Building a Better Brood Survey

Delta Waterfowl researchers work to improve methods to estimate duck production

Clutching a long-handled fishing net, Grant Rhodes strode forward, lunging through a thick mat of knee-high alfalfa toward an impossibly invisible orange marker stake that only a highly trained eye could ever spot on a green hillside sea of Saskatchewan grass. After five bounds, the Louisiana State University graduate student abruptly stopped his advance and slammed the net’s frame to the ground, trapping his intended quarry.

“I got her,” he affirmed, raising the net while carefully grabbing the wriggling body of a nesting hen blue-winged teal to separate her from the nylon mesh.

Carrying the duck firmly but gently against his side, Rhodes walked swiftly toward his parked all-terrain vehicle, where fellow graduate student Hannah Sabatier was earnestly preparing a makeshift operating table on the rear rack of the ATV.

Radio transmitters allow Delta researchers to track marked hens and their broods, while yellow nasal saddles identify individual hens.

Sabatier took the hen, tucking the bird’s head under a wing to calm it down. Rhodes applied a topical gel to the back of the duck between the wings, parting the feathers where he’d soon make a tiny hole to insert a radio transmitter. Then Rhodes carefully administered two lidocaine injections to numb the area. While waiting for the lidocaine to take effect, he attached a standard metal leg band and fitted the hen with a nasal saddle.

Rhodes pulled a custom-molded, PVC nasal saddle from a bag and slipped it over the duck’s bill. The bright yellow saddle—which is adorned with a solid black triangle and capital letter V on each side and the top—is held in place by an aluminum post that slides through the nares (nostrils) in the duck’s bill. The triangle signifies which study block the duck is nesting in, while the letter is a unique identifier to that specific hen. To complete the job, Rhodes crimped the ends of the post that holds the nasal saddle in place.

The device won’t inhibit the teal from feeding or carrying out any life functions. Even so, the marker won’t stay on the duck for long. It’s designed to fall off after six to eight weeks of wear.

“When she’s going through wetlands and dabbling and pecking at things, the aluminum rod holding the nasal saddle in place will deteriorate and it will eventually fall off,” Rhodes explained.

Sabatier repositioned the duck breast-down on the surgery table, and Rhodes poked the duck’s skin to insert a tiny, 4-gram radio transmitter with a wispy 9-inch antenna. He worked the radio’s metal prongs just below the skin of the hen. With a needle and surgical thread, Rhodes stitched the rear of the transmitter to the duck’s skin. The entire operation—from putting on a leg band to sewing the final suture—took about six minutes.

“You’re a good mama,” Sabatier said as she set the hen in the grass. After the duck regained her bearings for a few seconds, she launched skyward, then flew low over the Saskatchewan prairie, soon disappearing into the safety of a pothole.

Estimating Duck Production

An antenna-equipped Delta drone detects radio-tagged hens during its flight, then a second, camera-equipped drone moves in to “lay eyes” on the brood.

Rhodes is working toward a Ph.D. under adviser Kevin Ringleman, associate professor of waterfowl ecology and management at LSU. Rhodes’ research focuses on developing a more accurate, less invasive method to estimate duck production. In other words, he’s trying to design a better brood survey.

Waterfowl managers have long conducted brood surveys through visual observation, using an aircraft, boat, or truck to approach wetlands and simply look for ducks. Brood surveys are imprecise at best, mostly because ducks—especially protective mother hens and ducklings—instinctually evade detection from human eyes by hiding in the vegetation along wetland edges.

However, advances in technology—specifically in aerial drones and thermal-imaging cameras—have opened new possibilities to survey duck broods with much greater accuracy, without even disturbing or alerting the birds. In addition, using drones and cameras requires less people and time than the traditional beat-the-bushes method.

