As most of the world experienced this summer, plans got off track due to COVID-19. My original summer plan was to go to Guyana, South America for eight weeks to teach students how to band birds. (Bird banding is a mark-recapture technique used to gather information on bird populations, physiology, reproduction, and more.)
Then, Coronavirus happened. I was fortunate enough to be able to pivot my funding to go towards a five-year ongoing study about Hermit Thrushes breeding in North Carolina’s Appalachian Mountains. I helped a team of researchers collect data from 2016 through 2020 on Hermit Thrushes, including tracking thrushes using radio telemetry and collecting massive amounts of vegetation data on the field site. Spending this summer in Raleigh allowed me to begin to analyze this data for future publication purposes. One of the goals that I worked toward this summer was modeling the home ranges of nine breeding Hermit Thrushes. (If you’re not sure what a home range is, no worries: home range is the space that an animal uses to get all of the resources it needs for survival and reproduction.)
To begin, I had to ensure that the telemetry data was properly uploaded to Movebank (movebank.org), a global database for animal tracking data. I worked with a fellow NC State student and with a Movebank technician to correctly format the data. This involved extensive trial and error–I would look at each Thrush’s data that I had uploaded to Movebank, find issues, ask the Movebank technician how to remedy the issues, fix the issues…and then later find more problems. This taught me how valuable it is to standardize your data from the beginning and ensure that all of your spreadsheets have the same column names, with every cell containing data in the same format.
Once I had finally ensured that all of the telemetry waypoints were properly uploaded to Movebank, it was time to analyze the data. To turn GPS waypoints into home ranges for each bird, I had to import the Movebank data into R. I used a continuous-time movement modeling package in R Shiny to estimate the home ranges for each bird. This was just as tedious as getting the original telemetry data into Movebank, with seemingly incessant failures and corrections. Eventually, my colleagues and I were able to upload the telemetry data into R Shiny and estimate the home range for each bird. How exciting! It was amazing to see that all of the perseverance and hard work put into this project–capturing the birds to put little radio transmitters (trackers) on their backs, crashing through the forest to find the birds, struggling to manipulate the data into two different programs–finally paid off in the form of beautiful kernel density maps like the one pictured here.
So, what’s next? Now, I am continuing to work with the telemetry data in R to calculate the percentage overlap of the home ranges of each individual. I am looking forward to continuing the data analysis and publishing this data in an academic journal in 2021.