This repository stores the climatic niche estimation of 1422 plant and lichen species recorded in the International Tundra EXperiment database (ITEX).
The niches were estimated by first pulling the first 10,000 occurrences from the GBIF. After a filtering of the occurrences based on coordinates , we extracted the bioclimatologies from the latest version of the CHELSA V2.1 global climatology database. For each species, the mean, median, the quantile 5% and 95%, the interquantile range and the mode were computed from the distribution of each bioclimatic CHELSA variable where the species occurred. Species with less than 100 filtered occurrences were discarded from the niche computation. Estimations from the boreal and tundra ranges of 1256 species are also available.
To analyse species climatic niche in a multivariate framework, a PCA was also conducted to reduce the dimension of every CHELSA cells where at least one species occurred. The results of the PCA as well as the centroïds of each species projected in this PCA space are also available.
In ClimNiche_database, ClimNiche_climate_summary.csv presents the summary statistics for each climate variable of each species, ClimNiche_pca_summary.csv contains the centroïds of each species occurrence projected in the PCA and sampling_summary.csv doubles as a list of taxonomic synonym with the ITEX database and a count of how many occurrences were used in the niche computation.
The Complete_sampling folder contains all of the CHELSA cells (all_CHELSA_cells.Rdata) where at least one species occurrences, along with a unique cell_id identifier. species_cells.RData contains every cell_id where every species occurred, allowing for creating the full sampling back with a merge function.
In ClimNiche_database and Complete_sampling, "boreal_tundra" files are also available, corresponding to the same procedure but constrained to the tundra-boreal bioclimatic region to accommodate for ubiquitous species having cold-adapted sub-populations. In these files, 985 species are available.
The folder raw_occ contains the the 10,000 (or less) raw occurrence downloaded from GBIF before 15-feb-2015, in .RData format. The folder filtered_occ folder contains processed raw occurrences filtered by the occTest packages, removing duplicates within a CHELSA pixel, as well as test for various coordinates errors. chelsa_data_V2.1 contain the metadata behind the CHELSA bioclimatic name, and should contain the CHELSA rasters. They are not included but can be download with the script download_CHELSA_V2.1.R .
00_getting_occ_data.R downloads the raw occurrences from GBIF, 01_filtering_occ.R apply the geographical filters included in the occTest package. 02_getting_clim_niche.R extracts the climatic variable for each occurrences, performs the various summary metrics computation and the PCA-based analysis. None of these has to be ran again to gain access the final estimation of species climatic niches.
These script can be used to extend the analysis to non-vascular taxa.
We are grateful to the creators and the maintainer of the packages spocc, occTest, terra, sf and data.table that greatly facilitated our analysis, as well to the many data contributors to the GBIF and its sub-repositories (such as INaturalist). JB and IMS were funded by the ERC synergy grand n°101071417.
Jeremy Borderieux, Isla H. Myers-Smith
This dataset can be cited alongside the upcoming publication that motivated its creation. If the publication is not out yet, please get in touch with the authors.
Data are publicly available using a Creative Commons Attribution 4.0 International copyright (CC BY 4.0). Data are fully public but should be appropriately referenced by citing the paper. Although not mandatory, we additionally suggest that data users contact and collaborate with data contributors if this dataset will contribute a substantial proportion of observations used in a particular paper or analysis.