Reviving archived data to better understand humpback whale behavior

Breathtaking. This is the best way I can describe the synchronized hunt of bubble-netting humpback whales in Southeast Alaska. First, an eerie trumpet-like vocalization resonates across the dark green swells of Frederick Sound. Baranof Island’s jagged, snow-capped peaks are prominent against the bright blue sky, their reflection dancing in the waves. As the vocalization grows louder, a spiral of foamy bubbles appears at the water’s surface. Then, a true spectacle - a dozen or so humpback whales, mouths wide open, erupt from the water. Herring can be seen frantically jumping out of the way, but few escape the gaping jaws of these coordinated feeders. These whales have been working together for hours, herding herring schools into their nets. As an intern for the Alaska Whale Foundation (AWF) I spent many days floating in a small zodiac, observing this unique predation behavior.

Whales bubble-netting in Frederick Sound.

My path to Alaska and to marine mammal science was anything but straightforward. As an undergraduate, I studied Environmental Studies and Sustainable Design - a path to which I felt ready to commit. However, during my last semester, I had the life changing experience of studying in the Galapagos Islands. There, I was captivated by the tropical ecosystem full of manta rays, marine iguanas, and sea turtles almost twice my size. As well, I was in awe of the marine scientists I met whose passion for understanding and protecting the Islands’ ecosystems was infectious. Inspired by their passion, I decided to begin a new journey to become a marine scientist. Upon returning home from the Galapagos Islands, I declined my acceptance into Architecture graduate programs and began an internship that had me swimming alongside whale sharks in Baja California. I then worked in a small lab in the San Juan Islands, where I dissected herring under a microscope to identify their stomach contents. But eventually, I found myself in Alaska, where I spent two incredible summers as an AWF intern.

AWF interns on a hike in Warm Springs Bay.

AWF’s summer internship is no “whale watch”. We work hard and are often on the water for 10-12 hours a day collecting as much data as possible. We deploy a hydrophone (an underwater microphone) to record the vocalizations and “chatter” of the bubble-netters. We take detailed notes, calculating to the second how long the whales are submerged before lunging. We collect photographs of every fluke and dorsal fin we can in order to get photo-identification images - photos of the whales’ unique tails that allow us to identify individuals. We launch Unoccupied Aerial Systems (UAS, or ‘drones’), which allow us to collect information on whale body condition and size.

AWF team works to photograph every whale we see, in order to create a database of whale sightings over time.

Drone imagery allows researchers to estimate whale size and body condition, providing the AWF team with important information about whale health.

This dance of the field team - typically two or three graduate students and as many interns - is nearly as coordinated and efficient as the humpbacks we observe. Aboard our bright red 20-foot zodiac boat, our team works collaboratively to stay safe and efficient on the cold Alaskan ocean. Our boat driver keeps the zodiac at a safe distance from the whales, allowing the team to collect data without disturbing their behavior.

AWF has been monitoring Southeast Alaska humpback whales for nearly 30 years. The datasets we collect help us understand the habitat use and foraging needs of the whales in an environment that is increasingly impacted by anthropogenic pressures, such as vessel traffic and noise, and climate change. There are many research questions to ask and to aspire to answer, but the ones that interest me most are those surrounding cooperative bubble-net feeding.

Only a small subset of the Southeast Alaskan humpback population has been observed to engage in cooperative bubble-netting. How are these whales learning this behavior? If they are teaching it to each other, how do they choose to whom to teach it? Is bubble-netting a more efficient means of gaining energy than other forms of feeding? Do individuals time their return migration from their sub-tropical breeding grounds to meet up with other bubble-netters? There is a long road ahead of data collection and analyses to answer many of these questions, but one approach we are taking is to look to the past.

The AWF field office in Baranof Warm Springs, Alaska, where interns spend their time when not on the water.

Between 1991 and 2012, Dr. Fred Sharpe, Alaska Whale Foundation’s first scientist, recorded ~200 hours of video footage of humpbacks foraging in Southeast Alaska. More than three decades later, Annie Bartlett, a fellow AWF intern, and I, together with graduate student Dana Bloch and a large team of incredible volunteers took on the massive - yet satisfying - task of extracting as much information on individual whales from these videos as possible.

It was a year-long process to turn these archived videos and thousands of photographs into a humpback whale sighting history dataset. To do so, Annie and I trained 72 remote volunteers via Zoom to extract photo-identification images from the videos. Historically, researchers would manually compare these images to pre-existing “photo-ID catalogs”, noting the flukes’ black and white pigment patterns, scarring, and shape, to identify individuals. This was a painstaking - and very time intensive - process. Today, AWF has partnered with happywhale.com to use its automated online fluke matching system to accomplish the same task, not in hours, but in seconds. As a result, our volunteers could run their images through happywhale.com and almost instantly identify the whales in those images. They would then pass the identification results to Annie and I, and we compiled the data into a single, large dataset. All told, our team was able to create a rich dataset of 3,351 sightings of 558 individual whales observed over 22 years. As I looked through these data, one of the things that struck me most was that so many of the bubble-netters I observed during my tenure in 2021 and 2022 had been working together for decades. It is rare to find animals - humans notwithstanding - that form such enduring relationships.

Each whale fluke is unique, and by photographing whales as they dive, AWF helps to create a database of whale sightings throughout time.

This multi-decadal whale sighting history dataset is now allowing AWF researchers and collaborators to gain new insight into the social biology of bubble-netting humpback whales. And with new insight, come new questions: Is this group of bubble-netters growing? Or, as herring populations decline, is it shrinking? Do these groups form around kin? Is it only older animals that participate? Do males and females form tight bonds, or do they tend to avoid one another?

My time with AWF was a great introduction to critical methods for marine mammal research and the awe I felt in the field was heightened by participating in the revival of archived data. Since leaving Alaska, I have decided to begin my own graduate work. This fall, I will be attending the Marine Mammal Science MSc program at the University of St Andrews in Scotland. There, I hope to study socially-learned behaviors among marine mammals as a means to understanding habit use and potential conservation needs. I am grateful to AWF for introducing me to these areas of inquiry and helping me form the foundation of this new career.

The view from above of Warm Springs Bay.