How can we combine the geographic information from GBIF and the phylogenetic information from Open Tree?

In order to examine geographic distributions of species, it is useful to tie together information about phylogenetic relationships, which we can get from Open Tree of Life, with information about where individuals of that species are found.

library(rgbif)
library(rotl)
library(phytools)

Lets look for some fun mammal species, A polar bear, “Ursus maritimus”, a hippo “Hippopotamus amphibius”, a gorilla “Gorilla gorilla” and a wild boar “Sus scrofa”

polarbear <-"Ursus maritimus"
hippo <-"Hippopotamus amphibius"
boar <-"Sus scrofa"
gorilla <- "Gorilla gorilla"
mammal_spp <- c(polarbear, boar, gorilla, hippo)

We will use GBIF to search for records of these species, and arbitrarily chose the first record returned as the location for the species.

dat1 <- occ_search(scientificName = polarbear, fields=c('name','decimalLatitude', 'decimalLongitude'), limit = 1)
dat2 <- occ_search(scientificName =hippo, fields=c('name','decimalLatitude', 'decimalLongitude'), limit = 1)
dat3 <- occ_search(scientificName = boar,  fields=c('name','decimalLatitude', 'decimalLongitude'), limit = 1)
dat4 <- occ_search(scientificName = gorilla, fields=c('name','decimalLatitude', 'decimalLongitude'), limit = 1)
mammal_locs <- as.data.frame(rbind(dat1$data,dat2$data, dat3$data, dat4$data)) #Combine the data from each species
rownames(mammal_locs) <- mammal_locs[,1] # set the rownames to the taxon names
mammal_latlong <- as.matrix(mammal_locs[,c(3,2)]) #flip the columns to lat, long instead of long,lat

Instead of asking about each taxon one at a time, we can alternatively use a loop that gets the data for all taxa:

mammal_locs = data.frame(matrix(vector(), 0, 3))
for (taxon in mammal_spp){
  result <- occ_search(scientificName = taxon, 
                   fields=c('name','decimalLatitude', 'decimalLongitude'), limit = 1)
  mammal_locs <- as.data.frame(rbind(mammal_locs,result$data))
}
rownames(mammal_locs) <- mammal_locs[,1]  # set the rownames to the taxon names
mammal_latlong <- as.matrix(mammal_locs[,c(3,2)]) #flip the columsn to lat, long instead of long,lat

Now we have matrix of locations - we need a tree for our of species!

taxa <- tnrs_match_names(mammal_spp, context="Animals")
tr <- tol_induced_subtree(ott_id(taxa), label="name")
plot(tr)

enforce newick style tip names, by replacing spaces with underscores

rownames(mammal_latlong) <- gsub(' ','_',rownames(mammal_latlong))

Trees from OpenTree don’t automatically come with branch lengths - we need to infer some branch lengths for this tree in order to plot it.

tr_bl<-compute.brlen(tr)

Combine the phylogeny and the mapped locations

tdobj<-phylo.to.map(tr_bl,mammal_latlong,plot=FALSE)
plot(tdobj,direction="rightwards",ftype="i")

How to combine data from Open Tree of Life and GBIF

Emily Jane McTavish

2018-04-18

How can we combine the geographic information from GBIF and the phylogenetic information from Open Tree?