31

A REVIEW OF METHODS TO ESTIMATE AND MONITOR  
MOOSE DENSITY AND ABUNDANCE

Remington J. Moll1, Mairi K. P. Poisson1, David R. Heit1, Henry Jones2, Peter J. 
Pekins1, and Lee Kantar3

1Department of Natural Resources and the Environment, University of New Hampshire, 56 College 
Road, Durham, NH 03824, USA; 2New Hampshire Fish & Game Department, 200 Main Street, New 
Hampton, NH 03256, USA; 3Maine Department of Inland Fisheries and Wildlife, B State Street, 
Bangor, ME 04401, USA

ABSTRACT: Acquiring accurate and precise population parameters is fundamental to the ecological 
understanding and management and conservation of moose (Alces alces). Moose density is challenging 
to measure and often estimated using winter aerial surveys; however, numerous alternative approaches 
exist including harvest analysis, public observation, unpiloted aerial system (UAS) surveys, and cam-
era trapping. Given recent developments in a number of field and analytical techniques, there is value 
in reviewing and synthesizing the strengths and limitations of monitoring methods to best evaluate their 
respective tradeoffs in management scenarios. We reviewed 89 studies that included 131 estimates or 
indices of moose density. As expected, aerial surveys were the most common method of obtaining a 
moose density estimate (58%) followed by use of public data (e.g., harvest records = 27%); more recent 
studies employed novel methods including UAS. Most estimates (64%) failed to account for imperfect 
detection of moose (i.e., “sightability”) and this tendency has not improved over time. Density esti-
mates ranged from <0.1 to 10.6 moose/km2 (average = 0.7) and population precision, as measured by 
the 90% confidence interval, ranged from 6.5 to 120.0% of the density estimate (average = 37.4%). 
Correlations among estimates obtained for the same populations varied widely, with R2 values ranging 
from 0.02 to 0.99 (average = 0.58). Our review indicates that: 1) methods to estimate moose density 
have been dominated by aerial surveys but are diversifying, 2) precision of density estimates has been 
highly variable and on average lower than broadly accepted target benchmarks, and 3) many methods 
did not account for sightability and presumably underestimated moose density. We reflect on these 
trends and discuss how emerging methods, including camera trapping, UAS surveys, and integrated 
population modeling (IPM) can complement and improve traditional approaches. We suggest that no 
single “best” method exists, but rather the best method is one that accounts for sightability bias and 
yields target precision at reasonable cost, which vary by jurisdiction and goal.

ALCES VOL. 58: 31–49 (2022)

Key words: abundance, Alces alces, density, distribution, management, occurrence, survey

INTRODUCTION
Obtaining accurate and precise estimates of 
population parameters is fundamental to the 
ecological understanding, effective manage-
ment, and conservation of wildlife (Skalski 
et al. 2005, Sinclair et al. 2006, Silvy 2012). 
Such parameters are particularly important 
for moose (Alces alces) because this species is 
often managed for harvest (Jensen et al. 2020), 

has both negative and positive economic 
impact (e.g., vehicle collisions and ecotour-
ism, respectively; Storaas et al. 2001, 
Silverberg et al. 2003, Sample et al. 2020), 
and is hypothesized to be susceptible to global 
climate change (Murray et al. 2006, Jensen 
et al. 2020). The population parameters 
required to manage and conserve moose vary 
by location and context, but typically include 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

32

density, survival, recruitment, and composi-
tion (Gasaway et al. 1986, Krausman 2002, 
Van Ballenberghe and Ballard 2007). 
Estimates of density (hereafter, the term den-
sity includes the related term abundance) and 
proxies thereof are arguably the most funda-
mental because management is frequently 
focused on maintaining moose populations at 
specific densities over time (Leopold 1933, 
Franzmann and Schwartz 2007).

The estimated global moose population 
is ~2.2 million, with roughly half in Eurasia 
and half in North America (Timmermann and 
Rodgers 2017, Jensen et al. 2020). Recent 
reviews have highlighted variation in popula-
tion dynamics across management jurisdic-
tions, with some populations declining and 
others increasing or stable (Timmermann and 
Rodgers 2017, Jensen et al. 2020). Population 
dynamics are complex and geographically 
varied, but broadly reflect habitat composi-
tion, forest management, abiotic environ-
mental conditions, hunter harvest, predation 
pressure, and parasites (Boutin 1992, Messier 
1994, Rempel et al. 1997, Solberg et al. 1999, 
Musante et al. 2010, Jones et al. 2017, Pekins 
2020). Climate change has an underlying 
influence on these factors as well as moose 
behavior and susceptibility to parasites and 
disease (Joly et al. 2012, Tape et al. 2016, 
Montgomery et al. 2019, Pekins 2020), 
underscoring the need for techniques that 
accurately monitor moose populations over 
time (van Ballenberghe and Ballard 2007, 
Jensen et al. 2020).

