Seabirds have long been considered indicator species, useful for gauging the health of a marine ecosystem as a whole. They can be observed more easily than fish, whales, and other creatures living under the surface of the ocean. For most of our history, humans had very few ways to make observations beneath the waves. Seabirds, however, have always been visible to us. Though they live and breed above water, they feed almost exclusively on fish. They form an arm of the marine food web that sticks up above the water, where we can see it with the naked eye.

In recent centuries, technological advancements have made it just as easy to take measurements under the water, observing entire marine ecosystems in situ, but so too have they multiplied the number of ways in which we can study seabirds.

The Evolution of Tools and Methods

The 21st century saw a number of advancements that affected the field of seabird biology. There was a boom in the popularity of solar panels, improving remote camera setups. Drones, or remotely piloted aircraft systems (RPAS), were approved for commercial and private use. Machine learning, particularly neural networks, became mainstream tools. And of course, the quality of images taken on digital cameras continues to improve.

All of these tools came together on the project that became my master’s thesis (and that of three other students). Together we studied breeding success at three mixed Double-crested Cormorant and Pelagic Cormorant nesting colonies, but my project focused more particularly on comparing methods of surveying.

Remote Island Realities: Mitlenatch Island

The three colonies sit on a spectrum of remote to urban environments. Mitlenatch Island is a tiny island in the Salish Sea, only about one kilometer long by half a kilometer wide. It sits in the north of the Salish Sea between Vancouver Island and mainland British Columbia, Canada. Both Vancouver Island and the mainland are covered in rainforest, but Mitlenatch Island is in the rain shadow of Vancouver Island, making it a tiny semi-arid blip in the middle of the sea. It is accessible only by small boat and is a provincial park and nature reserve with only a couple public trails. I have been to Mitlenatch Island six times, but many volunteers in its stewardship trust have visited countless times over many decades.

Where they haven’t been, however, is the Double-crested Cormorant nesting colony atop the cliff at the southwest corner of the island. It is far from the public trails, and we needed a permit from BC Parks to access that part of the reserve. One week after moving to British Columbia, I was clambering up those rocks to install a camera to photograph the cormorants for my master’s degree research.

Many seabirds nest on cliffs because they afford protection from predators. But this cliff was accessible to numerous predators: crows, ravens, gulls, and even otters can get close enough to snatch eggs if the parent is away from the nest. And as we found in the photos on the camera that first year, the cormorant parents all left their nests nearly every day for the same reason: bald eagles. A bald eagle pair nests on Mitlenatch Island, less than 200 m away from the cormorant colony. Every time either of them overfly the colony, or any of their young that are still around (juvenile eagles may stay with their parents for a few years while they mature), every last cormorant flushes from the colony.

Years of Discovery and a Breakthrough

In 2021, the first year we set up the camera at the Double-crested Cormorant colony, we observed that not a single chick hatched. We didn’t get any photos of eggs either, but we assumed that some had been laid given the normal courtship and incubation posture we saw the cormorants in. When we returned that fall to collect the camera (not knowing what we would find on it, as it doesn’t transmit the photos), we spotted eggshells a stone’s throw away from the colony. Not somewhere they would have ended up after hatching. We assumed predation of those eggs. Some predation is expected, of course, in any population, but we were astonished to realize not a single egg had survived to hatch.

What happened to them? The camera showed evidence that pressure from the bald eagle family was causing the cormorants to flee from their nests daily, only returning once the threat had passed. It captured numerous photos of an empty colony with eagles seen in the background. In 2023, the camera finally captured proof: a photo of a crow, with an egg in its beak, flying past the camera while all the nests were empty.

Then, in 2024, for the first time since installing the camera, we saw chicks. Something had gone right for the colony this year, but what? The eagles were still around, with their family of four flying around the island. We don’t know what made the difference that year. Our options for observation were limited to our post-mounted camera and the kayak-based surveys done by the stewardship trust. Mitlenatch Island is too remote for frequent visits.

