Skip to main content

College-level lab exercise exploring raptor migration data

I wrote a lab exercise for an upper-level college course in conservation biology at Swarthmore College that was held in April 2013. The exercise is titled "GIS and conservation biology: A case study with animal movement data". After a 30-min presentation introducing GIS applications for biology, methods for tracking individual animal movements, and the Movebank project, the students worked in pairs on the exercise. There are three parts:

  1. Exploring the public study "Turkey Vulture Acopian Center GPS"
  2. Exploring the public study "Osprey Bierregaard North and South America GPS"
  3. Exploring both datasets annotated with land use data from the GlobCover 2009 dataset

Along with the lab handout I provided files containing reference data for both studies (for the ospreys especially, an important point is that many of them are juveniles—this information is stored in the reference data). In addition I downloaded the annotated datasets ahead of time, imported the both to an Excel spreadsheet, and summed the number of locations within each land use class for each study. Before parts 1 and 2 I briefly introduced both datasets (having talked with the data owners ahead of time and gotten some more details about the research).

The students had a good time checking out individuals on the maps, and the class was small enough that interesting questions and discoveries could be shared with the everyone. It took the students about 90 minutes to complete the exercise. This could be extended by making them compile the annotated data in Excel by themselves. If I were to do the exercise again, I would provide some additional material either during or after the lab to make sure that all of the students can answer the following questions:

  • The data show that osprey sometimes fly over water, but that turkey vultures do not. Why?
  • How do juvenile osprey's first migrations differ from those of adults? (Rob Bierregaard has some slides that show this better than can be illustrated in Movebank.)
  • What are causes of error or bias when using animal tracking data to understand the movements of an individual? a population? when looking at the annotated land cover data?

Also Part 3 could probably be improved by looking more closely at the land use data to come up with better questions.

Note that both of these datasets are currently being added to in Movebank, so if the exercise was used again, a few details would need to be updated.