Alaska is home to an estimated 175,000 to 200,000 moose, the largest moose population of any state or province in North America. That number, maintained by the Alaska Department of Fish and Game, sounds reassuringly precise, but arriving at it requires stitching together dozens of smaller surveys, each covering a fraction of Alaska’s vast landscape. Because no one can count every moose across 663,000 square miles of tundra, boreal forest, and mountain terrain, the statewide figure is always an informed approximation built from statistical models, aerial flights, and a growing toolkit of newer technologies.
Why Counting Moose Is Harder Than It Sounds
Alaska manages moose across roughly 26 separate Game Management Units, and biologists do not survey every unit every year. Instead, they rotate through areas on a schedule, sampling portions of each unit and using statistical frameworks to estimate totals. The statewide number you see reported is an aggregate of these unit-level estimates, some of which may be a year or two old at any given time. Think of it less like a census and more like a rolling series of snapshots combined into one picture.
The core challenge is straightforward: moose live in dense forest, thick brush, and steep terrain where they are easy to miss from the air. Snow cover helps enormously because dark-bodied moose stand out against white ground, which is why surveys are conducted in winter. Even so, a moose bedded down under a spruce canopy or standing in heavy willows can be invisible to a pilot and observer flying overhead. This means raw counts from a survey always underestimate the true number of moose present, sometimes substantially.
The Geospatial Population Estimator
The workhorse method across much of Alaska and northern Canada is the geospatial population estimator, or GSPE. The idea is to divide a survey area into a grid of small units, fly a sample of those units and count every moose visible within them, then use a statistical technique called finite-population block kriging to predict how many moose are in the units that were not flown. Nearby units tend to have similar moose densities because habitat quality, elevation, and food availability change gradually across the landscape, so predictions for unsampled units borrow strength from their neighbors.1The Journal of Wildlife Management. Bayesian spatial modeling of moose count data: increasing estimator efficiency and exploring ecological hypotheses
Researchers have been refining this approach. A Bayesian version of the spatial model allows biologists to fold in prior information about habitat, such as vegetation type, burn history, or elevation, and can produce tighter confidence intervals around the final estimate. In practical terms, that means managers get a more reliable population figure without necessarily having to fly more survey units, which saves fuel, flight hours, and the wear on small aircraft operating in harsh winter conditions.1The Journal of Wildlife Management. Bayesian spatial modeling of moose count data: increasing estimator efficiency and exploring ecological hypotheses
Correcting for the Moose You Do Not See
Even within the units you fly, you will not spot every animal. That is where sightability models come in. The basic idea, developed in the 1980s, treats each group of moose as having a probability of being detected that depends on measurable conditions: how much canopy is overhead, how many animals are in the group, whether they are moving, and so on. If a model predicts that a single moose under dense cover has only a 50 percent chance of being seen, then every lone moose actually counted in that habitat effectively represents two moose in the final estimate.2Biometrics. Sightability adjustment methods for aerial surveys of wildlife populations
Building a sightability model requires calibration data, typically from radio-collared moose whose locations are known. Observers fly the survey as usual without knowing exactly where the collared animals are, and afterward researchers check which collared moose were spotted and which were missed. One study in west-central Alberta found that sightability on the transect line was only about 67 percent in one year and dropped to 46 percent in another, meaning more than half of the moose directly along the flight path went undetected during the worse year.3Wildlife Society Bulletin. Contrasting aerial moose population estimation methods and evaluating sightability in west‐central Alberta, Canada Alaska’s conditions vary widely by region, but the takeaway is the same: without a sightability correction, raw aerial counts can miss a third or more of the moose that are actually there.
These correction factors ripple through the final population estimate. A small change in assumed sightability can shift a unit’s estimated moose population by hundreds of animals, and when you add those shifts across multiple units, the statewide total can move by thousands. This is one reason the statewide figure is reported as a range rather than a single number, and why biologists treat it as a best estimate rather than an exact headcount.
Newer Ways to Find and Count Moose
Traditional helicopter and fixed-wing surveys remain the backbone of moose monitoring, but several newer tools are being tested to supplement or eventually replace parts of the process.
Thermal infrared cameras mounted on aircraft pick up the body heat of large mammals against the cooler landscape. A study in Colorado evaluated this approach for Shiras moose in Rocky Mountain National Park and found that detection probability for groups on the survey line, and for individuals within observed groups, was high. The method also allowed the aircraft to maintain a greater altitude above the ground than traditional visual surveys, improving both safety and reducing disturbance to the animals.4Wildlife Biology. A high‐altitude thermal infrared method for estimating moose abundance and demography in Rocky Mountain National Park, USA
Drones carrying thermal cameras are being explored in Scandinavia’s boreal forest, which shares many habitat characteristics with interior Alaska. In one evaluation, researchers flew drones over GPS-collared moose to measure detectability under different weather conditions and then conducted grid flights to estimate local density. Moose detectability dropped in sunny conditions, likely because sun-warmed ground and vegetation created thermal clutter that made animals harder to distinguish. Under cloudy skies, detectability exceeded 80 percent. The drone-based density estimate came in at about 2.3 moose per square kilometer, higher than a parallel estimate from traditional pellet counts, suggesting drones may catch animals that other methods miss.5Authorea. Using drone-borne thermal infrared cameras to monitor moose in Scandinavia Drones are still limited by battery life and regulations, making them most practical for small, targeted areas rather than the sweeping landscapes Alaska surveys typically cover.
