When state biologists want to know how many moose are wandering around in different parts of Alaska, they usually get into small planes, take to the air and count the animals that stand out in the snow-covered landscape below.
Now climate change is threatening that practice.
A study led by University of Alaska Fairbanks scientists found that the time window for counting moose in the fall is being squeezed by later arrival of adequate snow. By midcentury, it said, there may be too little snow to continue to do these traditional fall aerial surveys in more than half of Alaska’s moose habitat.
The idea for the study was sparked by biologists’ complaints about surveys becoming more difficult to complete, said lead author Todd Brinkmann of UAF’s Institute of Arctic Biology.
“Totally, that’s the motivation,” Brinkmann said. “Being able to complete these in the fall is becoming more unpredictable.”
For now, the fall surveys remain “pretty effective,” but there is already some impact that is akin to “noise in the system,” he said. “If these patterns continue, we’ve got to start thinking about what we’re going to do.”
Statewide, Alaska’s snow season has shrunk, with snow arriving about a week later in the fall than in the 1990s and disappearing about a week earlier, on average, according to a 2019 report by the Alaska Center for Climate Assessment and Policy at UAF.
But mere arrival of snow is not enough to support a successful moose survey.
Snow cover must be adequate, which is at least 15 centimeters, or about 6 inches, Brinkmann said. Less snow than that may make the landscape look white, but not dependably so, and plants sticking up can make the ground look mottled, he said.
“It’s not just snow onset. It’s good enough snow to cover some of that ground vegetation so that those moose stick out. Because they’re brown, of course,” he said. “You need snow accumulation to cover up some of that vegetation, so that brown moose really jump out when these folks are in the plane at 500 feet or whatever it is.”
Also needed for surveying is adequate daylight, which gets scarce at Alaska’s high-north latitude and which makes midwinter moose-counting flights impossible.
If aerial surveying is delayed until daylight returns in the late winter or spring, biologists will not be able to spot from the air the difference between male and female moose, removing information needed to determine population sex ratios, Brinkmann said.. That is because the males, known as bulls, lose their antlers in the winter after growing them over late summer and fall.
Additionally, there are potential mismatch problems if surveys are delayed until spring, he said. Moose may have shifted locations since the start of winter, and it could be unclear what animals are being seen from the air, he said. “Are those the same moose, or does the distribution change?” he said.
Another problem with changing when surveys are done is that the state would lose consistency in its recordkeeping. Long-term data is from fall surveys, so the potential seasonal differences in moose distributions could mean a data interruption if biologists have to switch to spring surveys.
The concern goes beyond science and academia, Brinkmann said. The changes could affect state managers’ decisions and, ultimately, moose-hunting opportunities for the public, he said.
“If you don’t have really good data, you’ve got to be cautious to avoid overharvest. And the goal usually is to try to give hunters as many opportunities as they can so they can build their freezers and, and their families can consume this really renewable healthy protein source,” he said.
On the other hand, if hunting levels are set too low, there may be too many moose on the landscape for the habitat to support, he said.
Brinkmann’s study, which was co-authored by a UAF colleague and scientists at Colorado State University and Columbia University’s Lamont Doherty Earth Observatory, combined past records with climate projections through the middle of the century.
It examined snow and survey records at seven areas within different Alaska game management units, the regions where specific wildlife regulations and hunting limits are applied. The records were from surveys done between 1987 and 2019 in seven subunits, which are portions of management units. In that time period, 170 surveys were completed and 41 canceled. The average start date was Nov. 12, but there was a lot of variation by region. Late arrival of adequate snow cover corresponded to survey cancellation, the study found.
Snow records reveal a later onset of adequate snow cover in five of the seven game management subunits examined, with delays of as much as 14 days between 2005 and 2020. The most dramatic delay was in Game Management Subunit 14A, which encompasses the southern part of the Matanuska-Susitna Borough. In contrast, there was little change in fall snow cover over the period in Game Management Subunit 15A, which is on the Kenai Peninsula and was the southernmost of the areas in the study, the results found.
Continuation of those trends would mean too little fall snow for aerial surveys in most moose habitat within three or four decades, the study found.
In Southcentral and Western Alaska, that threshold is expected as soon as 10 years from now, according to the projections.
A time crunch has already emerged to cause problems in some places, said one of the biologists who goes airborne each fall to assess moose numbers.
Lincoln Parrett, the Fairbanks-based regional supervisor for the Alaska Department of Fish and Game’s Division of Wildlife Conservation – and a pilot himself – said fall snow is most reliable in Interior Alaska, where he works, and less reliable in areas closer to the coast.
But even in the Interior, where snow cover used to start reliably in October, there have been some challenges.
“As it turns out, what we found is that we’re just getting pinched, right?” Parrett said.
He was speaking on Nov. 21, the day after this group started one of the area surveys, which is relatively late in the year.
“And now we’re in a little bit of a race to get it done before a couple of things happen — before the bulls start to drop their antlers and before we run out of daylight. And so we just get pinched on that end,” he said.
The state’s fall surveys use a method called GeoSpatial Population Estimator, which samples boxes that are usually around 5 or 6 square miles, Parrett said. It generally takes about 45 minutes to survey each of those boxes, depending on terrain. The system allows for a margin of error, as it is nearly impossible to count all moose, he said.
Parrett, like the numerous other pilots working on the fall moose surveys, flies with an observer as a partner sitting behind him. Both look out the plane windows on either side to count the moose below.
Parrett normally flies about 600 feet above the ground, making patterns that vary with the terrain below. Over flat landscapes, his flight patterns are in straight lines; over hilly terrain and curvy mountains, the flight paths are contoured accordingly. Over forested and heavily vegetated areas, where moose are harder to spot, he has to make more and tighter flight passes; over open tundra, where sight lines are clear, he makes fewer passes.
Other necessities for successful fall surveys, beyond adequate snow cover, include safe flying conditions and an adequate number of pilot-observer teams and planes.
Uncertain fall snow cover is not new in some places of Alaska, Parrett said.
In some parts of Western Alaska, for example, biologists have been relying on spring surveys rather than fall surveys for about 20 years, despite the drawbacks of those spring counts, he said.
Spring surveys have also been used in part of Interior Alaska as well, where the climate is very dry and snow can be scarce, he said.
In the future, as later-arriving snow makes the old-school fly-and-count method less feasible, technology could fill in the gaps, Parrett said.
Biologists at the Department of Fish and Game are looking at the option of tracking moose populations through genetics, using a method called close-kin mark-recapture, he said. That method uses extrapolations to estimate population sizes by tracking the genetics that link kin relationships between animals. The method has been used with other species like bearded seals in the Bering, Chukchi and Beaufort seas off Alaska and Arctic grayling fish in Canada’s Yukon.
Moose surveying in the future could also incorporate the use of drones or infrared technology that can track the animals’ presence through their body heat, Parrett said.
For now, though, he and his colleague are still using the traditional method, as long as daylight and weather holds out.
While he was grounded by fog on Nov. 21, he was expecting better moose-spotting conditions in the days to come.
“Right now, the air is very moist, and so the one thing that it’s doing, I can see it happening before my eyes, is it’s frosting. So all the trees and all the willows are going to be covered with frost, which will produce excellent conditions,” he said.
Such conditions allow biologists to use the time-honored — and very low-tech — practice of seeking out moose by following the marks they leave in the stands of trees and bushes.
“That frost is amazing, because the moose knock the frost off as they walk around and they make these incredible trails,” he said.
