Decoding Killer Whale Communication From Above and Below
We are excited to share a new pilot project with Raincoast Conservation Foundation and Earth Species Project (ESP) to study the communication and coordination behaviors of killer whales.
Killer whales live in tight-knit matrilines, pass knowledge down through generations, and use complex vocal repertoires to hunt, share food, and stay together. These actions require coordinated, precise, and context-specific communication. And while we know that killer whales have rich vocal cultures, we still don’t fully understand what their calls mean, especially in the context of social behavior or ecological change.
This partnership brings together Raincoast’s decades of field expertise and individual whale identification with ESP’s frontier AI tools, designed to analyze more-than-human communication. By combining aerial drone footage and underwater hydrophone recordings, the team aims to build a synchronized, high-quality dataset linking killer whale vocalizations to specific behaviors, individuals, and environmental contexts—including underwater noise.
This research will help us build the training data needed to develop AI tools that can annotate and decode decades of existing killer whale bioacoustics data from research efforts around the world. In doing so, we hope to answer some big questions: How do killer whales coordinate their actions across distances using sound? How are their vocal behaviors shaped—or disrupted—by rising noise levels in the ocean? And how can we better protect the acoustic lifelines their survival depends on?
During a recent test of the equipment for the pilot, we sat down with Dr. Valeria Vergara and Dr. Lance Barrett-Lennard, co-directors of Raincoast’s Cetacean Conservation Research Program, on RCF's research vessel Achiever to learn more about the science behind the project, what makes orca communication so compelling, and what this partnership might unlock.
Q: How did each of you come to work in this field?
Valeria: I’ve always been fascinated by intelligent, social mammals—how they communicate and use sound to navigate their complex social lives. I began my career studying red foxes and other canids, then spent two decades researching beluga whale communication.
Lance and I have worked together for over 20 years, and when we both joined Raincoast, I became more involved in his long-standing killer whale projects. These animals have incredibly rich social structures and vocal traditions that make them especially compelling to study.
Lance: I’ve been interested in both genetic evolution and in the evolution of culture in animals since I was an undergraduate student. When I learned at that time that killer whales have distinct, overlapping populations that behave differently and have different food preferences, I was hooked.
I’ve spent most of my career studying the species–its social structures, genetics, mating systems, and acoustic behaviour—and how these things are impacted by humans. What continues to inspire me is how much we still don’t know—and how collaboration across disciplines can unlock new insights.
Q: You’ve both spent decades studying whales. What makes this particular species so interesting?
Valeria: Cetaceans are incredibly acoustic species. And orcas in particular have a complex and rich communication system. There's a lot that we know about them already. For example, we know that they have dialects that are used to identify each particular pod and even matrilines, similar to regional accents. But we don't know what the different calls in their repertoire mean.
Lance: Most mammals can produce relatively few types of sounds and are poor at vocal imitation—humans and some of the toothed cetaceans being the rare exceptions. The communication system of killer whales in particular is rich and complex, and it clearly plays a central role in how they live. But we’ve only scratched the surface in understanding how these sounds function in their social lives and decision-making.
Q: Tell us more about the pilot project.
Valeria: This summer marks the start of a two-week pilot field season aboard our research vessel, Achiever, in Johnstone Strait. This will be the first year of a collaborative project between Raincoast Conservation Foundation and Earth Species Project, where we will attempt to understand killer whale communication dynamics from above and below.
To say more, we’re synchronizing drone footage, hydrophone recordings, and behavioral observations from the field. The precise time-alignment of these different data streams is what allows us to start linking specific calls to specific actions or interactions.
Lance: We're especially interested in how killer whale communication is affected by underwater noise. Boat traffic interferes with their ability to echolocate and may also mask these social sounds. If we don’t understand what these vocalizations mean, we can’t truly understand what’s being disrupted.
.jpg)
Q: What are some of the big research questions you're hoping to answer?
Lance: We know that underwater noise interferes with killer whale echolocation – their ability to use biosonar to find food. It must also interfere with their communication as well. But we don't really know the consequences. For example, are their calls important for coordinating fish hunts? Are they important for coordinating the way that the killer whales share their prey, which they do often? Do they change the calls they use in response to high levels of underwater noise?
Valeria: We're trying to answer pretty fundamental questions about how killer whales use sound and make decisions. Are there particular sounds that we hear consistently in the recordings when, for example, different matrilines join or split, or when they change directions? Are there particular short distance sounds that are more subtle and softer that the whales might use in very intimate social interactions, like those between mothers and calves?
Q: And where does AI come in?
.jpg)
Valeria: Working with ESP is really neat because machine learning gives us the ability to decode these relationships, to decode syntax, and to work with massive amounts of data in a way that we haven't been able to do before.
We’re going to be out there collecting terabytes and terabytes of sound and video. Machine learning gives us the opportunity to work with these very complex, very large data sets and discover patterns that we would not be able to discover by hand.
The couple of decades that I spent studying beluga whales, we did this work by hand. We didn't have the AI or the machine learning abilities that we have currently, and it was painstaking and slow. So I'm very excited about the potential.
Lance: In a study like this, you quickly accumulate huge amounts of video data, huge amounts of acoustic data. Nowadays, we have such good equipment that we record these things in high fidelity, but that takes up a lot of computer space. So what Earth Species Project brings to this initiative is the ability to find patterns in those data, to analyze those data, to help us decode this relationship between function and form, if you like, of acoustic calls.
AI also allows us to look back. With enough ground-truthed data—where we know exactly which whale made which sound and what they were doing—we can build AI systems that annotate historic acoustic recordings automatically. That’s incredibly exciting because there are decades of bioacoustic data from around the world that may have answers we’ve never been able to look for.
So we’re not just collecting data for this summer—we’re unlocking years of data that already exist.
Q: Why is this work urgent from a conservation perspective?
Lance: We don't really know what the consequences of underwater noise might be for masking the sounds of killer whales, but we do know they use sound to maintain social bonds. So we're trying to better understand the functional significance of their calls and how important they are for maintaining contact with each other. Are they important for coordinating fish hunts? Are they important for coordinating the way that the whales may share their prey, which they do very often?
Valeria: Healthy killer whale populations are an indicator of a healthy ecosystem. So while the ocean is becoming noisier every year, it’s increasingly important for us to understand how they use sound. Without that, we can’t assess what’s at stake. By learning about what all of these various sounds mean, we can get a much better understanding of the impacts of noise on the ability of these animals to communicate effectively.
Q: Any final thoughts?
.jpg)
Valeria: Killer whales are strikingly similar to us in some ways—they have strong family bonds, dialects, and the females even go through menopause. A species’ value isn’t determined by how similar they are to humans, but the fact that killer whales have so many traits that are reminiscent of humans makes it a lot easier to not only empathize with them, but to get people to understand them and love them.
Lance: Killer whales and humans are also similar in having large brains and well-developed cognitive abilities. One of the ways we differ is that only humans have hands and can use tools to extensively alter and shape their environments. What do killer whales use their huge brains for? We know that they have well-developed mental maps of their environment and extensive memories of social relationships and interactions and of environmental features. It seems very likely that they have sophisticated ways of sharing at least some of their extensive knowledge among family and group members. The thing that excites me most about this study is that it will help us determine if and how acoustic communication contributes to this essential information transfer.
…