Moose density is challenging to estimate 
and monitor for logistical, financial, and eco-
logical reasons. Moose are highly mobile and 
inhabit large ranges that make monitoring 
difficult by enlarging the scale required for 
adequate sampling (Krebs 2006, Singh and 
Milner-Gulland 2011, Harris et al. 2015) and 
rendering survey efforts expensive (Bontaites 
et al. 2000, Peters et al. 2014, Boyce and 
Corrigan 2017). Behavior frequently reduces 

detection or “sightability” because moose 
use dense forest cover, avoid human distur-
bance (including activities associated with 
surveying populations), and are mostly cre-
puscular or nocturnal when active (Frid and 
Dill 2002, Harris et al. 2015). Further, there 
are practical difficulties associated with sur-
veying populations in remote regions with 
high topographic relief (van Ballenberghe 
and Ballard 2007, Kellie et al. 2019). Thus, 
monitoring moose populations is an “evolv-
ing art” (Krebs 2006, p. 367) that is shaped 
by the advent and implementation of emerg-
ing technologies and methods (Boyce and 
Corrigan 2017, Oyster et al. 2018, McMahon 
et al. 2021).

Winter aerial surveys are a common 
method used to estimate and monitor moose 
density, and often conducted by helicopter 
with observers counting moose on snow 
(Gasaway et al. 1985, 1986, van Ballenberghe 
and Ballard 2007, Timmermann and Rodgers 
2017). Many jurisdictions have employed sur-
veys for decades as the backbone of monitor-
ing efforts (e.g., Alaska, USA, and Alberta, 
Canada, Alberta Environment and Parks 2016, 
Kellie et al. 2019). Traditionally conducted 
with a simple single or double-observer 
method, surveys have evolved to include a 
distance sampling approach (Gasaway et al. 
1986, Alberta Environment and Parks 2016, 
Oyster et al. 2018). Challenges associated 
with moose monitoring include high cost, 
accounting for sightability bias, and danger to 
aviators and observers (Sasse 2003, Peters 
et al. 2014, Oyster et al. 2018).

Alternatively, where sufficient harvest 
occurs, moose populations have been moni-
tored using harvest data analysis (Solberg 
et al. 1999, Skalski et al. 2005, DeCesare 
et al. 2016). Hunter observations have also 
been used as a cost-efficient index for moose 
density, although such observations require 
 calibration with other data to achieve reli-
ability (Bontaites et al. 2000, Boyce and 



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

33

Corrigan 2017). Less common approaches 
include snow tracking, pellet surveys, cam-
era trapping, and aerial surveys using 
unpiloted aerial systems (UAS) (Bobek et al. 
2005, Krester et al. 2016, Pfeffer et al. 2018, 
McMahon et al. 2021).

The challenges and importance of obtain-
ing accurate moose density estimates are 
clear and best management requires continu-
ous assessment and adaptation to evolving 
methodology. Thus, it is valuable to review 
and synthesize current methods and evaluate 
their associated strengths and limitations to 
identify and navigate their trade-offs. Here, 
we review methods used to estimate and 
monitor moose density with an emphasis on 
studies that directly analyzed a method of 
monitoring moose to derive a density esti-
mate or index. We summarize the results of 
our literature survey, the limitations of each 
method, and discuss future direction in mon-
itoring and estimating moose density.

METHODS
We performed a literature survey to review 
methods for estimating and monitoring 
moose density with an emphasis on studies 
with management application. In May 2022 
we used the Web of Science to search all col-
lections using the following Boolean string 
of terms: TITLE: (“Alces alces” OR moose) 
AND TOPIC: (abundance OR density) AND 
TOPIC: (management); the search yielded 
453 studies. We eliminated irrelevant stud-
ies including those focused on other species, 
purely mathematical or simulation-based 
studies, and reviews. We only retained stud-
ies that were primary sources and omitted 
those using density estimates from other 
sources. We retained studies that used densi-
ties to test ecological hypotheses, but only if 
they met the other criteria and included suf-
ficient details regarding study design and 
density estimation methods. We included 
the term “management” in the search to 

emphasize studies relevant to practitioners 
and to help eliminate irrelevant studies (e.g., 
those that used moose density estimates 
from other sources).

