But after nine years, he saw with his own eyes that Old Timer had survived.
“It was heartwarming, because I realised it wasn’t just the old whales who were perishing,” Pack said. “Some of them were resilient.”
Historically, tracking the whereabouts of the whales has been done the hard way: by scientists using their own eyes to compare new fluke photos with old ones. But future studies of Old Timer and other humpbacks of all ages are set to be accelerated with artificial intelligence. And Pack hopes it will help him learn how, and why, some whales can withstand tough conditions.
Multiple humpback populations dwell in the North Pacific. Old Timer is part of a population that spends winters breeding in the waters around Hawaii and summers in southeastern Alaska, filling up on fish and tiny shrimplike animals known as krill. These humpbacks have been the subject of an ongoing scientific study, which began in 1976, when a marine mammal researcher, Louis Herman, began photographing the whales and their distinctive flukes.
Herman conducted annual surveys, amassing an enormous collection of tail pictures that allowed scientists to keep tabs on individual whales over the course of their lifetimes. These fluke photos, which now number more than 30,000, have provided new insight into the lives of whales, from their migration patterns to their social behaviours.
“It’s one of the longest ongoing scientific studies of humpback whales in the world,” said Pack, one of Herman’s former students and colleagues and now leader of the whale project.
The study is now entering the age of machine learning, with the help of an online platform called Happywhale, which collects whale fluke photos from scientists and members of the public from around the world. The Happywhale database currently contains roughly 1.1 million images of more than 100,000 individual humpbacks, said Ted Cheeseman, a co-founder of Happywhale and a doctoral candidate at Southern Cross University in Australia.
Artificial intelligence-powered photo matching algorithms help automatically identify the whales in submitted photos, aiding scientists in the field or others who need to look up previous sightings of a given animal.
“Happywhale has revolutionised our field and has made large-scale collaborations possible,” Pack said.
Earlier this year, Cheeseman, Pack and dozens of other researchers used Happywhale’s image recognition tool to estimate humpback whale abundance in the North Pacific from 2002 through 2021. Initially, the population boomed, climbing to about 33,500 whales in 2012.
But then it dropped sharply. This population decline coincided with the severe marine heat wave, when Pack last spotted Old Timer. It lasted from 2014 to 2016 and slashed the supply of fish and krill. “There’s a lot more we want to learn about the event, but it is quite clear: warmer waters mean food is less available overall, and what is available is more dispersed and deeper,” Cheeseman said in an email.
The Hawaii humpback population was especially hard hit, falling by 34% from 2013 to 2021. Although there had been some sightings of Old Timer reported after 2015, Pack was excited to finally set eyes on the whale himself. That excitement soon gave way to curiosity: why had Old Timer survived, when so many others had perished?
Now, Pack is hoping to dive deeper himself, with the help of Happywhale. He plans to investigate how humpbacks survived the lean years and whether there are any discernible patterns. Could Old Timer’s age have been an advantage?
“It is possible that Old Timer’s been around enough to be adaptable when certain food resources are limited,” Pack said.
The idea remains speculative, and it is not yet clear whether Old Timer was the exception or the rule. “How many whales like Old Timer were resilient to this devastation of marine resources?” he said.
This article originally appeared in The New York Times.
Written by: Emily Anthes
©2024 THE NEW YORK TIMES