Scientists find space-for-time substitutions exaggerate urban bird--habitat ecological relationships

While studies have found bird populations are on the decline -- Canada and the United States have lost nearly three billion birds over the last half century -- measuring waning populations isn't simple, particularly in the absence of longitudinal datasets that measure the same variables over time. That is why scientists often use space-for-time substitutions -- a method for predicting future species counts using habit associations measured at one point in time.

Harold Eyster, a climate scientist for the Nature Conservancy and lead author of the study, analyzed datasets of breeding bird species surveyed in the Vancouver, Canada metro area from 1997 to 2020 and land cover change assembled using photos, satellite imagery, and machine learning methods, to test the validity of space-for-time methods. His team found total bird populations declined 26 percent, with the biggest reductions -- up to 80 and 90 percent -- in house sparrows, barn swallows, and starlings. Land cover changes only accounted for part of population shifts, suggesting some ecological models may be overvaluing the strength of habitat-bird relationships.

"It's not consistent," explains Eyster, who led this research while a postdoc at UVM's Gund Institute for Environment. "It might be a signal that maybe we should be thinking about expanding some of the surveys that we are supporting. …We are measuring one thing -- breeding birds -- but birds have to survive the rest of the year too. So, we have to also think about protecting areas and conserving landscapes that are going to help them the rest of the year."

In 1997, the most common bird species in the metro Vancouver area were the European starling, house sparrow, American crow, house finch, and violet-green swallow. By 2020, chickadees and bushtits pushed swallows and sparrows out of the top five, and scientists found large increases in species such as the northern flicker. The swings in population growth and decline across bird species don't align with predicted space-for-time substitutions, meaning scientists may also be overestimating and underestimating the losses of certain bird populations.

Brian Beckage, a professor with joint appointments in UVM's Department of Plant Biology and Computer Science, and a co-author of the study, says the study's results reflect the nature of complex systems. "Whenever you're fitting a model to a dataset at one point in time, there are a lot of other things going on in the environment … and so you're conditioning on a lot of things that you're not really aware you're conditioning on."

For instance, while there may be large-scale regional changes to an environment, there are also local landscape changes that can mitigate -- or exacerbate -- the effects on wildlife. This is where longitudinal datasets can help scientists tease apart the relationships that matter most and understand what levers humans can pull to buoy wildlife populations in a period of global change.

"Longitudinal data allows you to start seeing how these relationships change over time, how variable they are, what some of the underlying factors that drive the relationships," Beckage explains. "You can begin to understand when those variables are important, and how they are important, and when they interact."

The study was funded by Environment and Climate Change Canada to investigate potential nature-based solutions to climate change, such as city planting initiatives, and their effects on bird populations in urban landscapes. Eyster and Backage previously found that conifers may be better at cooling the Vancouver area during heat waves, which benefits both humans and wildlife. But planting trees has myriad health advantages.

"It can improve the quality of the environment for people; it can directly mitigate climate change because trees absorb more carbon; it can mitigate local effects of climate change like heat because it provides more shade," Beckage explains. "It can also provide more habitat for birds. It also gives people agency to make the world better in all these different ways by doing something that's not that hard to do."

While many of the bird species in metro Vancouver exhibiting the steepest population losses are considered nonnative to North America, they often live in the densest areas and may represent some of the only interactions people have with wildlife. As more than half of the world's population lives in cities, urban environments provide a critical opportunity for fostering positive human and wildlife interactions.

In 2020, Eyster replicated birding surveys in the Vancouver area from the '60s and '70s conducted outside the breeding season and found that many bird species have more flexible and complex behavior than reflected in breeding bird surveys. These species may be more adaptable to changes in land cover than previously understood.

"A lot of our ideas about birds are based on people going out in the 1900s, or even earlier," Eyster explains. "But … birds are smart. They are flexible. They are able to culturally adapt. Part of what my research is to be really open to that and not assume that birds are going to eat the same things that they did in the '60s when forests were different."

For instance, certain birds, like the Pacific Wren, appear finicky. During the breeding season, they nest in large forest expanses in British Columbia. However, Eyster found the wrens venture into new territory in the fall and winter. "They're moving into people's cedar hedgerows in their yards. They move into these completely different landscapes," he says, "and we've actually seen an increase in how they're using the rest of these cities."

Perhaps they are adapting to a changing landscape, he continues. "… And maybe what we do is not just protect those big parks, but also actually the whole landscape matters."

He argues urban planners should not ignore the importance of these in between spaces in cities. Large expanses of forest and fields may be important for species during the breeding season, but residential backyards and neighborhood shrubbery also provide sources of food such as berries and seeds that are vital at different times of the year.

It's easy to initially point to climate change as the primary culprit of bird population declines. But it's harder to pin down their disappearances to climate change alone. Habitat loss, changes in food supply, or competition from newly introduced species are factors, too.

"One of the things that I am interested in is thinking about ways to disentangle some of those things," Eyster says.

For him, unraveling the complexity involved in species decline means better understanding the web of relationships affecting them -- especially the human ones. And cities may be an ideal place to start.

The study was co-authored by Kai M.A. Chan of the University of British Columbia, Morgan E. Fletcher, an undergraduate at UVM, and Brian Beckage.