Getting your Trinity Audio player ready...

By Dan Gunderson, Minnesota Public Radio News

The red-headed woodpecker was once commonplace across Minnesota, but the population has dwindled here and across North America.

“Red-headed woodpeckers in Minnesota have declined by about 95 percent over the last 50 to 60 years, probably due to habitat loss and conversion to agriculture and development,” University of Minnesota researcher Elena West said.

Some red-headed woodpeckers will winter in Minnesota and others will migrate to nearby states like Missouri or Illinois, West said.

The birds favor oak savanna habitat, an area of transition between tallgrass prairie and oak forest. They will also inhabit hardwood forest areas across the state.

Red-headed woodpeckers need: dead trees where they can carve out nests, trees that produce nuts they can gather for food and open areas to catch insects in flight.

West has been studying the birds for five years at the Cedar Creek Ecosystem Science Center in East Bethel, a managed habitat with a robust red-headed woodpecker population.

Habitat is critical

But she realized relatively little is known about where they live and what habitats they use across the state.

“The ultimate goal here is to try to inform management efforts that are aimed at restoring oak savanna habitats that appear to be really, really rare in the state,” West said.

The red-headed woodpecker doesn’t just use the habitat — it creates habitat for other species.

“We know that at least 30 other species use the cavities that red-headed woodpeckers create,” she said, “From amphibians to other birds to small mammals.”

There were once wide expanses of oak savanna, primarily across southeastern and central Minnesota, but only fragments remain.

Out in the field

One of those rare spots is at Crane Meadows National Wildlife Refuge near Little Falls, Minn. where field technicians Kate Sibila and Paul Zeitz are placing small audio recording devices on trees in pre-selected locations.

This spot and hundreds of others across the state are picked by scrutinizing satellite images.

“Then we see if it’s reachable,” Sibila said. “If there’s better suited habitat nearby then maybe we can move a point.”

After picking their way through tall grass and downed trees, they find an oak that’s a good size for mounting an audio recorder.

The device, about two inches square, is in a sealed plastic bag that’s zip tied to the tree.

It will capture bird calls within about a 150 meter radius, Zeitz explains, and the recorders are programed to switch on in the morning from 7-11 a.m. when birds are most vocal. They run for three days before being collected and moved to a new location.

West learned during a pilot study two years ago that if the recorder was near a red-headed woodpecker territory, it would record the bird within 24 hours.

So two teams keep busy moving the 115 recorders available for this project to locations around the state.

“We don’t have that many devices, but we’re able to cover a lot of ground because we’re rotating them very, very quickly,” explained West.

The project recorded at only about 250 sites last year because the international computer chip shortage limited the availability of the recording devices.

This year they expect to at least double the data collection.

Artificial intelligence a critical tool

The team is using artificial intelligence to develop an algorithm to identify the red-headed woodpecker calls in hundreds of hours of recordings.

Using machine learning, a computer will be trained to identify a specific call.

“It’s actually called a queeah call,” West said. “It’s a territorial call that red-headed woodpeckers produce when they’re defending a territory. Males do it a lot when they’re setting up a territory and trying to attract a mate.”

A human ear still needs to confirm each woodpecker call flagged by the computer.

Wildlife researchers have been deploying recording devices for years, but West believes advances in machine learning will greatly expand the potential uses.

“We’re sort of on the forefront, I think, of our ability to survey species at much larger spatial and temporal scales. And that’s really why this this technology is exciting,” she said.

This project is an example. Using human observation to conduct a statewide survey of red-headed woodpeckers is not feasible. But acoustic monitoring allows a small team to cover a large geographic area.

The audio recorders also capture every species that vocalizes within range of the device.

“And it’s essentially like a museum specimen,” West said. “You can store it to come back to it later. And maybe you can analyze it for another target species or a community of species to understand how has that changed under land management changes or climate change? So in that sense, I think it’s really powerful to collect data statewide.”

Project funding from the Minnesota Environment and Natural Resources Trust Fund will end next year, but West hopes to continue and expand the data collection. One idea she wants to pursue is using citizen scientists to deploy and collect audio recorders at specific locations all over the state.

The first couple of years of data will give researchers a better idea of where red-headed woodpeckers still live and reproduce in Minnesota, what kind of habitat they need to be successful.