For each relevant study, we recoded 
information according to the framework 
described below (Fig. 1). We recorded the 
study location, the spatial scale at which 
inference was desired (km2; typically, the 
study area), and the method(s) employed to 
estimate or monitor the target population 
density. We then classified each method 
as follows.

We first identified the goal type (n = 3) of 
the method according to Timmermann and 
Buss (2007): 1) a census, where an attempt 
was made to count all animals within an area, 
2) a sample, where inference regarding den-
sity was achieved through sampling and sta-
tistical analysis, or 3) an index, where a 
relative measure representing density was 
desired. Next, we classified the primary 
method as either aerial, ground-based, or 
based on public observations. Aerial methods 
included fixed-wing aircraft, helicopters, and 
unpiloted aerial systems (UAS, or “drones”); 
ground methods included pellet counts, snow 
track surveys, and camera trapping; public 
methods included the use of harvest data and 
public observations (e.g., hunter or citizen 
science observations). For all methods, we 
recorded whether the survey design was sys-
tematic, random, stratified random, or 
non-random (e.g., a survey targeting a spe-
cific management area). For aerial methods 
we also recorded whether the sampling 
scheme was conducted using block searches 
(an area-based design where “blocks” might 
also be called “quadrats”, “plots”, or “sample 
units”) or strip-transects (a line-based 
design). For public observations, we recorded 
whether data reporting was mandatory (typi-
cal for harvest data) or voluntary.

We recorded the timing (i.e., seasons) of 
each method, the duration of data collection 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

34

(in years), and whether each method 
accounted for the imperfect detection of indi-
viduals (i.e., an individual is present but not 
detected; MacKenzie et al. 2002), often 
referred to as sightability (Gasaway et al. 
1986). We note that the duration of data col-
lection might not always reflect the duration 
of monitoring programs, but rather is indica-
tive of the dataset’s use for a particular analy-
sis. Thus, this value represents the duration of 
data collection used to inform the scientific 
literature rather than the true duration of 
moose monitoring efforts. Seasons were set 
as fall (September through November),  winter 
(December through April), and spring-sum-
mer (May through August). We also recorded 
whether the method distinguished the age or 
sex of individual moose. For aerial and 
ground surveys, we considered a study to 
have accounted for sightability if it formally 
accounted for undetected individuals, includ-
ing the use of double-observation methods, 
mark-resight methods, distance sampling 

methods, or previously calculated sightability 
correction factors. For public observations, 
we considered a study to have accounted for 
sightability if it formally corrected for imper-
fect reporting (e.g., a moose was observed but 
not reported) or undocumented harvest. We 
then used a logistic regression to evaluate 
whether there was a temporal trend in 
accounting for sightability, where a binary 
response of accounting for sightability was 
modeled as a function of the publication year; 
p < 0.05 was the threshold for inference. 

When reported, we recorded the mean 
population estimate and converted it to moose 
density (individuals per km2). We acknowl-
edge the substantial variation in moose densi-
ties across their circumpolar range. We report 
the densities of the studies in our review to 
quantify this variation broadly and provide to 
context for the various monitoring methods, 
which might be informed by local moose 
densities (e.g., tailoring survey methods for a 
low-density population; Hinton et al. 2022). 

Fig. 1. The framework used to organize and summarize methods used to estimate and monitor moose 
(Alces alces) density and abundance. See text for detailed descriptions of categories. UAS = 
unpiloted aerial system.



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

35

We recorded the precision of that estimate 
following Gasaway et al. (1986), according 
to the equation:

 u p
p

ˆ
ˆ
−  (1)

where u is the upper value of the 90% confi-
dence interval and p̂ is the estimated popula-
tion value (density or abundance). This 
precision metric is widely recognized and 
cited as a target benchmark for management 
decisions (e.g., Timmermann 1993, Bontaites 
et al. 2000, Peters et al. 2014). For studies 
that reported multiple density estimates and 
precisions (e.g., across management units or 
years), we recorded the average density and 
precision across all estimates. In cases where 
only the standard error was reported for pre-
cision, we converted this error into a confi-
dence interval by multiplying it by 1.645 (the 
Z-value for a 90% confidence interval). 
Lastly, we recorded whether a study formally 

compared one population estimate to an esti-
mate produced via another method, and if so, 
the Pearson’s correlation coefficient (r) or R2 
value of each such comparison.

