Global Biodiversity

Mapping the geographic distribution of species using the GBIF API

C. Ribeiro
Data Analyst

Global Biodiversity Information Facility (GBIF)

GBIF's vision

"A world in which biodiversity information is freely and universally available for science, society and a sustainable future."

GBIF is an international open data infrastructure funded by governments.
It provides a single point of access to more than 400 million records related to evidence of more than one million species.
It is the biggest biodiversity database on the Internet.
rgbif is an R package to search and retrieve biodiversity data using the public GBIF API.

Example: Ursus americanus

The American black bear (Ursus americanus) is a medium-sized bear native to North America. It is the continent's smallest and most widely distributed bear species.

name_backbone() is used to search against the GBIF backbone taxonomy:

suppressPackageStartupMessages(library(rgbif))
sp <- name_backbone(name='Ursus americanus', rank='species')
cat(paste(sp[c("kingdom","phylum","class","order","family","genus","species")]),sep=",")

## Animalia,Chordata,Mammalia,Carnivora,Ursidae,Ursus,Ursus americanus

Count the number of georeferenced occurrences with occ_count():

(max_occ <- occ_count(taxonKey=sp$speciesKey, georeferenced=TRUE))

## [1] 2859

Example: Ursus americanus

Search for georeferenced occurrences using occ_search():

dat <- occ_search(taxonKey=sp$speciesKey,
        fields=c("name","key","decimalLatitude","decimalLongitude","basisOfRecord"),
        limit=max_occ, return='data',hasCoordinate= TRUE)

head(dat[,3:5])

##       basisOfRecord decimalLongitude decimalLatitude
## 1 HUMAN_OBSERVATION          -103.29           29.23
## 2 HUMAN_OBSERVATION           -72.53           43.74
## 3 HUMAN_OBSERVATION          -103.29           29.28
## 4 HUMAN_OBSERVATION          -103.32           29.27
## 5 HUMAN_OBSERVATION          -103.30           29.28
## 6 HUMAN_OBSERVATION           -80.07           37.80

Example: Distribution of Ursus americanus

library(googleVis)
dat$Loc <- paste(dat$decimalLatitude, dat$decimalLongitude, sep=":")
plot(gvisMap(dat,locationvar="Loc",tipvar="basisOfRecord"))