MEDIS: spatial data for Mediterranean islands
Data files
May 07, 2024 version files 90.08 MB
-
medis.geojson
-
README.md
Abstract
The intrinsic characteristics of islands make them unique for studying ecological and evolutionary dynamics. The Mediterranean Basin, a biodiversity hotspot, is rich in islands, hosting a significant global biodiversity proportion. Despite extensive research, a comprehensive spatial dataset for these islands is lacking. This study presents the first comprehensive spatial dataset of all Mediterranean islands larger than 0.01 km2, aiding ecological investigations and interdisciplinary research on economic, environmental, and social issues. The MEDIS spatial dataset offers detailed information on 36 geographic, climatic, ecological, and land-use variables, including island area, perimeter, isolation metrics, climatic space, terrain data, land cover, paleogeography, road networks, and geological information, providing a multifaceted view of each island's characteristics. The study encompasses 2212 islands in the Mediterranean Basin larger than 0.01 km2. The spatial grain varies, with datasets like CHELSA-BIOCLIM+ and EU-DEM providing high-resolution climatic and terrain data. The spatial dataset incorporates various datasets, each with its own timeframes, such as the Global Shoreline Vector from 2014 Landsat imagery and the WorldCover dataset from 2021. Historical data like the Paleocoastlines GIS dataset offer insights into island configurations during the Last Glacial Maximum. While not focusing on specific taxa, the study lays the foundation for comprehensive research on Mediterranean islands, facilitating comparisons and investigations into the distribution of native, endemic, or alien species. The level of measurement is extensive, encompassing a wide range of variables and providing polygonal features rather than centroids’ coordinates.
README: MEDIS: spatial data for Mediterranean islands
https://doi.org/10.5061/dryad.b8gtht7k5
This repository contains the MEDIS dataset as well as the R scripts and files necessary to create the MEDIS database.
The "data" directory includes the spatial layer featuring island names and IDs (island_polygons.gpkg), alongside several subdirectories housing auxiliary files used for variable extraction. Climatic variables must be downloaded only once using the cumpute_climate.R script.
Description of the data and file structure
The complete dataset is provided in a .GeoJSON format, with reference system ETRS89-extended / LAEA Europe (EPSG: 3035). A complete list of all the variables can be consulted in the table below.
Name | Type of variable | Unit of measure | Source | Description |
---|---|---|---|---|
id | Qualitative | NA | NA | Unique identifier. Defined by ISO 3166-1 alpha-3 country codes and island number. Islands are ordered for each country according to decreasing area size |
name | Qualitative | NA | expert-based | Name of the island |
name_simple | Qualitative | NA | expert-based | Name of the island without accent marks |
country | Qualitative | NA | various sources | Island country |
area_km2 | Quantitative | km² | GSV (Sayre et al., 2019) | Area value of the island (planar EPSG: 3035) |
perimeter_km | Quantitative | km | GSV (Sayre et al., 2019) | Perimeter value of the island (planar EPSG: 3035) |
lgm_id | Qualitative | NA | Paleocoastlines GIS dataset (Zickel et al., 2016) | Polygon ID of the Paleocoastlines dataset to which the island belonged |
area_lgm_km2 | Quantitative | km² | Paleocoastlines GIS dataset (Zickel et al., 2016) | Area of the Polygon ID of the Paleocoastlines dataset to which the island belonged (planar EPSG: 3035) |
volcanic | Qualitative | NA | expert-based | The field indicates if the island is entirely volcanic |
dist_mainland | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the mainland (planar EPSG: 3035) |
dist_closest_island_km | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the closest island (planar EPSG: 3035) |
dist_closest_larger_island_km | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the closest larger island (planar EPSG: 3035) |
dist_closest_10x_larger_island_km | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the closest 10 times larger island (planar EPSG: 3035) |
dist_closest_100x_larger_island_km | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the closest 100 times larger island (planar EPSG: 3035) |
dist_closest_1000x_larger_island_km | Quantitative | m | GSV (Sayre et al., 2019) | Distance to the closest 1000 times larger island (planar EPSG: 3035) |
bio_01_mean | Quantitative | °C | CHELSA-BIOCLIM+ (Brun et al. 2022) | Mean value of mean annual air temperature |
bio_01_stdev | Quantitative | °C | CHELSA-BIOCLIM+ (Brun et al. 2022) | Standard deviation value of mean annual air temperature |
avg_bio_04_mean | Quantitative | °C/100 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Mean value of temperature seasonality |
sd_bio_04_stdev | Quantitative | °C/100 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Standard deviation value of temperature seasonality |
avg_bio_07_mean | Quantitative | °C | CHELSA-BIOCLIM+ (Brun et al. 2022) | Mean value of annual range of air temperature |
sd_bio_07_stdev | Quantitative | °C | CHELSA-BIOCLIM+ (Brun et al. 2022) | Standard deviation value of annual range of air temperature |