RESULTS
A total of 89 (20%) of 453 studies returned 
by the literature search met the criteria for 
review (see Appendix 1), with most (74%) 
conducted in North America and the remain-
der in Fennoscandia and Eurasia (Fig. 2). Of 
these, 65 provided details regarding the spa-
tial scale of desired inference that ranged 
from as small as 6.0 km2 to >13.6 million 
km2 (see Appendix 1); the average and 
median values were 278,000 km2 and 
3,456 km2, respectively, indicating that the 
distribution of these scales was non-normal 
and heavily right-skewed.

The 89 studies contained 131 estimates 
or indices used to monitor moose density. 
However, the proportional statistics below 

Fig. 2. Geographic distribution of studies included in a literature survey of methods used to estimate 
and monitor moose (Alces alces) density and abundance conducted in May 2022. Note that the sum 
of studies here exceeds the total number of studies reviewed because several studies were conducted 
in more than one country.



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

36

omit certain studies lacking sufficient detail 
to adequately characterize the methodology 
(i.e., NAs in a given category were dropped; 
see Appendix 1). Across all methods, the 
most common goal type (n = 79, 61%) was a 
sample or statistical representative of the 
broader population density; less common 
were the index (n = 29, 22%) and census 
(n = 22, 17%) goal types. A single method 
employed a cohort analysis that combined 
multiple data types (see Appendix 1). 

Of the three primary methods, aerial sur-
veys were most common (n = 76, 58%), fol-
lowed by public reporting (n = 35, 27%) and 
ground surveys (n = 19, 15%) (Fig. 3). 
Helicopters (n = 35, 46%) and fixed-wing 
aircraft (n = 26, 34%) were the most com-
mon flight modes, in which block (n = 56, 
76%) and strip-transect surveys (n = 17, 
23%) were used exclusively, other than a sin-
gle exception, being their combination. 
Harvest (n = 20, 57%) and public 

observation (n = 14, 40%) were the most 
common public reporting methods with man-
datory (n = 19, 56%) and voluntary (n = 15, 
44%) reporting used in both. Pellet counts 
(n = 10, 53%) and direct observation (n = 6, 
32%) were the most common ground survey 
methods; snow tracking methods (n = 2, 
11%) were used less frequently (Fig. 3). 

The most common survey design was 
non-random (n = 60, 49%), followed by 
stratified random (n = 37, 30%), systematic 
(n = 19, 15%), and random (n = 7, 6%). Most 
(67%) non-random survey designs were 
associated with public observations, while 
all stratified random survey designs, except 
one, were associated with aerial methods 
(Table 1; Appendix 1). Approximately one-
third of all methods formally accounted for 
sightability (n = 45, 36%). The application 
of a sightability correction for density esti-
mates in studies did not change linearly over 
time (β = 0.004, df = 122, p = 0.77; Fig. 4). 

Fig. 3. Primary and secondary methods used to 
estimate and monitor moose (Alces alces) 
density and abundance according to a literature 
survey conducted in May 2022. Primary 
methods are represented in the central circle; 
associated secondary methods are represented 
in the outer circle. One method employed a 
cohort analysis that combined multiple methods, 
and thus was not included in this figure.

Fig. 4. Model predictions of the probability that a 
study estimating or monitoring moose (Alces 
alces) densities accounted for sightability, or 
the imperfect detection of individual moose, as 
a function of the year the study was published. 
The slope of the model is not significantly 
different from zero (p = 0.83). Model was fit to 
data collected from studies returned from a 
literature survey of methods used to estimate 
and monitor moose (Alces alces) density and 
abundance conducted in May 2022.



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

37

Helicopter surveys were the only method 
that accounted for sightability most (59%) of 
the time (Table 1). Fifty-four studies reported 
68 mean population densities ranging from 
<0.1 to 10.6 moose/km2 (average = 0.7; 
Fig. 5A). The precision of 30 estimates 
reported or determinable in 24 different 
studies ranged from 0.07 to 1.20 moose/km2 
(average = 0.37; Fig. 5B); half of these esti-
mates were larger than the 0.25 target bench-
mark suggested by Gasaway et al. (1986).

