Match scientific names with the GBIF Backbone Taxonomy

Introduction

This tutorial will explain how you can match a list of scientific names against the GBIF backbone taxonomy.

It is important that you have the most recent version of inborutils installed and available:

remotes::install_github("inbo/inborutils")   # install inborutils

library(tidyverse)    # To do datascience
library(rgbif)        # To lookup names in the GBIF backbone taxonomy
library(inborutils)   # To wrap GBIF API data
library(knitr)

Read data file containing the scientific names

Read file containing the scientific names you want to check against the GBIF taxonomic backbone:

species_df <- read_csv("https://raw.githubusercontent.com/inbo/inbo-pyutils/master/gbif/gbif_name_match/sample.csv", trim_ws = TRUE, col_types = cols())

Take a look at the data:

kable(species_df)

Request taxonomic information

Given a data.frame, you can match the column containing the scientific name against GBIF Backbone Taxonomy, using the gbif_species_name_match function from the inborutils package. You need to pass a data.frame, df and a column name, name:

species_df_matched <- gbif_species_name_match(df = species_df, name = "name")

## [1] "All column names present"

As the name argument has "name" as default value, the code above is equivalent to:

species_df_matched <- gbif_species_name_match(species_df)

or using pipe %>%:

species_df_matched <- species_df_matched %>% gbif_species_name_match()

By default gbif_species_name_match returns the following GBIF fields: usageKey, scientificName, rank, order, matchType, phylum, kingdom, genus, class, confidence, synonym, status, family.

Take a look at the updated data:

kable(species_df_matched)

Notice that GBIF fields whose name is already used as column name are automatically renamed by adding suffix 1. In our case, input data.frame species_df contains already a column called kingdom. The GBIF kingdom values are returned in column kingdom1:

species_df_matched %>% select(kingdom, kingdom1)

## # A tibble: 5 x 2
##   kingdom  kingdom1
##   <chr>    <chr>   
## 1 Animalia Animalia
## 2 Plantae  Plantae 
## 3 Plantae  Plantae 
## 4 Plantae  Plantae 
## 5 Plantae  Plantae

You can also specify which GBIF fields you would like to have:

species_df %>% 
    gbif_species_name_match(
      gbif_terms = c(
        'scientificName', 
        'family',
        'order',
        'rank',
        'matchType',
        'confidence',
        'status')) %>%
    kable()

## [1] "All column names present"

The function inborutils::gbif_species_name_match is a wrapper around rgbif::name_backbone, so you can pass any argument of name_backbone. For example, you can set strict = TRUE to fuzzy match only the given names, but never a taxon in the upper classification:

species_df %>% 
    gbif_species_name_match(strict = TRUE) %>%
    kable()

## [1] "All column names present"

These are all accepted parameters of name_backbone: ‘rank’, ‘kingdom’, ‘phylum’, ‘class’, ‘order’, ‘family’, ‘genus’, ‘strict’, ‘verbose’, ‘start’, ‘limit’, ‘curlopts’. See ?name_backbone for more details.

For Python users, there is a similar (but no longer maintained) function in inbo-pyutils.