avg_bio_12_mean | Quantitative | kg m-2 year-1 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Mean value of annual precipitation amount |
sd_bio_12_stdev | Quantitative | kg m-2 year-1 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Standard deviation value of annual precipitation amount |
avg_bio_15_mean | Quantitative | kg m-2 year-1 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Mean value of precipitation seasonality |
sd_bio_15_stdev | Quantitative | kg m-2 year-1 | CHELSA-BIOCLIM+ (Brun et al. 2022) | Standard deviation value of precipitation seasonality |
max_elev_m | Quantitative | m | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Maximum elevation value (WGS84 ellipsoidal vertical datum) |
mean_elev_m | Quantitative | m | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Mean elevation value (WGS84 ellipsoidal vertical datum) |
sd_elev_m | Quantitative | m | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Standard deviation of elevation values (WGS84 ellipsoidal vertical datum) |
mean_rug | Quantitative | adimensional | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Mean of Terrain Ruggedness Index values (TRI; Riley et al., 1999) (WGS84 ellipsoidal vertical datum) |
sd_rug | Quantitative | adimensional | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Standard deviation of Terrain Ruggedness Index values (TRI; Riley et al., 1999) (WGS84 ellipsoidal vertical datum) |
mean_tpi | Quantitative | adimensional | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Mean of Topographic Position Index values (TPI; Weiss, 2001) (WGS84 ellipsoidal vertical datum) |
sd_tpi | Quantitative | adimensional | NASA SRTM Digital Elevation 30m (Farr et al., 2007) | Standard deviation of Topographic Position Index values (TPI; Weiss, 2001) (WGS84 ellipsoidal vertical datum) |
n_lc | Quantitative | count | ESA WorldCover 10m v200 (Zanaga et al., 2022) | Number of unique land cover types in the island |
even_lc | Quantitative | adimensional | ESA WorldCover 10m v200 (Zanaga et al., 2022) | Evenness of land cover types in the island |
perc_cropland | Quantitative | % | ESA WorldCover 10m v200 (Zanaga et al., 2022) | Cropland cover fraction in the island |
perc_built | Quantitative | % | ESA WorldCover 10m v200 (Zanaga et al., 2022) | Built up cover fraction in the island |
major_roads_length_km | Quantitative | km | OSM (OpenStreetMap contributors, 2024) | Length of major roads and road links in the island (planar EPSG: 3035) |
major_roads_density_km_1 | Quantitative | km^-1 | OSM (OpenStreetMap contributors, 2024) | Density of major roads (major roads length/island area) (planar EPSG: 3035) |
all_roads_length_km | Quantitative | km | OSM (OpenStreetMap contributors, 2024) | Total road length in the island (excluding hiking trails and minor paths) (planar EPSG: 3035) |
all_roads_density_km_1 | Quantitative | km^-1 | OSM (OpenStreetMap contributors, 2024) | Totoal road density (all roads length/island area) (planar EPSG: 3035) |
Sharing/Access information
Data was derived from the following sources:
- The Global Shoreline Vector was downloaded from Sayre, R., Noble, S., Hamann, S., Smith, R., Wright, D., Breyer, S., Butler, K., Van Graafeiland, K., Frye, C., Karagulle, D., 2019. A new 30 meter resolution global shoreline vector and associated global islands database for the development of standardized ecological coastal units. J. Oper. Oceanogr. 12, S47–S56.
- Climate data was derived from CHELSA-BIOCLIM+ https://chelsa-climate.org/bioclim/
- European Digital Elevation Model (EU-DEM) version 1.1 Copernicus Land Monitoring Service. 2021. Available online: https://land.copernicus.eu/. See also: https://land.copernicus.eu/imagery-in-situ/eu-dem/eu-dem-v1.1/view (accessed on 1 July 2023).
- Land cover information was extracted from the European Space Agency (ESA) WorldCover dataset; Zanaga, D., Van De Kerchove, R., De Keersmaecker, W., Souverijns, N., Brockmann, C., Quast, R., et al. (2021). ESA WorldCover 10 M 2020 V100. Zenodo. doi:10.5281/zenodo.5571936
- Road networks were derived from the Global Roads Inventory Project (GRIP) Meijer, J.R., Huijbregts, M.A., Schotten, K.C., Schipper, A.M., 2018. Global patterns of current and future road infrastructure. Environ. Res. Lett. 13, 064006.
Code/Software
Within the "R" directory, you'll find scripts employed to extract various sets of variables (geographic, landscape, distance, roads, and climate) for each island, as well as the script used for assembling the MEDIS database and creating the article's figures.
The "output" directory stores the sets of variables extracted using the aforementioned R scripts.
To execute the code, launch the R project file (medis.Rproj) in R Studio. Run each script within the R folder to generate the variable outputs. Note that files in the output directory will be overwritten. The order in which you execute the scripts is inconsequential. Some extractions and calculations may be time-consuming and are executed in parallel (execution time varies based on your machine). Once all output sets have been created, merge them with the spatial layer using the make_medis.R script to generate the MEDIS database in the main folder (medis.gpkg and medis.geojson).