Terrestrial example: Panthera tigris
Inspect the taxon concept
Before downloading, get_status() shows which accepted
name and synonyms get_gbif() will use internally, and
retrieves the current IUCN Red List status from the GBIF backbone.
# Accepted name and direct synonyms only (the default)
get_status("Panthera tigris")
# Also include infra-specific taxa (subspecies, varieties) —
# these are the keys actually used by get_gbif()
get_status("Panthera tigris", level = "children")level = "all" additionally returns alternative name
representations for manual inspection, but those extra entries are not
used for occurrence retrieval.
Download occurrences
get_gbif() works without GBIF credentials. It is built
on top of rgbif (Chamberlain et al. 2022) and harmonizes
the query to the accepted GBIF taxon key, applies a dynamic
moving-window tiling strategy when the geographic extent contains more
than 10,000 records, and runs 13 configurable post-processing filters
based on custom logic and CoordinateCleaner (Zizka et
al. 2019). The function originates from the occurrence retrieval
workflow first employed in Chauvier et al. (2021, Ecological
Monographs).
countries <- terra::vect(
ext_file("world_countries.shp")
)
terra::plot(countries, col = "#bcbddc")
graphics::points(obs_pt[, c("decimalLongitude", "decimalLatitude")],
pch = 20, col = "#99340470", cex = 1.5)
Note that some records of likely captive individuals remain (e.g., in
Europe, the U.S., and South Africa) — the default
CoordinateCleaner-based filters do not remove all zoo or
botanical garden records. The get_gbif() help page
documents the available post-processing arguments for stricter
cleaning.
Build the range map
get_range() implements the ecoregion-constrained range
inference algorithm originally developed by Hagen et al. (2019) and can
accept any types of observations as long as
’decimalLongitude,decimalLatitude` columns are valid:
# Download and read the packaged terrestrial ecoregions (The Nature Conservancy 2009)
eco_terra <- read_ecoreg(ecoreg_name = "eco_terra", save_dir = tempdir())
# Range (default)
range_tiger <- get_range(
occ_coord = obs_pt,
ecoreg = eco_terra,
ecoreg_name = "ECO_NAME",
degrees_outlier = 5,
clust_pts_outlier = 4,
format = "SpatVector"
)
# Plot
terra::plot(countries, col = "#bcbddc")
terra::plot(range_tiger$rangeOutput, col = "#238b45",
add = TRUE, axes = FALSE, legend = FALSE)
degrees_outlier and clust_pts_outlier
control how isolated clusters of observations are handled before the
ecoregion lookup. Increasing either parameter produces a more
conservative range that excludes more distant clusters; the defaults
(~550 km and ~440 km respectively) already removed the most obvious
anomalies in Europe, the U.S., and South Africa for this example. Note
that default parameters are usually recommended for creating range at
the global scale — see the “Tuning the main range arguments”
section of
vignette("ecoregion-constrained-range-inference").
Optionally, format controls the geometry type of the
output range. Three are available: SpatVector,
sf, and SpatRaster. If polygons, then
merge_range can be optionally used to aggregate the
geometry:
By default, the range output has its number of features defined by
ecoreg_name for exploratory purposes. If not needed,
merge_range can be applied to dissolve them into a single
polygon.