On the genetic side, researchers are experimenting with DNA collected non-invasively from fecal pellets and even from saliva left on browsed twigs. By genotyping individual animals from these samples and applying capture-recapture statistics, biologists can estimate population size without ever seeing or disturbing a moose. One recent study tested whether salivary DNA from browsed twigs could complement fecal samples for individual identification and population estimation.6European Journal of Wildlife Research. Estimating the population size of a semi-isolated moose (Alces alces) population from two sources of non-invasively collected DNA Fecal pellets themselves can serve as a DNA source for mapping the distribution of individual animals, though the timing of collection matters because DNA degrades quickly in warm or wet conditions.7Journal of Fish and Wildlife Management. Timing Moose Pellet Collections to Increase Genotyping Success of Fecal DNA These genetic methods are labor-intensive and work best for small or semi-isolated populations; they are a long way from replacing aerial surveys across Alaska’s interior, but they fill niches where flying is impractical or where managers need fine-scale data on individual movements and relatedness.
What Pushes the Population Up or Down
The number of moose in any given part of Alaska depends on a balance among food supply, predation, disease, weather, and human harvest. Understanding these drivers matters because they explain why moose can be abundant in one drainage and scarce in the next, even when the habitat looks similar on a map.
Predation is the leading cause of calf mortality in most moose populations. In areas where bears and wolves overlap, calves face pressure from multiple predator species. A study on the Grand Portage Indian Reservation in Minnesota, where black bears and gray wolves are the primary moose-calf predators, found that implementing a spring black bear harvest was associated with a 68 percent reduction in bear predation on calves, with no compensatory increase in wolf kills despite rising wolf density during the study period. In years without the spring hunt, the proportion of calf deaths attributed to bears was roughly five times higher.8The Journal of Wildlife Management. Spring black bear harvest and predation pressure on moose calves in a multi‐predator system Alaska has its own long history of predator-management programs, and the tradeoffs are real: predators can hold moose populations at lower densities, which reduces the number available for harvest but tends to promote larger-antlered individuals valued by guides and the tourism industry.9PubMed Central. Tradeoffs Among Predator Control, Moose Harvests, and Trophy Antlers: Principles Pertinent to Managing Alaska’s Wildlife
Wildfire is another major force shaping moose habitat. Moose thrive on the early-successional browse that sprouts after a burn: willows, birch, and aspen regenerating in the open canopy left by fire. Given the increasing prevalence of wildfires across Alaska’s boreal zone, researchers have been examining how moose harvest patterns change in the immediate aftermath of a fire.10Canadian Journal of Forest Research. Limited immediate effects of wildfire on Alaska moose harvest The short-term effects appear limited, but over a decade or more a large burn can create a pulse of high-quality habitat that temporarily boosts local moose density. Conversely, forests that have not burned in a long time tend to mature past the point of producing good moose browse, and local populations decline.
Climate Change and Winter Ticks
One of the more ominous pressures on moose across their southern and eastern range involves winter ticks, parasites that can infest a single animal by the tens of thousands, draining blood and causing the host to rub off its insulating fur. In eastern Canada, the number of heavily infested moose has been rising, and researchers tie this to milder climatic conditions that improve tick survival. Warmer temperatures earlier in spring and later arrival of hard frosts in autumn extend the window during which ticks can find a host.11The Journal of Wildlife Management. Linking weather conditions and winter tick abundance in moose
Mathematical modeling of the tick-moose relationship suggests that even a modest shift in season length, such as a single-month delay in the first autumn frost, could push some moose populations toward local collapse. The cyclical die-offs already observed in some areas may be an early warning that the population is approaching a tipping point.12Ecological Modelling. A mathematical model of the impacts of climate change on the winter tick epizootic in moose Alaska’s interior is still cold enough that winter ticks have not been a major factor there, but as climate continues to warm, the parasite’s range could expand northward. For Alaska’s moose managers, this is a watch-and-prepare situation rather than a current crisis, but it illustrates how a population estimate is not just a snapshot of the present; it is also a baseline against which future declines or shifts can be measured.