Middle Ground: Gabriola Island

Gabriola Island is much larger at 14 km long by 4 km wide and is less remote—being accessible from the city of Nanaimo by a 25-minute ferry ride. It is also between Vancouver Island and the mainland, but unlike Mitlenatch Island it has permanent human inhabitants, and is bypassed constantly by boat traffic both large and small. The Gabriola Island cormorant colony is on a sheer cliff face facing the city of Nanaimo and a busy shipping channel, but atop the cliff there are only a few quiet dwellings. Here, too, we installed a single static camera to photograph the colony from above. Unlike Mitlenatch Island, however, we had a vantage point 1 km away in Nanaimo from which we could take long-distance photos, too. We took these once or twice per breeding season. The Gabriola Island colony successfully reared chicks throughout my time working on the project, but wasn’t nearly as productive as the third colony site.

Urban Success Story: The IMSN Bridge

The largest colony not only in our study, but in the entire Salish Sea, is found on the mainland at the Ironworkers’ Memorial Second Narrows (IMSN) Bridge in Vancouver. The IMSN Bridge connects the city of Vancouver with the Lower Mainland’s north shore, and is one of only two road bridges that do so. The cormorants nest on the crisscrossing support beams below the road surface and have vehicle traffic passing overhead 24/7. A rail bridge passes parallel to the IMSN Bridge less than 100 meters away, and a shipping channel passes right below. It’s unlikely for the IMSN Bridge cormorant colony to ever have a quiet day. But there’s one thing that doesn’t seem to give them much trouble: eagles.

In all our observations of the bridge colony, 2–4 times per week for five breeding seasons, we have never observed the cormorants flush from their nests. There are certainly bald eagles in the area, with at least one nest within 1 km of the bridge. But they don’t seem to frighten the cormorants on the bridge like they do those on the cliffs.

Why not? The question of why animals do anything is always difficult. It’s possible the horizontal surface overhead makes the cormorants feel safer, or even that it hides the eagles from their view when they are close enough to normally be a concern. Whatever the mechanism might be, it appears the bridge offers some measure of predator protection to the cormorants.

The Power of Access and AI

Aside from apparently being less accessible to eagles, the IMSN Bridge colony was more accessible to us as researchers. We students took it in turn to visit the colony (averaging 2–4 visits per week between the four of us) and photograph it from a ground-level vantage point below the bridge. We used a robotic mount, tripod, and 200x zoom lens to take a mosaic of photos, to be stitched together into a panorama later. We also experimented with the use of a drone, or RPAS, to photograph the nests from above. Because the bridge colony was so accessible, we ended up with more panoramas than we could process. Finally, we developed a convolutional neural network (CNN) to count the cormorants and nests in the panoramas. The CNN can scan a panorama much faster than a human, and after months of tuning, is about as accurate as a human too. With our Nanaimo-based vantage point, it might be possible to someday make a similar model for Gabriola Island, but Mitlenatch Island is too far from any other landmass for it.

A Changing Field with Expanding Possibilities

The three colonies we studied have varying degrees of accessibility and different sets of survey methods. Ultimately, there is no single method that can be used at every colony. At the IMSN Bridge, which is owned by the provincial government, we were unable to get permission to install a stationary camera. We didn’t have the time or resources for frequent visits to Gabriola Island or Mitlenatch Island. Hiring a licensed RPAS pilot was prohibitively expensive to do for the duration of the project. Any seabird colony study will run into limitations on what methods are feasible for them. Even so, we are spoiled for choice in comparison to past decades. New methods can improve the quality of data, the amount and types of data it’s possible to collect, and the speed at which they can be analyzed. With an abundance of data and software tools to process it, we can investigate new and more complex questions with the time we would once have spent on manual entry and calculations. Where our options as seabird biologists were once limited to recording data with pencil and paper from a boat or bird blind, a whole world of new possibilities has opened up.

---

About the Author

Rose Wilkin is a lesbian conservationist based in Vancouver, BC. She monitors the Great Blue Heron colony in Stanley Park and previously studied various cormorant colonies in the Salish Sea. Rose received her Master’s degree in Ecological Restoration from Simon Fraser University in collaboration with the British Columbia Institute of Technology. Her thesis focused on cormorant monitoring methods. Her forthcoming paper based on this work will appear in Ornithological Applications. Seabirds have been her lifelong passion, and she considers it a privilege to study them in her career.