Reproducible Occurrence-to-Biome Classification Using 31 Global Biome Schemes

Reproducibly classifies occurrence records into biomes using 31 published global biome schemes compiled by Fischer and colleagues (2022) , provided as harmonised raster layers at 10x10 km resolution globally. Includes functions to choose the most suitable biome scheme for a dataset by a data-driven ranking, to classify occurrence records, and to tabulate and visualise the result. Works with user-provided occurrences or a taxon name, in which case occurrences are downloaded from GBIF (< https://www.gbif.org>) and cleaned automatically.

biomes: An R package for reproducible occurrence-to-biome classification using 31 global biome schemes

biomes ships spatially explicit raster layers of 31 published global biome schemes (compiled by Fischer et al. 2022) in one harmonised format, at a native 10 × 10 km resolution, together with functions to classify occurrence records into biomes and to choose the most suitable scheme for a dataset in a transparent, reproducible way.

Because the layers are global at 10 × 10 km, biomes is built for macroecological and large-scale biogeographical questions (continental to global). It is not intended for fine-grained, national-extent analyses, where the 10 × 10 km grid is coarse.

The workflow follows four steps (see the four vignettes below):

Assembly of occurrence records and biome schemes.
Choice of a biome scheme: rank the 31 schemes for your data.
Occurrences-to-biome classification: assign records to biomes.
Output and visualisation: tabulate and visualise the result.

A single wrapper, biomes_full(), runs all four steps in one call.

Terms used in this package

biome definition: the concept on which a scheme delimits its biomes, e.g. by climate, by vegetation, by land cover, by ecoregions, by a synthesis of several criteria (integrative), or by human land use (anthropogenic). Schemes sharing a biome definition form a group; the definition argument ranks within one group.
biome scheme: one of the 31 published classification systems (e.g. the vegetation scheme of Ramankutty & Foley 1999).
biome: a single category within a scheme (e.g. savanna, tropical evergreen woodland).
biome scheme number: the number (1-31) that identifies a scheme; it is the value you pass to the scheme argument. Numbers follow the order of the biome inventory (Fischer et al. 2022).

Installation

The released version is available from CRAN:

install.packages("biomes")

The development version with the latest changes can be installed from GitHub:

# install.packages("devtools")
devtools::install_github("azizka/biomes")

biomes depends on terra and installs ggplot2, sf, viridis and tidyterra automatically. The combined figure (cowplot) and the GBIF download path (rgbif, CoordinateCleaner) live in Suggests and can be installed on demand.

Quick start: the one-call workflow

biomes_full() runs the whole four-step workflow. Give it an occurrence dataset (or a taxon name); it chooses the best-fitting biome scheme, classifies the records, tabulates them, and (optionally) builds a figure. No figure is drawn by default (plot = "none", the fastest option). plot = "all" returns the combined figure in res$plot, while a subset of c("rank", "map", "barplot") returns the panels individually (no panel letters) in res$rank, res$map and res$barplot.

library(biomes)

data(bombacoideae_occurrences)

# from an occurrence dataset:
res <- biomes_full(x = bombacoideae_occurrences)   # scheme = "best"; no figure by default

res            # short summary
res$scheme     # the chosen biome scheme number
res$table      # occurrence records per biome

# add a figure with `plot`; the combined figure is returned in res$plot:
res <- biomes_full(x = bombacoideae_occurrences, plot = "all")
res$plot
ggplot2::ggsave("biomes_figure.png", res$plot, width = 9, height = 12, dpi = 300)

# the panels individually (no a/b/c letters) in res$rank / res$map / res$barplot:
res <- biomes_full(x = bombacoideae_occurrences, plot = c("rank", "map", "barplot"))
res$rank; res$map; res$barplot

# from a taxon name instead of a dataset (downloads from GBIF), e.g. a whole order:
res <- biomes_full(taxon = "Fagales", plot = "all")

To force a specific scheme, pass its biome scheme number; to rank within one methodological group, pass a biome definition:

biomes_full(x = bombacoideae_occurrences, scheme = 1)            # fixed scheme
biomes_full(x = bombacoideae_occurrences, scheme = "vegetation") # best vegetation scheme

The same pipeline as individual building blocks, following the four workflow steps:

# Step 1: Assembly of occurrence records and biome schemes
data(bombacoideae_occurrences)
schemes <- biomes_get()

# Step 2: Choice of a biome scheme
ranking <- biomes_rank(bombacoideae_occurrences, definition = "vegetation")
best <- attr(ranking, "best_scheme")

# Step 3: Occurrences-to-biome classification
cls <- biomes_classify(bombacoideae_occurrences, scheme = best)

# Step 4: Output & visualisation
biomes_tab(cls)
biomes_visualise(bombacoideae_occurrences, scheme = best)

biomes_visualise() draws up to three panels (rank, map and barplot) combined into one figure; select any subset with panels, e.g. panels = "map".

Vignettes: one per workflow step

Read them on the package website, or open them locally with browseVignettes("biomes").

Citation

Groß HC, Fischer J-C, Walentowitz A & Zizka A (2026) biomes: An R package for reproducible occurrence-to-biome classification using 31 global biome schemes. bioRxiv, doi:10.64898/2026.09.25.754345, https://github.com/azizka/biomes. R package version 0.9.5.
Fischer J-C, Walentowitz A, Beierkuhnlein C (2022): The biome inventory – Standardizing global biogeographical land units. Global Ecology and Biogeography 31(11): 2172-2183. https://doi.org/10.1111/geb.13574