Of the 131 estimates, 87% (n = 114) 
were conducted during a specific season, 
most in winter (n = 63, 55%) and fall (n = 
31, 27%); spring-summer estimates were 
uncommon (n = 8), as were year-round esti-
mates (n = 5). Aerial estimates more fre-
quently occurred in winter (80% of the 
time), while public estimates typically 
occurred in fall (77% of the time; Table 1). 
Eighty-one estimates (62%) distinguished 
age (calf vs. adult) and/or sex of moose; 
however, those studies distinguishing sex 

did not necessarily distinguish age or vice 
versa (see Appendix 1).

Study duration ranged from 1 to 
73 years. Twenty-four methods (18%) 
employed a single year of study, typically a 
single season; 38 methods (29%) were 
employed for two to five years, 17 (13%) for 
6–10 years, 15 (11%) for 11–20 years, and 
15 for >20 years. Twenty-two studies did not 
report duration or varied so greatly in dura-
tion and/or spatial coverage (due to funding 
limitations or other barriers) that they were 
omitted from the summary (see Appendix 1).

Seventeen studies provided Pearson’s 
correlation coefficient (r) or R2 values to 
compare methods. The most frequent com-
parisons included moose observations by 
hunters, hunter success rate, density esti-
mates from aerial surveys, and harvest data 
analysis (see Appendix 2). The R2 values 
from these comparisons ranged from 0.02 to 
0.99 (average = 0.58). Sample size for these 
values ranged from 6 to 111 (average = 22). 

Table 1. A summary of the number of methods used to estimate and monitor moose (Alces alces) density 
and abundance according to a literature survey conducted in May 2022.

 
 

 
 

Aerial Ground Public

Fixed-wing Helicopter UAS Pellets Tracks Harvest Obs.

Goal Census 8 9 0 0 1 0 2

Index 2 0 0 3 0 8 14

Sample 16 26 1 7 0 12 4

Survey design Non-random 6 7 0 4 1 18 18

Random 3 1 0 1 0 1 0

Strat. Rand. 12 16 0 0 0 1 0

Systematic 3 9 1 4 0 0 1

Timing Fall 2 0 0 0 0 15 12

Spr-Su 1 0 1 6 0 0 0

Winter 17 32 0 2 1 1 1

Other 5 1 0 1 0 0 6

Sight-ability Yes 13 20 1 2 0 2 1

No 12 14 0 8 1 16 19

UAS = unpiloted aerial system, Obs. = observation, Stat. Rand. = stratified random.



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

38

The lowest R2 values were associated with 
comparisons between hunter harvest per unit 
effort and minimum population counts from 
aerial surveys, and between change-in-ratio 
abundance estimates and rates of moose-
train collisions (Appendix 2). A comparison 
between hunter harvest per unit effort and an 
aerial density estimate yielded the highest 
correlation using a non-linear model fit with 

small sample size (n = 6; Appendix 2). The 
complete references for reviewed studies 
are provided in Appendix 1 available on 
the ALCES website https://alcesjournal.org/
index.php/alces.

DISCUSSION
As expected, this literature survey revealed a 
historical reliance on aerial methods to 

Fig. 5. The mean density of moose/km2 (Panel A) and the density estimate precision (Panel B) 
according to a literature survey of methods used to estimate and monitor moose (Alces alces) density 
and abundance conducted in May 2022. For clarity, one outlier density estimate of 10.6 moose/km2 
from a local-scale study was omitted in Panel A (see Appendix 1).

https://alcesjournal.org/index.php/alces
https://alcesjournal.org/index.php/alces


ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

39

estimate moose density. Many authors noted 
limitations to conducting an ideal aerial sur-
vey, especially financial and logistical costs 
and weather (e.g., Peek 1974, Nygrén and 
Pesonen 1993, Bowyer et al. 1999, Harris 
et al. 2015, Kellie et al. 2019). Although 
sightability bias was acknowledged fre-
quently as an issue, almost two-thirds of 
density estimates across studies did not 
account for it (Fig. 4). We also found high 
variation in the precision of density esti-
mates, with only half achieving the bench-
mark management target of Gasaway et al. 
(1986). Landscape, habitat, and social fac-
tors all appeared to play a role in method 
choice. For example, aerial surveys were 
particularly common in Alaska, USA, where 
habitat is more open and individuals tend to 
cluster across the landscape, whereas har-
vest and hunter-based approaches were com-
mon in Scandinavia where harvest is high 
and reporting is strong (Appendix 1). Finally, 
several recent studies highlighted new tech-
nologies and modern analytical develop-
ments that hold promise for moose 
population monitoring, although key ques-
tions and challenges remain regarding their 
reliability and overall efficacy (Table 2).

