Marine scientists have discovered that using drones to monitor the physical condition of dolphins is nearly as accurate as examining a dolphin in person — but much easier to practice.
By analyzing unoccupied aerial system photographs of stranded pygmy killer whales, an elusive oceanic dolphin species, and comparing their data to necropsy samples, scientists have proven that so-called UAS photogrammetry is a reliable and convenient alternative to directly viewing dolphins nearby in the field or in the lab. This breakthrough may illuminate undiscovered threats to vulnerable dolphin populations, according to the researchers who took these images and detailed their findings in an April 14 study from Scientific Reports.
Unoccupied aerial system photogrammetry is a specialized type of drone photography that has been used in related research to track various whale species, such as humpbacks, baleens and gray whales. Less clear was whether the same techniques could be used on smaller organisms.
"The real beauty here is that this can be applied to almost essentially any dolphin species in any particular region," said lead author Jens Currie, chief scientist at the Hawaii-based Pacific Whale Foundation. "Having this validated as an approach to monitoring weight loss and body condition, which we know is related to health — it has huge ramifications for our ability to begin to model the impacts of climate change and human threats on many dolphin species."
Pygmy killer whales inhabit deep tropical and subtropical waters around the globe, and they truly are "pygmies" compared to their famous namesake: Where pygmy killer whales weigh up to 500 pounds, a real killer whale weighs up to 11 tons.
Pygmy killer whales also face various threats such as oceanic noise pollution and entanglement in fishing gear, according to the National Oceanic and Atmospheric Administration, but the International Union for Conservation of Nature's Red List categorizes them under "least concern," and not endangered or vulnerable.
Currie and his colleagues flew a drone over pygmy killer whales that entered the Hawaiian island of Maui's Maʻalaea Bay on Sept. 13, 2019, with some remaining stranded until the following Oct. 3. Sensing a unique opportunity to try the UAS approach, the researchers made four photo runs over a 17-day period from September to October to monitor the dolphins' behavior and physical condition.
While Currie could not initially be certain, the drone images appeared to show the dolphins growing skinnier. Starving dolphins, for instance, begin to grow "peanut heads," as the fat and muscle around the ovular dome of their skull withers away, revealing a divot that cuts a peanut-like figure, rather than one more egg-like, according to Currie.
These dolphins were visibly tired, appearing to rest or mill about, when they would usually be looking for food over a larger area than where they remained in Maʻalaea Bay.
Analysis of the images would eventually show that the dolphins were losing about 2% of their total body mass per day, equivalent to 2.9 to 3.4 kilograms. This added up to a 27% loss of body mass over 17 days.
To discern the size of these dolphins from photos alone, the researchers needed to know the altitude of the drone when it took each picture. From there, they could equate the size of one pixel to a defined measurement on the stranded dolphins. For example, one pixel could equal five centimeters of space on a dolphin.
With that information, Currie and his team could determine the length and volume of each dolphin, though the latter required additional steps and was ultimately less accurate because the researchers could not account for fin size using the drone image.
They tested the hypothetical volume of the stranded dolphins with a water displacement experiment. They created a crate-like object whose volume matched the measurements from the image, and recorded how much water it displaced when placed in a tank.
The researchers could eventually verify their findings when they received two of the dolphins that died during the stranding they photographed. These dolphins, to Currie's surprise, showed extensive scarring, perhaps as a result of swimming near the sea floor, and their stomachs contained plant matter, rather than the fish or squid they normally eat.
"That's essentially indicative of a starving animal, that they're just trying to opportunistically consume anything," Currie said in an interview with The Academic Times. "These certainly aren't herbivores."
Drawing on measurements taken from post-stranding necropsies, the team could see that the UAS readings for each respective dolphin had only a 1.27% and 1.29% rate of error while measuring length, and a 10% and 13% rate of error for volume, mostly because the volumes of a dolphin's fins and tail are difficult to estimate with a formula.
Because the drone measurements were so accurate, scientists such as Currie can now apply this technology to see how various dolphin species are responding to threats and foraging disturbances, whether from human activity or climate change.
"What we can do now is compare disturbed populations to undisturbed populations to see if there's a change in body condition," Currie said. "And if there is, we can then begin to provide a good line of evidence that this potential threat or activities are having an impact on that population."
Anthropogenic activity is a well-recognized scourge for whales and dolphins alike. An analysis of killer whale deaths published in December showed that humans, especially via boats, are responsible for a significant share of killer whale deaths.
In dolphins, prolonged exposure to vessel traffic, competition with fisheries or declines in prey populations can disrupt feeding, and possibly lead to starvation or stranding. Now, using drones, scientists can closely monitor how these events are affecting dolphin health, opening up new avenues of study without needing a dolphin on hand.
The study, "Rapid weight loss in free ranging pygmy killer whales (Feresa attenuata) and the implications for anthropogenic disturbance of odontocetes," published April 14 in Scientific Reports, was authored by Jens J. Currie and Stephanie H. Stack, Pacific Whale Foundation; Kristi L. West, Hawaii Institute of Marine Biology and College of Tropical Agriculture and Human Resources; Martin van Aswegen and Fabien Vivier, University of Hawaii at Manoa; and Lars Bejder, University of Hawaii at Manoa, Aarhus University and Murdoch University.