Mic’ed Up: Reindeer Edition

Turnip the reindeer is beautiful, and she knows it. She has a nearly-white coat, fading smoothly into darker patches along her sides, pale, nearly-symmetrical antlers, and an imperious look in her eye. Turnip and 41 other reindeer hold court at the University of Alaska Fairbanks Large Animal Research Station, situated on an old homestead with a stunning view of the Alaska Range mountains and Denali. Turnip was one of eight LARS reindeer recently outfitted with collars carrying audio-recording devices as part of PhD student Megan Perra’s work developing methods to monitor wildlife using audio data. Audio recorders have been used to study vocal and hard-to-track wildlife like songbirds, bats, and frogs for several decades. Advances in technology and a growing call for non-invasive methods to monitor sensitive wildlife have led researchers like Megan to develop new kinds of recording equipment and computer programs to process the many hours of audio recordings that come from them. Megan’s “animal-borne acoustic recorders” measure about 4-inches long and weigh around 175 grams, small enough to be mounted onto GPS tracking collars that biologists use to study wild caribou and other large mammals. Tracking collars are invaluable for gathering movement patterns but are limited in how detailed they can be. Acoustic recorders can provide highly-detailed insight into the day-to-day experience and behavior of wild animals, documenting behavior (e.g., how much time per day did they spend eating?), disturbance (e.g., did their behavior change when they heard that plane fly by?), and their environment (e.g., how intensely are caribou being harassed by biting insects?). Megan has spent several years developing and troubleshooting her travel-sized acoustic recorders. Last December, she and a few colleagues traveled to LARS to test her latest version on Turnip & Co. First, we fitted the reindeer with the tracking collars carrying Megan’s recorders, with the help of LARS staff. The recorders have a 3D-printed casing which holds the microphone, batteries, and a memory card. The casing is sealed with epoxy and attached to the collars. Each recorder also has an external antenna that gives off a radio signal so it can be found when it falls off the reindeer or caribou. Soon, Turnip, Navajo, Suki, Lily, Ginger, Wiki, Ruth, and Katara were back in their pasture, new bling in place. The next day, we returned, decked out in head-to-toe snow gear, to film the collared reindeer. Amidst some serious side-eye from the herd, we spent the next few afternoons taking video of them eating, napping, scraping their antlers on an unfortunate sapling, and doing the things reindeer do. These videos are critical to Megan’s work, because they allow her to compare each behavior in the videos to the sounds on the audio recordings. Each time we started a new video recording, we blew a unique pattern on a whistle (or, in my case, a duck hunting call) which would be picked up by the audio recorders. That way, Megan could go back later and line up each video clip with the audio recorded by the acoustic devices. Our biggest challenge was keeping the camera batteries alive in the sub-zero temperatures! By the end of the third day, we had taken about 20 hours of video and convinced the reindeer we were too boring to pay much attention to. The acoustic monitors came off the reindeer in early January, and now Megan’s real work begins: she has to go through each recording and identify every time a sound was recorded that correlates to a behavior of interest. She is working with other acoustic ecologists to develop computer programs that can do some of this work, sifting through thousands of hours of audio and tagging distinct sounds like reindeer vocalizations, chewing, or urinating. However, the computer programs have to be “taught” what each behavior sounds like, so, for now, Megan will be spending a lot of quality time listening to and categorizing the ambient sounds of Turnip and her comrades.

This work is funded by Government of Northwest Territories, World Wildlife Fund, National Science Foundation INTERN grant, and National Science Foundation Navigating the New Arctic grant number 2127271.