“The end goal of this research is to find less invasive, more accurate ways of measuring broods,” Rhodes explained. “In a perfect world, Delta can go out there with just the thermal and optical camera drones, and be able to say, ‘We counted this many broods and our detection probability was this, and so we’re way more comfortable with the survey number now based on previous research.’ It’ll only get better as the technology improves. We’re doing all of this research to obtain a better, more efficient survey design so we can get a more accurate number of how many ducklings are being fledged (reaching flight).”

To develop such a survey, Rhodes and his crew need to determine how many ducklings survive. They began by dragging chains behind ATVs to find nests in the grass on four study blocks in Saskatchewan. Once located, the nests are monitored periodically until the eggs are nearing hatch. Technicians document the progress of each nest, checking the stage of the eggs by “candling” them. In blue-winged teal, incubation is generally about 24 days. Bluewings are later nesters, so many clutches hatch at the end of June or early July.

Ideally, Rhodes wants to capture the hen bluewing on her nest a day before her ducklings hatch. Predation of duck nests is high, so marking the hen late in incubation provides the best chance of a successful brood to monitor. Also, the hen is much less likely to abandon a nest when it is within a day or less of hatching.

During the 2023 nesting season, Rhodes and Sabatier caught and marked 46 hen blue-winged teal. Each was fitted with a nasal saddle and implanted with a radio transmitter. Of those 46 hens, 27 successfully hatched a nest of at least one duckling.

Nest predation rates are high, so Delta marks hens as late in the incubation process as possible to boost brood monitoring successes.

“In the past, a lot of duckling survival studies would mark individual ducklings right when they hatch on the nest,” Rhodes said. “They would capture the hen and mark all of the ducklings to track them. That gives you a lot of good information about what causes duckling mortality, but that can also affect the ducklings. Our goal was to capture the hens and be able to track them well enough so we wouldn’t need to mark the individual ducklings to track them.”

Sabatier, a master’s degree candidate at the University of Wisconsin—Stevens Point where she’s advised by Dr. Ben Sedinger, is studying nest success and brood survival.

Playing Hide-n-Seek

Delta Waterfowl has been using drones and cameras to find duck nests since 2017, when graduate students Roald Stander and Jacob Bushaw piloted their use over the prairie nesting grounds.

In its earliest stage, Delta experimented with drone-mounted thermal-imaging cameras to detect heat signatures of nesting ducks and their eggs.

Through his master’s work, Bushaw advanced Delta’s use of drones to include conducting spring pair counts, nest searches, and brood surveys in Manitoba. During the same time, drones and camera technology continued to improve.

So now, armed with a hangar of souped-up drones, sharper cameras, and state-of-the-art software programs, Rhodes is taking on the latest frontier in waterfowl survey work.

“This first season was a pilot study to determine proof of concept and see if these things worked,” Rhodes says. “We were trying to see if we could track broods using the drone technology. Starting off, we had to learn on the fly how to operate the drone to find ducks.”

During a typical day during the height of hatch, Rhodes and Sabatier catch nesting hens and perform surgeries in the morning. By late morning, Rhodes breaks out his fleet of drones, then climbs inside a pickup truck to fly each drone and stare intently at his computer monitors.

The operation is much like an interactive video game.

First, Rhodes sends up a drone carrying four antennae. Flying over a block of grass about 200 feet up, the drone detects the VHF signals of any ducks implanted with radios. The flight path of the drone displays on a monitor, along with directional indicators of where any radios are located. From that, Rhodes has a solid idea where to search for any marked hens.

Rhodes returns the antenna drone to his launch pad, then sends up a different drone that’s equipped with a thermal-imaging camera and an optical camera. Now that he knows which wetland to search, he flies the second drone, toggling between views of the thermal camera and regular view.

“There she is!” he declares, excitedly zooming the camera view tightly to spot the hen marked with “triangle J” easily visible on the screen. Rhodes quickly flips back to the thermal-camera view, and the heat outlines of nine ducklings loafing with mama appear.

“We use the first drone as a detection cue, then actually find her hiding in the vegetation and then get a count of her ducklings using the thermal and optical cameras,” he explains. “It took a bit at first to get a system down, but toward the end of the field season, we could just do a quick loop around all of the wetlands with the radio telemetry drone to locate broods.”

