Pigeon Guillemot banding
Scott Pearson, WDFW
Researchers collect data on a Pigeon Guillemot on Protection Island before tagging and releasing the bird.

When we think of wildlife ecologists, we might envision researchers traipsing through meadows, fording rivers, and tracking elusive predators on daring field expeditions. While some of these images may be accurate, those who work in quantitative ecology and conservation know that some of the most groundbreaking and essential ecological research takes place behind the computer screen, using statistics, mapping, and mathematical models.

For the most part, wild animals are fickle, making them inherently tricky to study. So when a researcher returns home after weeks of pursuing a pack of wolves in grueling conditions with only one collared animal, the real work of drawing conclusions and generating knowledge based on limited data begins.

Ecological data are messy

Over the years, Sarah Converse — Unit Leader of the USGS Washington Cooperative Fish and Wildlife Research Unit and Associate Professor in the School of Environmental and Forest Sciences and the School of Aquatic and Fisheries Sciences — has studied myriad wildlife species, from brown tree snakes and gray wolves to beluga whales and pigeon guillemots.

“Ecological data tend to be messy and noisy,” she said, which is one of the main reasons quantitative methods come in handy. “It’d be amazing if we had perfect data on our study species, but we almost never have that. All ecology requires some kind of quantitative methods of various sophistication.”

The models Converse produces enable her to make predictions about a given population based on those incomplete, messy data, which she can then share with wildlife managers and decision makers so they can move forward with the best available information.

Pigeon Guillemot nest
Kevin Cole
Pigeon Guillemots often nest in burrows on steep cliff faces.

How do you count nests you can’t see?

Converse and her lab members frequently develop new versions of mark-recapture models. Mark-recapture models account for the fact that ecologists can almost never get regular observations of individual animals over time. Sometimes when ecologists visit an area, the animals aren’t there, or they’re around but hiding. In fact, most animals actively avoid humans, making them challenging to study.

Development of novel modeling methods is exemplified by their research on Pigeon Guillemots, petite seabirds with dark plumage and bright scarlet feet, which are found throughout the Salish Sea. Despite the fact that Pigeon Guillemots are an indicator species in the region, there has been little demographic research done on the Salish Sea population. Part of the problem is that they nest primarily in steep cliff faces, hidden from predators and prying human eyes.

To get around this challenge, Converse’s lab collaborates with community science volunteers from the Salish Sea Guillemot Research Network who have been visiting the beaches below colonies since 2008. Observers record when a bird flies into a nesting burrow and whether it carries a fish. Using that information, Converse and her former PhD student, Amanda Warlick, along with collaborators, were able to develop a new type of mark-recapture model to estimate the number of chicks produced in a given colony.

Now, Converse and her student Liam Pendleton are starting a new project with Pigeon Guillemots on Protection Island in the eastern Strait of Juan de Fuca, where they are marking and recapturing adults in addition to following nests. By combining all the existing information on Pigeon Guillemot populations, Converse hopes to get a better idea of what is happening with this ecologically important species.

Gray wolf
Milo Weiler
The first breeding wolf pack in Washington since the 1930s was identified in 2008. Since then, wolf populations have continued to increase.

When humans meet wolves

Other quantitative ecologists approach ecological problems with a social science lens, while still employing rigorous quantitative methods. In his study of human-wildlife interactions, Alex McInturff — Assistant Leader of the Washington Cooperative Fish and Wildlife Research Unit and Assistant Professor in the School of Environmental and Forest Sciences — combines qualitative methods with statistics and mathematics to investigate patterns of thinking and behaving in human populations. Rather than building models that predict what a population of animals might do, McInturff builds models to explain psychological phenomena, such as the ways people balance the benefits and risks of a given hazard.

This fall, McInturff’s team will begin researching community perceptions and tolerance of wolves in eastern Washington — in other words, how well humans and wolves will get along. As wolf populations have started returning to the state, researchers have been modeling habitat suitability and population dynamics to provide insight on recovery efforts.

McInturff predicts that including human tolerance levels of wolves will enhance our understanding of how humans and wolves interact. “We are bringing our understanding of wolf ecology into conversation with people’s willingness to tolerate them,” he said. “We often study those things a little bit independently, so putting them into dialogue is really exciting.”

Quantitative ecologists use models and statistics in many different ways, giving us important insights into questions of behavior and conservation. As the field and the technology that powers it evolves, it will continue to push the frontiers of our understanding, and open up new possibilities for protecting critical ecosystems.


Story by Amelia Wells