Using AI and Citizen Science to Reveal the Secrets of Seabirds
Seabirds are top predators in ocean ecosystems. They link the land and the ocean and act as indicators of ocean environmental change. By studying seabirds, researchers can learn a lot about the ocean ecosystems and its overall health. Jonas Hentati Sundberg, Associate Senior Lecturer at the Department of Aquatic Resources, SLU – Swedish University of Agricultural Sciences in Lysekil, Sweden, from SLU, led the construction of the Stora Karlsö Auk Lab in Stora Karlsö, a small island in the Baltic Sea about six kilometers west of the island of Gotland. The island is known for its huge colony of common murres (Uria aalge), or common guillemots. Jonas has been studying seabirds for more than 20 years. A few years back, he and his colleagues realised the limit of traditional techniques for biological or ecological studies, and the idea of building a giant nest box arose.
(Credit: Baltic Seabird Project)
The Auk Lab is a steel construction jutting out from Stora Karlsö’s cliff edge about 50 meters above sea level – the world’s largest nest box with space for 600 pairs of breeding seabirds in the middle of the common murre colony. It’s constructed in two parts: the inner part is the researchers’ hut, where equipment is mounted; the outer part mimics limestone cliffs, common murres’ natural habitat. Researchers can either be there physically to observe common murres from a close range, or sit remotely in their office looking at real-time videos captured by CCTV cameras connected over 4G.
In collaboration with WWF, a citizen science project was launched. A live stream was linked to WWF’s website during the early summer of 2019 and the public could watch live stream of common murres in their breeding season. A simple user interface was built to guide the reporting of behaviours such as copulations, fights and socialising or other unusual signs of distress. This attracted around 15,000 viewers within four weeks in 2019 and more than 2000 observations were submitted.
The massive amount of data generated during the 2000 hours of streaming made it impossible for researchers to study in a traditional sense. A two-day hackathon was held at AI Innovation of Sweden in Gothenburg at the end of 2019. The winner of the hackathon, Team SkyRoads, developed a strong algorithm to detect, identify and analyse the seabird data, which helped the researchers immensely. To ensure the success of any hackathon, says Jonas, it is important to have clearly defined tasks and create an environment that is inclusive; but for him, a hackathon is not a competition, rather a joint learning activity where people with different backgrounds and knowledge come together.
(Credit: Jonas Hentati Sundberg)
For researchers like Jonas, making use of AI and citizen science revolutionises seabird monitoring and ways to understand and answer questions about the ocean, and he believes opportunities will expand dramatically. They are also working in parallel on a marine drone project to monitor fish stocks as well as other oceanographic parameters. He admits that AI in research is a long-term and labour-intensive investment – a lot of time and effort are required for annotation, model development, testing and validation before one can start to reap the benefits. Citizen science is a possible way to help validate and annotate, but he stresses that citizen scientists are by no means cheap labour – researchers need to make it a reciprocal process, both fun and meaningful so it is a learning experience for the citizen scientists as well.
For more information on the Baltic Sea Project, see: http://www.balticseabird.com
On June 15, 2021, common murre will go live again on WWF website: https://www.wwf.se/english/baltic-guillemots-live-stream/.
ODF Sweden’s Spring 2021 Webinar Series continues on April 30, “Law as facilitator (or obstacle) for science-based marine governance”.