caper’ package (Orme et al. 2013). All analyses were performed in R 3.2.3 (R Core Development Team 2015) and later versions.

Access information

Other publicly accessible locations of the data:

• 1.       The database of Atlas Florae Europaeae. Available at: https://www.luomus.fi/en/database-atlas-florae-europaeae. Last accessed 30 June 2015.

Data was derived from the following sources:

• N/A