Surprisingly, the critical issue of 
sightability bias was usually unaccounted 
for, with no evidence of improvement over 
time (Fig. 4), despite its influence on the 
accuracy and precision of moose population 
estimates (Evans et al. 1966, Caughley 1974, 
Gasaway et al. 1986, Peters et al. 2014, 
Harris et al. 2015). Multiple studies using 
aerial methods have documented declines in 
moose sightability with increasing canopy 
cover, especially in conifer patches. For 
example, using thermal drone surveys, 
McMahon et al. (2021) found that sightabil-
ity declined from near 100% in open habitat 
to < 25% in 75% canopy cover, and Peters 
et al. (2014) estimated sightability along tran-
sects to be as low as 46% during helicopter 

flights over non-ideal snow conditions. 
Regarding public data, sightability can influ-
ence physical observation rates or manifest 
as imperfect reporting or undocumented har-
vest (e.g., illegal harvest). In these cases, the 
number of moose present within or removed 
from the population would be underesti-
mated, and could be a common scenario. 
Such underestimation is likely pervasive for 
moose and other species where imperfect 
detection of individuals is the norm 
(Caughley 1974, MacKenzie et al. 2002, 
Stephens et al. 2006). Therefore, a key area 
of improvement for moose monitoring meth-
ods is to explicitly and rigorously account 
for sightability bias.

Studies accounting for sightability bias 
varied in approach, ranging from extensive 
field efforts and sophisticated statistical 
modeling (e.g., Oyster et al. 2018) to correc-
tion factors based upon previous work in the 
same study area using similar methods (e.g., 
Bontaites et al. 2000). To increase sightabil-
ity, aerial surveys are typically conducted in 
snow conditions that accentuate the contrast 
between moose and their background envi-
ronment (Gasaway et al. 1986); however, 
this reliance on snow cover is problematic 
from a scheduling perspective. The expecta-
tion is that this problem will worsen presum-
ing that climate change reduces the length 
and timing of snow cover, especially along 
the southern range of moose (Bormann et al. 
2018, Kellie et al. 2019, Jensen et al. 2020). 
Another approach to quantify sightability in 
aerial methods is to search for a subset of 
VHF- or GPS-collared individuals and quan-
tify their detection probabilities as a function 
of habitat or environmental conditions (e.g., 
Peters et al. 2014). This approach is effective 
but usually part of a separate ecological 
study given the substantial costs of capture, 
collars, and tracking individual moose. An 
approach generally absent in our literature 
survey, but common for other taxa, is using 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

40

repeated surveys in space or time to quantify 
detection probabilities (e.g., as in occupancy 
modeling; MacKenzie et al. 2002, Tyre et al. 
2003). For moose, this method could be 
applied by using repeated sampling over 
time with camera traps, with multiple 
observers in ground-based surveys or among 
the public, or by flying transects more than 
once in rapid succession such that popula-
tion closure assumptions are reasonably met 
(Adams et al. 1997, Rota et al. 2009). 
Depending on context, such strategies are a 
cost-effective alternative to using radio-
marked animals. However, we note that 

sightability is method-specific and the cor-
rection from one method might not apply to 
another. Whatever the method, it is crucial to 
consider sightability bias because a failure to 
do so likely results in density underestima-
tion and leads to bias in other types of infer-
ence (e.g., wildlife-habitat relationships; see 
Kéry et al. 2010). For example, the early 
successional habitat that moose prefer is 
associated with decreased sightability. 
Therefore, a survey that does not separate 
the observational effects of habitat on moose 
detection (i.e., a negative effect) from the 
ecological effects (i.e., a positive effect) 

Table 2. A comparative summary of the predominant methods used to estimate and monitor moose (Alces 
alces) density and abundance.

