A simulation-based framework to detect marine animal-vessel interactions using tracking data
This paper introduces a simulation-based framework, implemented in the R package *intersimR*, that effectively distinguishes genuine behavioral responses (attraction and following) from incidental co-occurrence in marine animal-vessel interactions, revealing that while fishing vessels constitute a small fraction of traffic, they drive the majority of significant interactions with Scopoli's shearwaters.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine the ocean as a giant, bustling highway where two very different groups of travelers are constantly crossing paths: the wild animals swimming or flying above, and the massive metal ships cutting through the waves. For a long time, scientists have worried that these crossings aren't just random accidents. They fear that when animals and ships get too close, it can lead to dangerous situations like animals getting caught in fishing nets or being hit by boats. To understand if this is really happening, researchers use special "backpacks" (tags) on animals to track their every move, and they use digital logs from ships to see where the boats are going. The big question is: when an animal and a ship are near each other, is the animal actually paying attention to the ship, or are they just two strangers who happened to be in the same place at the same time? It's like trying to figure out if a person waving at a bus is actually waiting for that specific bus, or if they are just waving at a friend who happens to be standing near the bus stop.
This is exactly the puzzle a team of researchers tackled in a new study. They created a clever computer game—a simulation—to help them tell the difference between a real connection and a lucky coincidence. They took real data from 2,705 foraging trips by Scopoli's shearwaters (a type of seabird) and matched it with the paths of thousands of ships in the northwestern Mediterranean. First, they flagged every time a bird and a ship got close, finding 1,567 "candidate" moments. But here is the trick: they didn't just assume the birds were following the ships. Instead, they ran a simulation where they created thousands of fake bird paths that moved randomly, with no interest in ships at all. They then compared the real birds to these fake, clueless birds.
The results were eye-opening. The computer showed that many times when a bird and a ship were close, the bird was actually just minding its own business, and the ship had simply drifted by, or the bird had stopped moving and the ship had approached it. In fact, the simulation proved that about 46.5% of the close encounters were just "incidental proximity"—pure chance. However, the other 53.5% of the time, the birds were genuinely attracted to the ships or were actively following them.
The study also revealed a fascinating pattern about which ships the birds cared about. Even though fishing vessels made up only about 3% of all the ships recorded, they were responsible for a huge chunk of the action: 69% of the attraction events and 74% of the following events. The birds seemed to ignore most non-fishing ships, treating them as background noise. The researchers also noticed that birds were more likely to follow ships during the day, and that following behavior dropped when the birds spent more time sitting still.
To make sure their new method was solid, the team tested it against older, simpler ways of counting interactions. They found that the old methods were very sensitive to how close the animals and ships had to be to count as an "interaction," often flagging too many false alarms. In contrast, their new simulation-based approach remained steady and reliable, correctly identifying the real behavioral responses. They even packaged this tool into a free software program called intersimR so other scientists can use it to separate real animal behavior from random chance, helping to protect wildlife by understanding exactly how they react to the busy human world on the water.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.