Homework 13 — Map Your Own Species with GBIF

Download real occurrence records with rgbif and map where a species lives

Students choose a species, download its real occurrence records straight into R with the rgbif package, clean the coordinates, and map them with sf and geom_sf — the same recipe as the Bigfoot lecture, on data they pick.

Map Your Own Species 🌍

In the lecture you mapped Bigfoot sightings. Bigfoot is (probably) not real — but GBIF is: the Global Biodiversity Information Facility holds billions of real species-occurrence records, each with a latitude and longitude. The rgbif package pulls them straight into R. This week you pick a species and map where it actually lives.

Tip📥 What you’ll turn in

A rendered Quarto document (HTML or Word) that contains:

1. The species you chose and its scientific name
2. A map of its occurrences on the US states basemap
3. Short answers to the four questions in Part 5

Submit the .qmd and the rendered file to Canvas.

---

Part 1 · Pick a species

You need the scientific name (Genus species). Find it on https://www.gbif.org by searching the common name, or use one of these fun starters:

Common name	Scientific name
Monarch butterfly	Danaus plexippus
Bald eagle	Haliaeetus leucocephalus
American black bear	Ursus americanus
Eastern red-backed salamander	Plethodon cinereus
Coyote	Canis latrans
Venus flytrap	Dionaea muscipula

✏️ Your choice:

Common name:
Scientific name (Genus species):
Why you picked it:

---

Part 2 · Download occurrences with rgbif

Install rgbif once, then load your packages.

▶ Run this (install once, in the Console):

install.packages("rgbif")

▶ Run this at the top of your script:

library(tidyverse)
library(sf)
library(maps)
library(rgbif)     # download GBIF occurrence records

▶ Adapt this — put YOUR scientific name in:

# Pull up to 1500 US records that HAVE coordinates ------
gbif_result <- occ_search(
  scientificName = "Danaus plexippus",   # <- change to your species
  country        = "US",
  hasCoordinate  = TRUE,
  limit          = 1500
)

occ_df <- gbif_result$data   # the records live in $data
nrow(occ_df)

⚠️ Watch out! If occ_df has 0 rows, your name may be misspelled. Check it with name_backbone(name = "Danaus plexippus"), or search the name on gbif.org.

---

Part 3 · Clean and make it spatial

GBIF coordinate columns are decimalLatitude and decimalLongitude.

▶ Run this:

# Keep rows with real coordinates in the lower 48 -------
occ_df <- occ_df %>%
  filter(!is.na(decimalLatitude), !is.na(decimalLongitude)) %>%
  filter(decimalLongitude > -125, decimalLongitude < -66,
         decimalLatitude  >   24, decimalLatitude  <  50)

# Build the sf object (x = longitude first!) -----------
occ_sf <- occ_df %>%
  st_as_sf(coords = c("decimalLongitude", "decimalLatitude"), crs = 4326)

nrow(occ_sf)

✏️ Your turn:

Records with usable US coordinates:

---

Part 4 · Map it

▶ Run this:

# US states basemap (same as the lecture) --------------
states_sf <- st_as_sf(maps::map("state", plot = FALSE, fill = TRUE)) %>%
  st_set_crs(4326)

# Points on the map ------------------------------------
species_map <- ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = occ_sf, color = "darkgreen", alpha = 0.3, size = 0.7) +
  coord_sf(crs = 5070) +
  labs(title = "Occurrences of Danaus plexippus",   # <- your species
       caption = "Data: GBIF via rgbif") +
  theme_void()

species_map

# Save it ----------------------------------------------
ggsave("figures/my_species_map.png",
       plot = species_map, width = 7, height = 5, units = "in", dpi = 300)

---

Part 5 · Interpret (answer these)

✏️ Your turn — write 1–2 sentences each:

1. Where in the US does your species occur? Does the pattern match what you
   expected (range, habitat, climate)?

2. GBIF records come mostly from where PEOPLE look (cities, parks, universities).
   How might that sampling bias shape your map?

3. How is this map the SAME recipe as the Bigfoot map from class? Name the shared
   steps (get geometry -> get data -> ... -> plot).

4. One thing you'd explore next if you had another week (e.g., color by year,
   a per-state choropleth, or a different region).

---

Stretch (optional, +extra)

Turn your points into a per-state choropleth, exactly like the Bigfoot lecture: count records per state (stateProvince, lowercased), left_join() to states_sf, and shade with geom_sf(aes(fill = n)).

# Sketch — adapt the lecture's choropleth code ---------
occ_df %>%
  mutate(state = str_to_lower(stateProvince)) %>%
  count(state, name = "records")
# ... then left_join() to states_sf and geom_sf(aes(fill = records))

---

Getting unstuck

1. 0 records → check the scientific name with name_backbone(name = "...").
2. could not find function "occ_search" → library(rgbif).
3. Empty map → your coordinate filter may be too tight, or the species isn’t in the US — widen the box or drop the country = "US" filter.
4. Points off the map → confirm coords = c("decimalLongitude", "decimalLatitude") — longitude (x) first.
5. GBIF docs / rgbif — https://www.gbif.org/tool/81747/rgbif

💡 Key idea: you just downloaded real biodiversity data from a global database and mapped it — the exact workflow ecologists use in published papers.

---

Homework 13. Bring your map to class — we’ll compare species ranges.