Method Advantages Limitations Outlook

Fixed-wing and 
helicopter surveys

Easily understood and 
trusted by many 
stakeholders; sometimes 
considered the “gold 
standard”

Logistical and financial costs; 
danger to aviators; often 
dependent on weather and 
snow, and sensitive to terrain 
variation

Useful method in 
jurisdictions where 
resources enable consistent 
surveys; overall prevalence 
likely to decline as other 
methods develop 

Camera trapping Relatively low field effort 
and expense; non-invasive

Must process voluminous 
image data efficiently; 
analytical methods are 
rapidly changing

Increase in prevalence as 
analytical methods continue 
to develop

Harvest data  
analysis

Inexpensive; produces 
reliable estimates of 
population when 
appropriate statistical 
methods are applied

Limited to areas where 
harvest is substantial; bias in 
non-random hunter behavior 
must be accounted for

A foundational method 
when harvest is substantial

Pellet and track 
surveys

Relatively inexpensive; 
non-invasive

Requires calibration with 
other density estimates to 
achieve reliability; low 
precision; field intensive in 
terms of person-power; 
tracking requires snow

Pellet surveys could become 
more powerful when 
combined with genetic 
analyses; tracking a useful 
low-tech index at local 
scales

Public or hunter 
observations

Inexpensive; engages 
multiple stakeholders

Study design must be 
carefully considered to avoid 
reporting bias; indices do not 
always track with population 
dynamics

Viable management tool 
when calibration using other 
methods occurs regularly

UASs Less expensive and safer 
than traditional aerial 
surveys

Spatial extent and locations 
of surveys often limited by 
regulations; unproven as 
broadscale monitoring 
method; non-trivial initial 
purchase costs

Current applications are 
most effective for local 
scale studies; regulatory 
changes could precipitate 
rapid increase in capacity

UAS = Unpiloted Aerial System. Italicized methods indicate those appearing in recent literature.



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

41

risks conflating observation error with eco-
logical inference.

Indices, especially those based on hunter 
observation and success rate, were a rela-
tively common approach in tracking moose 
population trends (Appendix 2). Indices are 
often employed because they are cheaper – 
sometimes much cheaper – than more field 
intensive and potentially hazardous methods 
such as helicopter surveys (Sasse 2003, 
Krebs 2006). However, indices must be 
carefully calibrated and regularly reviewed 
to ensure reliability (Caughley 1974, 
Bontaites et al. 2000, Hatter 2001, Ueno 
et al. 2014). Although several papers reported 
high correlations between an index and pop-
ulation density, sample sizes were often 
small (e.g., n < 10) and a number of compar-
isons had only moderate correlation (see 
Appendix 2). For example, moose observa-
tions made by hunters in Maine, USA exhib-
ited only a modest correlation with densities 
obtained from aerial helicopter surveys (R2 = 
0.32; n = 13; Kantar and Cumberland 2013). 
Recent work in Ontario, Canada showed that 
the relationship between harvest and abun-
dance varies by age and sex class, thereby 
highlighting the need to calibrate indices by 
demographic categories (Priadka et al. 
2020). This study also found non-linear rela-
tionships between harvest effort, harvest, 
and abundance, suggesting that harvest 
might underestimate abundance when har-
vest effort is high. These observations 
emphasize that such indices must account 
for non-randomness and change in hunter 
activity relative to spatial coverage, effort, 
participation rate, weather, and technique, as 
well as potential non-linear relationships 
with hunter effort, harvest success, and 
moose abundance (Fryxell et al. 1988, 
Kantar and Cumberland 2013, Larson et al. 
2014, Priadka et al. 2020).

More broadly, our literature survey sug-
gests there could be untapped utility to use 

indices (e.g., hunter observations) in more 
comprehensive integrated population mod-
els (IPMs). Such models combine multiple 
data sources to improve the precision of 
population parameter estimates, thereby 
improving management efficiency (Schaub 
and Abadi 2011, Arnold et al. 2018). For 
example, Månsson et al. (2011) used simula-
tions to demonstrate how the combination of 
hunter observations and pellet counts could 
more accurately inform management than 
more expensive aerial surveys. Similarly, 
Marrotte et al. (2021) integrated aerial sur-
vey data with hunter-reported data into a sin-
gle model to estimate moose population 
trends in relation to harvest and predation. 
Each of these data sources had unique lim-
itations; the aerial surveys were infrequent 
but had higher accuracy while the hunter 
reports were more frequent but less stan-
dardized. The IPM developed by Marrotte 
et al. (2021) partially overcame these chal-
lenges, resulting in increased confidence in 
overall moose population trends. 

