For months, the data looked like an ordinary shark. A pop-up satellite archival tag attached to a pregnant porbeagle shark off Cape Cod tracked the animal through its usual pattern: diving deep during the day, sometimes past six hundred metres, then rising back up to shallower water, roughly a hundred to two hundred metres down, at night, tracking temperatures that shifted with depth the way the surrounding ocean’s temperatures do. Then, for roughly four days before the tag detached and floated to the surface, the pattern broke.
The depth readings narrowed into a tighter band. The daily vertical migration stopped. And the temperature, instead of falling and rising with depth the way seawater does, held in an unusually narrow, elevated band, roughly 16 to 25 degrees Celsius, well above what the surrounding water at those depths would predict. Something was keeping that tag warm and moving it up and down independent of the ocean around it, and it wasn’t the shark it had been attached to.
The finding, from a team led by Brooke Anderson at Arizona State University with James Sulikowski at Oregon State University and colleagues at the Atlantic Shark Institute, was published in Frontiers in Marine Science in September 2024. The porbeagle in question was a 223-centimetre female, confirmed pregnant, tagged in Northwest Atlantic waters southeast of Cape Cod.
Reading the data as a crime scene
Anderson’s own first instinct was to doubt what the numbers were telling her. “I honestly kept trying to talk myself out of it being a predation,” she told Popular Science. That scepticism is a reasonable starting point for anyone reading tag data, since equipment can drift, sensors can fail, and unusual readings are just as often instrument error as they are evidence of anything happening in the water.
What ruled out equipment failure here was the specific shape of the anomaly. A malfunctioning tag tends to produce noise, not a clean, stable signal. What the researchers saw instead was consistent, elevated, and detached from the ambient temperature cues that had tracked the shark’s real depth changes for months beforehand. That combination, a warm and constant reading unrelated to the true water temperature at depth, matches what would be expected if the tag were sitting inside the stomach of a predator capable of keeping its core substantially warmer than the surrounding sea.
A handful of shark species, including white sharks and shortfin makos, have this capacity, sometimes called regional endothermy: specialised blood vessel arrangements that let them retain metabolic heat in muscle and organs rather than losing it all to the surrounding water. The researchers favour a white shark as the likely predator over a mako, based on the depth pattern during those four days, which more closely resembled the extended time white sharks are known to spend in deeper, mid-water zones rather than the faster, shallower oscillations typical of mako diving behaviour. This is pattern-matching, not confirmation, and the paper presents it as the best-supported explanation rather than a settled one.
What this method has shown before, and what it can’t show
Using tag data this way to infer predation after the fact isn’t new. Researchers have applied similar logic to detect predators eating tagged fish in earlier work on pop-up satellite archival tags, and comparable cases have turned up in Mediterranean spearfish research. What makes this instance notable is the animal on the receiving end: porbeagles are themselves a substantial predator, listed as vulnerable to endangered across parts of their range depending on the population, and direct evidence of one being eaten by something else had not been documented before, according to the paper’s authors.
The four days of anomalous data can’t say precisely when or how the shark died, only that whatever swallowed the tag maintained an unusually stable, warm internal environment for that stretch before the tag was expelled or the carcass decomposed enough to release it. Nor does a single tagged animal establish how often this happens to porbeagles generally. One recorded predation event, however clearly diagnosed, is a single data point.
An ordinary tool doing an unexpected job
What’s striking about the case is less the shark than the tag itself, doing a job it wasn’t built for. Pop-up satellite archival tags exist to answer questions about where and how deep animals travel. That it can double as an inadvertent witness to a predation event, recording temperature and depth data that outlives the animal it was attached to, is a reminder that the instruments researchers deploy often end up measuring more than the thing they were designed to track.
The porbeagle’s own tag told a story its makers never anticipated, and the closest anyone has come to reading it is a plausible reconstruction built from patterns that don’t quite add up any other way.