Moose in the City
Not all of Alaska’s moose live in the backcountry. Anchorage, home to roughly 300,000 people, also hosts an estimated 1,500 or more moose that wander through neighborhoods, greenbelts, and along roadways. This coexistence comes with a cost. A study of moose-vehicle collisions in Anchorage during the early 1990s found the collision rate rising significantly over the study period, reaching about 49 incidents per 100,000 residents per year. Of 519 reported collisions, nearly a quarter resulted in injuries to 158 people, though no human fatalities occurred. Most crashes happened after dark, with a concentration between 6:00 p.m. and 2:00 a.m. More than half occurred on slick roads, and nearly a third were clustered in December and January, when darkness dominates Anchorage’s day and moose move along plowed roads to avoid deep snow.13PubMed. Urban encroachment on the wilderness: moose-vehicle collisions in Anchorage, Alaska, 1991-1995
Suggested prevention measures from that era included reducing speed limits around greenbelt corridors, installing brighter streetlights in known moose-crossing zones, building wildlife underpasses, and clearing snow berms that funnel moose onto the road. Some of these have been implemented in the decades since, but the fundamental dynamic persists: moose are attracted to the same valley-floor corridors that humans build roads through, and Anchorage’s expanding footprint keeps pushing the two together.
Moose Have Been in Alaska a Very Long Time
It is sometimes assumed that moose expanded into northern Alaska only recently, following boreal forest that crept northward after the last ice age. Radiocarbon dating of moose bones from the Ikpikpuk and Titaluk River drainages on Alaska’s North Slope complicates that story. The oldest bone from those sites dates to roughly 14,100 years ago, and the collection spans millennia, showing that moose inhabited the region at least intermittently through major climate fluctuations, including the Little Ice Age. Of 34 dated bones and antlers, only one was definitively modern, with a handful of others possibly from the last few centuries. There is no evidence the bones were transported from farther south.14Canadian Journal of Zoology. Prehistoric perspectives can help interpret the present: 14 000 years of moose (Alces alces) in the Western Arctic
This long archaeological presence matters for management because it reframes the baseline. If moose have used Arctic watersheds for thousands of years, their current presence on the North Slope is not an anomalous northward expansion driven by recent warming; it is a continuation of deep-time use. Managers setting population objectives for northern Alaska can look to this record as evidence that moose belong in the landscape, and that fluctuations in their presence are normal rather than alarming.
How Harvest Quotas Connect to Population Estimates
Alaska issues moose-hunting permits through a system of registration, drawing, and general-season hunts that vary by Game Management Unit. The number of permits issued and the structure of the season, whether it targets bulls only, allows antlerless harvest, or restricts antler size, all hinge on the population estimate and the observed trend in each unit. When surveys suggest a population is growing, managers may open antlerless hunts to slow growth and prevent habitat degradation. When surveys show a decline, they tighten restrictions.
Moose are deeply important to rural Alaska. Subsistence harvest provides a significant source of calories and cultural continuity for Indigenous and rural communities, and the state constitution requires that subsistence use be given priority when wildlife populations cannot support all uses. This creates a direct link between the accuracy of population estimates and the food security of remote villages: if an estimate is too high, overharvest can follow; if too low, unnecessary restrictions can cut off a critical food source during a short hunting season. The stakes of getting the count right extend well beyond wildlife biology.
Commercial guiding and wildlife tourism also depend on healthy moose numbers. Guides operating in areas where predator pressure keeps moose density low may benefit from larger-antlered bulls, but their clients still need a reasonable chance of encountering an animal.9PubMed Central. Tradeoffs Among Predator Control, Moose Harvests, and Trophy Antlers: Principles Pertinent to Managing Alaska’s Wildlife The economic and social pressures on moose management are substantial, and they all funnel back to the same question: how many moose are actually out there?
The Uncertainty You Should Expect
If you see a headline stating that Alaska has exactly 200,000 moose, treat that as a rounded midpoint of a range with real uncertainty on both sides. Individual unit estimates carry confidence intervals that can be wide, especially in heavily forested regions where sightability is poor or in units that have not been surveyed recently. Rolling those unit-level uncertainties up into a statewide figure compounds the imprecision.
That said, the trends are often more reliable than the point estimates. Even if the exact population in a given unit is uncertain, repeated surveys using the same methods can show whether the population is growing, stable, or declining. Managers often care more about the direction and rate of change than about the absolute number, because those trends dictate whether harvest regulations need to be adjusted. A unit might have “somewhere between 3,000 and 4,500 moose,” but if three consecutive surveys show the estimate climbing, that is actionable information regardless of whether the true number is 3,200 or 4,100.
The push toward Bayesian spatial models and newer detection technologies is partly about narrowing those confidence intervals so that managers can make finer-grained decisions. But the fundamental reality of counting large mammals across a landscape the size of Alaska is that some uncertainty will always remain. The goal is not a perfect census; it is an estimate good enough to manage a wild population sustainably, keep moose on the landscape for the people and ecosystems that depend on them, and detect problems before they become irreversible.