In RS-eco/traitdata: Easy access to various ecological trait data

knitr::opts_chunk$set(collapse = TRUE, comment = "#>")

Getting started

Load the traitdata, dplyr and ggplot2 library

library(traitdata)
library(dplyr, quiet=TRUE)
library(ggplot2)

Data query

To connect to a table within our database, we use the data() function.

data(passerines)

Now we can use standard R calls to have a look at the table. Here, we check the class and first 6 rows of the table:

class(passerines)
head(passerines)

Basic verbs

Select certain columns, here: Genus, Species and Winglength

data(passerines)
select(passerines, Genus, Species, Wing) %>% head()

Only get bird species from the Auk family

Note: Only passerines has a column named Family. elton_birds and elton_mammals also contain information on Family, but the specific column has a different name.

passerines %>% select(scientificNameStd, Genus, Species, Family, Wing) %>% 
  filter(Family == "Cardinalidae") %>% head()

Order the data according to certain variables, here by Family

passerines %>% select(scientificNameStd, Genus, Species, Family, Wing) %>% 
  arrange(Family) %>% head()

Summarise the number of entries (not number of species) for each family (as some species might have multiple records).

passerines %>% select(scientificNameStd, Family, Genus, Species) %>% group_by(Family) %>% 
summarise(n_family = n())

We select only male animals, extract their Bill length, width and depth and arrange the data according to bill length.

# Get passerine data
data(passerines)

# Run queries on data
passerines %>% filter(Sex == 1) %>% 
  select(`Bill L`, `Bill W`, `Bill D`) %>% 
  arrange(`Bill L`) %>% head()

Grouping

This query groups all the taxa by Family, counts the number of Species per Family and removes all Families that have less than 10 species.

Note: With n_distinct we now get the actual number of species, not counting duplicate entries of one species.

by_family <- group_by(passerines, Family)
count <- summarise(by_family, n = n_distinct(scientificNameStd))
count <- filter(count, n > 10)
count

We can pipe dplyr operations together with %>% from the magrittr R package. The pipeline %>% takes the output from the left-hand side of the pipe as the first argument to the function on the right hand side.

passerines %>% group_by(Family) %>% summarize(n=n_distinct(scientificNameStd)) %>% 
  arrange(desc(n)) %>% head(10)

We can also plot this

passerines %>% group_by(Family) %>% tidyr::drop_na(Family) %>% 
  summarize(n=n_distinct(scientificNameStd)) %>% 
  arrange(desc(n)) %>% head(15) %>% collect() %>% ggplot(aes(x=Family, y=n)) +
  geom_bar(stat='identity',color='skyblue',fill='#b35900') + xlab("") + 
  ggtitle('Top ten families with highest number of species') + coord_flip() + ylab('Total number of species')

Wing length and kipp distance for females and males

Get mean wing length, kipp distance and sample size per species and sex

#Get passerines data
data(passerines)

#Get mean wing length and sample size per species and sex
passerines %>% group_by(scientificNameStd, Sex) %>% filter(Sex %in% c(1,2)) %>% 
  summarise(wing_length = mean(Wing), 
            sample_size = n()) %>% head(5)

Now we want to plot a boxplot of wing length

passerines %>% group_by(scientificNameStd, Sex) %>% filter(Sex %in% c(1,2)) %>% 
  head(15) %>% collect() %>% ggplot(aes(x=as.factor(`scientificNameStd`), y=Wing, fill=Sex)) + geom_boxplot() + coord_flip() + 
  xlab("") + ylab("Wing length")

Joins

Let's join passerines and migbehav_birds by scientificNameStd, only retaining species for which both datasets have entries.

migbehav_passerines <- inner_join(passerines, migbehav_birds, by="scientificNameStd", copy = TRUE)

Note: You need to be careful when merging datasets, as some datasets have unique values for one species, while other datasets have multiple records for one species. I also haven't cross-checked if data entries occur multiple times in different datasets. Please also have a look at the glossary before merging multiple datasets, as some datasets contain variables with the same names.

Now, we can identify the ten bird species with the longest bill length in our dataset and check their migratory status

migbehav_passerines %>% arrange(desc(`Bill L`)) %>% select(scientificNameStd, `Bill L`, Migratory_status) %>% unique() %>% head(10)

Alternatively, we might want to join datasets, which contain the same variables. For example, we want to get the Body mass of all species from elton_birds and elton_mammals.

data(elton_birds); data(elton_mammals)
eltontraits <- full_join(elton_mammals, elton_birds)
eltontraits %>% select(scientificNameStd, BodyMass.Value) %>% 
  arrange(desc(BodyMass.Value)) %>% head(10)