Professional Documents

Professional documents should be easy to follow and get the point across to the reader in a concise manner.

Helpful components include:

document header with title, name, and date
- ideally, the date reflects the day the document is most recently rendered
all packages are loaded together at the top of the document
include code comments when appropriate
clean code indentation for lists, parameters, functions within functions, etc.
data visualizations such as maps and plots contain all necessary components such as a legend, axes labels, units, and appropriate palette (divergent, colorblind friendly, etc.)
folding code or sourcing separate scripts when appropriate to direct reader’s attention to the important output or to condense code
hiding unnecessary output such as warning messages or loaded packages
succinct documentation between steps, such as section headers, descriptions for an analysis step, and map/plot interpretation
complete and detailed data citations
function definitions should include warnings and errors that ensure the intended operations are executed

Elephant tracking in Krugar National Park

Elephants are intelligent and gregarious animals that tend to travel as a herd and seek out resources such as shrubs and water. Elephants in Krugar National Park were collared with GPS tracking devices from August 2007 through August 2009 (Slotow et al. 2019).

Objective: Use GPS point locations of individual elephants to visualize their movement throughout Krugar National Park in South Africa between 2007 and 2009.

Load packages and source functions

Show the code

library(tidyverse)
library(here)
library(sf)
library(tmap)
library(rnaturalearth)
library(osmdata)

# source external script that defines custom functions
source(list.files(path = here("course-materials"),
                  pattern = "functions.R",
                  recursive = TRUE,
                  full.names = TRUE))

Import data

Import the elephant movement data and configure its spatial components

metadata_df <- load_csv(filename = "elephants_metadata.csv",
                        directory = "course-materials",
                        communicate = FALSE)

elephants_df <- load_csv(filename = "elephants.csv",
                         directory = "course-materials",
                         communicate = FALSE)

# convert the elephant observations to sf objects
elephants <- to_spatial_points(data = elephants_df,
                               latitude_col = "location-lat",
                               longitude_col = "location-long")

Visualize data

Import the spatial boundaries for Kruger National Park to serve as a basemap for the elephant tracks

# set bbox to approximate boundaries of Kruger NP
kruger <- osmdata::opq(bbox = c(16, -35, 33, -22)) %>%
          add_osm_feature(key = "boundary", 
                          value = "protected_area") %>%
          add_osm_feature(key = "name", 
                          value = "Kruger National Park") %>%
          osmdata_sf()

# subset imported list of elements to just the geodataframe
kruger_bounds <- kruger$osm_multipolygons

Produce a map of elephant observations to visualize the individual movement of elephants

Show the code

title = "Elephant Observations\nKruger National Park\n2007-2009"

# basemap: Kruger National Park geometry
tm_shape(kruger_bounds) +
  tm_borders(col = "black", lwd = 0.5) +
  tm_fill(col = "black") +
  # overlay all elephant GPS locations 
  tm_shape(elephants) +
  tm_dots(col = "individual-local-identifier",
          palette = "Set3",
          # small dots to distinguish between individuals
          size = 0.001,
          border.col = "black",
          title = "Individual") +
  tm_layout(bg.color = "darkgrey",
            legend.bg.color = "white",
            title = title,
            frame = TRUE,
            title.position = c("center", "top"),
            title.size = 1.2,
            title.fontface = "bold",
            legend.frame = TRUE,
            legend.outside = TRUE,
            legend.position = c(0.08, 0.3),
            legend.title.size = 0.8,
            legend.text.size = 0.6) +
  # add compass with only North arrow
  tm_compass(position = c(0.0, 0.01),
             show.labels = 1)

Map interpretation

Elephants in Kruger National Park fitted with GPS collars are primarily located in the southern part of the park. The clustered tracks imply that this species is highly gregarious. Tracks that expand beyond park boundaries, notably on the western side, imply that the elephants could be seeking resources such as food or water outside of the park. Alternatively, error in the GPS location for these points could explain this perceived movement.

Futher analysis steps could include:

zooming in further to the realized niche
partitioning the data temporally
using metadata to advise how to subset the data to view fewer elephant locations at a time
overlaying other vector or raster data layers such as surface water and vegetation to understand if the elephants’ movement is driven by these resources

Data Citations

Data	Citation	Link
Elephant observations	Slotow R, Thaker M, Vanak AT. 2019. Data from: Fine-scale tracking of ambient temperature and movement reveals shuttling behavior of elephants to water. Movebank Data Repository. https://doi.org/10.5441/001/1.403h24q5	MoveBank Repository: Fine-scale tracking of ambient temperature and movement reveals shuttling behavior of elephants to water