WELLFLEET — The unique sounds known as “signature whistles” that all the dolphin species found in Cape Cod waters use to identify themselves may be the key to finding out about dolphin strandings — like the one that took place at the Gut in Wellfleet Harbor on June 28 — before they even happen.

Cape Cod is a global hotspot for dolphin strandings. Stacey Hedman, communications director at the International Fund for Animal Welfare (IFAW), which manages stranding responses on the peninsula, said the 10-year average is 315 strandings per year. And most of them happen here: 75 percent of Cape Cod’s mass strandings (in which more than two dolphins strand) occur in Wellfleet, Hedman said.

Scientists do not know why dolphins strand. Without a clear cause, these events are very difficult to anticipate, according to Molly Dent, an instrument specialist at Mote Marine Laboratory and Aquarium in Sarasota, Fla. And because they are usually a surprise, it takes time for responders to arrive, which puts dolphins at increased risk of injury or death.

Preventing dolphin deaths was the motivation behind Dent’s research into Wellfleet Harbor strandings as an undergraduate at Hampshire College in 2021. Her goal was to identify factors that preceded strandings to help rescuers predict when they would happen.

A key component of her research was a collection of acoustic recordings from a buoy located about a quarter mile east of Jeremy Point. The recordings, made between 2014 and 2020, were mostly collected by her adviser, Laela Sayigh, who is a professor of animal behavior at Hampshire and a research specialist at the Woods Hole Oceanographic Institution.

Dent compared the recordings with data from IFAW on nearby strandings. What she found was that, except in cases of very few dolphins stranding, the sounds dolphins made during the 48 hours before strandings were different from the animals’ ordinary calls.

The key difference was those so-called signature whistles. Though these noises sound like any other dolphin sound to the human ear, Dent said, marine mammals use them to identify themselves to one another. They give these calls in many circumstances, such as between a mother and calf, when a dolphin is returning to its pod, or when they experience stress.

Dent found that, on recordings from the day of or immediately before a stranding, an average of 82 percent of dolphin whistles are signature whistles. On other days, the average is only 54 percent. Dent thinks this is likely because dolphins are more stressed in the lead-up to a stranding.

Dent and Sayigh wanted to use their findings to develop a warning system that would send an alert to IFAW when signature whistles are more prominent, the idea being to learn about imminent dolphin strandings before they happen. But identifying the whistles was time- and labor-intensive. Dent had to slow the recordings down to 25 percent of normal speed and scan the results through a spectrogram — a visual depiction of the sound’s frequency — to find the whistles. Then there was more work to be done to identify which of the whistles were signature whistles specifically. If there was a high concentration of dolphin sounds, five minutes of a recording could take a whole day to analyze, Dent said.

A Feasible Approach

A warning system would need to be automated, Dent said, with a computer trained through machine learning to identify different dolphin noises in real time. Before the researchers could begin that step, though, they ran out of funding to pursue the project further.

Their idea, however, is “quite feasible,” said Christopher Clark, senior scientist emeritus at the Cornell Lab of Ornithology. He has studied marine mammal bioacoustics for 48 years and was not involved in Dent’s research.

“We aren’t having to invent new technology,” he said, pointing to the Lab of Ornithology’s Merlin Sound ID app, which can identify 1,382 species of birds from their vocalizations in real time. He suspected that such an early warning system would cost in the range of six figures. “We’re not talking about tens of millions of dollars,” he said.

Clark did note that the study was ambiguous as to whether signature whistles are an effective predictor of dolphin strandings, but he said that implementing such a system would be a good way to determine whether they are. Even knowing that a pod of dolphins is in the harbor at all could be useful information to responders, he said.

Brian Sharp, the director of marine mammal rescue at IFAW, told the Independent in an email that he, too, believes there is potential in a system like this. Knowing that a stranding may be about to occur could allow IFAW volunteers to herd animals into deeper water before low tide and stage equipment and responders in key areas.

Sharp thought one challenge might be for a system to distinguish between multiple simultaneous strandings from the same group of dolphins. One lone dolphin stranded on Eastham’s Thumpertown Beach in the early morning hours before the mass stranding happened in the Gut.

Five Days of Strandings

After the mass stranding in the Gut on June 28, a few dolphins continued to come ashore in the following days, Sharp wrote.

On June 30, approximately 20 dolphins stranded in Brewster, with three dying before IFAW could intervene. The following day, IFAW was monitoring five or six animals in Wellfleet Harbor; one stranded and was transported in a dolphin rescue van and released into water at Herring Cove Beach in Provincetown.

Tuesday, July 2, saw more strandings, with one dolphin on Rock Harbor Beach in Orleans, one on Boat Meadow Beach in Eastham, and 11 more in Wellfleet. The Orleans and Eastham dolphins were “humanely euthanized after showing signs of distress,” Sharp wrote, as were two of the Wellfleet animals. The other nine were released at Herring Cove.

Of the 125 dolphins that stranded over a five-day period, approximately 20 died naturally and seven were euthanized. One dolphin was released at Herring Cove with a tracker and was later reported tracking around four miles northeast of Marshfield. IFAW believes it to be part of the same group and that this particular stranding is now over.

Editor’s note: An earlier version of this article, published in print on July 11, incorrectly defined a mass stranding as involving more than one dolphin rather than more than two. The mass strandings in Wellfleet occur not just in the harbor but also in other locations in town.