Even using today’s modern technology, tracking broods is often difficult, Rhodes says.

“We learned a lot about brood behavior,” he says. “There were times where I knew exactly where a hen was with her brood—she was on a wetland with no other wetland around—but I could not get a visual ID of her. They are so well hidden in those cattails. They’re in an unobservable state. That, in turn, helps us learn about our brood abundance surveys when trying to increase detection of these broods. We hypothesize that movement, time of day, and surveyor experience all plays a part, but there are just some times when you’re not going to find them.”

Rhodes tries to track each brood for 30 to 35 days, and he’s been able to follow a couple broods for 40 days.

“We don’t know a whole lot about blue-winged teal broods,” he says. “We did see instances of brood adoption and, with the drones, we were able to confirm that by capturing video and photographic proof. Having individual marked hens is giving us two-fold information: We’re getting estimates of duckling survival, but we’re also learning about brood abundance survey designs. We have this known population of marked birds, so we know there are 27 marked birds out there, but the technicians don’t. If they only observe 20 of the 27, we can look at why they missed the other seven.”

Goals for Year 2

Rhodes and his crew will return to the Saskatchewan prairie in April for the second of three field seasons.

“Our first field season was good proof of concept,” Rhodes said. “We know a lot more, and we can hit the ground running this next season. Our goal is to mark 100 blue-winged teal this season.”

Dr. Chris Nicolai, a waterfowl scientist who oversees Delta’s research program, said the research holds strong promise to impact the way wildlife managers survey duck production.

“The biggest thing we learned this year was developing a whole new set of field methods,” Nicolai said. “We had very happy results. We became good at catching and following birds, and the new technology worked great. We did not get the sample sizes we wanted, but that’s where we’ll improve next year. I’m exceedingly happy with our data collection—we just need more of it.”

Even using today’s modern technology, tracking broods is often difficult.

For 2024, the crew will expand nest searching to ditches and anywhere there’s grass/nesting cover.

Drones are proving to be useful wildlife management tools, and Delta researchers keep pushing the technology in new ways.

“The new drone with its higher-quality cameras exceeded our expectations,” Nicolai said. “The ability to detect and read the nasal saddles is blowing everyone’s minds, and then having that drone-based VHF receiver took it even higher. Being able to follow broods three different ways is awesome. We want to develop an unmarked brood survey, but we were able to compare unmarked broods to those marked with nasal saddles and those with VHF radios. If we can assume that VHF broods are very close to truth, then we can see how far short of truth we are with the other methods and can develop correction factors to move forward with unmarked broods, which is way cheaper and less work.

“We can find a specific hen with a nasal saddle and determine, ‘That’s her.’ And if we can’t find her visually just using the nasal saddle, then we turn the VHF on. So, we couldn’t find her, but now we can. What’s she doing and why did we miss her (with visual surveys)? Did she move over a wetland? Did she lose her ducklings? Or was she simply hiding and no one with binoculars could ever detect her?”

Ultimately, Rhodes’ research could hold the key to how waterfowl managers determine duck production and population sizes, which affects management decisions.

“We want to develop a very rapid, accurate, corrected brood survey that’s more efficient and less invasive,” Nicolai said. “That would give us a much better way to enumerate production. To me, a perfect approach would be to do pair counts to set the stage for what’s available, and then later on, do the brood surveys to show what happened, and then relate that to any type of management action under the sun.

“To develop a state-of-the-art brood survey gets us closer to a better prediction of the fall flight. When you only conduct breeding population surveys, you’re missing production that year. It lets us move our population estimate closer to the initiation of harvest (hunting season). Right now, the Breeding Waterfowl Population and Habitat Survey is conducted about five months before hunting season. As managers, we want to be as close to accurate as we can.”

Paul Wait is senior manager of communications for Delta Waterfowl.