Skip to contents

pre-processing-data

pre-processing-data.Rmd

Background

In this tutorial I will present how data on longevity and fossil occurrence can be used to investigate the imprints of species competition in different levels. This will be done with the help of the new R package Buccaneer

Data

We will use the data from Faurby et al (2024) to demonstrate how the new Buccaneer R package can be used in processing and data analysis.

This data contains information on fossil data of all extant and extinct species of mammalian carnivores and related extinct groups (Carnivoramorpha, Hyaenodonta and Oxyaenidae)retrieved from Paleobiology Database and New and Old Worlds database.

Loading data 

library(dplyr)
df.faurby <- read.csv(here::here("data", "data_faurby.csv"), sep = ";") # dataset from https://royalsocietypublishing.org/doi/10.1098/rspb.2024.0473
df.body.mass <- read.csv(here::here("data", "Inferred_Size_Carnivoramorpha.csv"), sep = ",") # dataset from https://royalsocietypublishing.org/doi/10.1098/rspb.2024.0473
taxonomic_placement <- read.csv(here::here("data", "taxonomic_placement.csv"), sep = ";")

General cleaning and processing

Here I will filter the original data keeping only Carnivora species from suborders Caniformia and Feliformia and only records from Eurasia

taxonomic_placement2 <- 
  taxonomic_placement |> 
  filter(Suborder == "Caniformia" | Suborder == "Feliformia")


# adding a new column with genus and species to taxonomic placement and occurrence records
taxonomic_placement2$genus.species <- paste(taxonomic_placement2$Genus, taxonomic_placement2$Species, sep = "_")
df.faurby$genus.species <- paste(df.faurby$Genus, df.faurby$Species, sep = "_")


df.faurby2 <- 
  taxonomic_placement2 |> 
  left_join(df.faurby, by = c(genus.species = "genus.species")) |> 
  select(-Family.y, -Genus.y, -Species.y) |> 
  rename(Family = Family.x, Genus = Genus.x, Species = Species.x)

# keeping only Europe and NA
df.faurby3 <- 
  df.faurby2 |> 
  filter(Continent == "Europe" | Continent == "Asia")

Keeping the site ID from PBDB. I will use NOW ID only when PBDB is absent

df.faurby4 <- 
  df.faurby3 |> 
  mutate(site = ifelse(is.na(PBDB_ID), NOW_ID, PBDB_ID))

Processing data on body mass by calculating mean value for each species

df_body_mass <- 
  df.body.mass |> 
  group_by(Genus) |> 
  summarise(mean.size = mean(Size_Est))

Processing and flagging data with Buccaneer

I will use the function clean_occ_fossil to make simple data cleaning, processing and flagging the occurrence records. In this example I will remove any record with NA in any of the columns and flag records with the maximum and minimum age range greater than 5 myr, calculate midpoint age, time of origination (TS) and time of extinction (TE)

df_occ_faurby <- 
  clean_occ_fossil(df.occ.fossil = df.faurby4, 
                 method.ages = "midpoint",
                 thresh.age.range = 5, 
                 species = "Genus",
                 Max.age = "Maximum_Age",
                 Min.age = "Minimum_Age", 
                 site = "site",
                 lat = "Latitude",
                 lng = "Longitude",
                 remove.sub.species = TRUE, 
                 comp.TS.TE = TRUE, 
                 group = "Suborder")
df_occ_faurby <- 
  df_occ_faurby |> 
  filter(species != "Delete")

# filtering TS and TE per species

df_TS_TE_faurby <- 
  df_occ_faurby |> 
  dplyr::distinct(species, .keep_all = T)

# joining TS and TE with body mass

df_TS_TE_mass <- 
  df_body_mass |> 
  left_join(df_TS_TE_faurby, by = c(Genus = "species")) |> 
  rename(species = "Genus") |> 
  filter(!is.na(TS) & !is.na(TE)) |> 
  filter(!is.na(site))

# creating time intervals (bins)

bins <- seq(from = max(df_occ_faurby$Max.age), to = min(df_occ_faurby$Min.age), by = -5)