The target precision for moose density 
estimates was only achieved half of the time 
in the studies we analyzed (Fig. 5B; 
Appendix 1). In their seminal paper, 
Gasaway et al. (1986) suggested a target pre-
cision with a confidence interval width of 
±25% of the population estimate; however, 
the rationale for this target was not provided 
and they noted that higher precision was 
desirable, but often prohibitively expensive 
(p. 5). These authors also endorsed a strati-
fied random sampling design as a means to 
achieve target precision, which was an 
approach used in 50% of the aerial surveys 
we reviewed. Nonetheless, certain studies 
found that stratification did not increase pre-
cision, especially in low-density populations 
(e.g., Crete et al. 1986). The expanding suite 
of remote-sensing products (e.g., land cover 
and digital elevation maps) is making strict 
stratification less important than a random 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

42

design that captures the full range of spatial 
conditions in a study area (Fletcher and 
Fortin 2018). Spatial modeling using remote-
ly-sensed covariates can improve accuracy 
and precision, while enabling prediction of 
moose density beyond sample units (Ver 
Hoef 2008, Michaud et al. 2014). 

The degree to which target precision 
(Gasaway et al. 1986) is acceptable for man-
agement objectives is an open and con-
text-dependent question. Ideally, initial 
target precision would be informed by power 
analyses and simulation, then updated with 
field data in an adaptive management frame-
work (Steidl et al. 1997, Lyons et al. 2008). 
The utility of this approach was demon-
strated by Boyce et al. (2012), where 
tradeoffs between infrequent but accurate 
aerial surveys and frequent but less accurate 
kill-per-unit-effort harvest data were mathe-
matically explored using population projec-
tion analysis. Simulation can also inform 
study design by enabling researchers and 
managers to estimate required sample sizes 
for particular precisions (Hinton et al. 2022). 
For example, Gasaway et al. (1986) recom-
mended an initial aerial survey to stratify 
landscapes by coarse-scale moose popula-
tion densities. However, these flights can be 
expensive, thus simulation using known, 
assumed, or preliminary data (e.g., habitat 
suitability) represent an attractive alternative 
to traditional Gasaway-type stratification. 
Precision can also be improved through the 
use of IPMs (Marrotte et al. 2021; discussed 
further below).

It is often difficult to monitor wildlife 
population trends over sufficient periods 
because of the many sources of variation in 
ecological systems and the short-term fund-
ing cycles that support monitoring efforts 
(Field et al. 2007). Indeed, population moni-
toring efforts with other taxa are often biased 
in site selection or not conducted long 
enough to reliably detect population trends 

(Fournier et al. 2019, White 2019). Likewise, 
many studies uncovered by our literature 
survey lasted just one year, or when they 
occurred across multiple years were often 
limited by cost that influenced the survey 
area in any given year (Appendix 1). Such 
constraints emphasize the need to creatively 
and effectively design monitoring programs 
that inform stated management objectives in 
a given landscape context. In locations with 
substantial harvest, cohort analysis and sta-
tistical population reconstruction represent 
powerful methods that can offer high accu-
racy and precision across broad spatiotem-
poral scales (Solberg et al. 1999, Skalski 
et al. 2005). In addition, the impacts of 
sightability bias related to unknown sources 
of harvest should be considered (Skalski 
et al. 2005, Timmermann and Rodgers 
2017). For low-density populations or those 
with little to no harvest, multiple methods 
that complement each other and account for 
sightability bias are paramount and will 
likely co-evolve with technological and ana-
lytical developments. In particular, the 
advantages of advanced statistical modeling 
coupled with robust geospatial data should 
be used to “get the most” out of sparse data 
that are expensive to collect for low-density 
populations. Hinton et al. (2022) provide an 
example of this approach by combining 
non-linear generalized additive models, 
adaptive sampling, and informative geospa-
tial covariates to monitor a low density pop-
ulation over a large (~25,000 km2) landscape 
in New York, USA.

We conclude that several methods 
deserve additional consideration, and pend-
ing evaluation, could be implemented 
increasingly in moose density estimation 
and monitoring, foremost camera traps, 
UAS, and IPM. Technological advance-
ments have increased wildlife detection 
capability of camera traps and improved 
their field reliability (Burton et al. 2015, 



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

43

Steenweg et al. 2017). Statistical models 
that estimate population density with cam-
era trap data have also advanced (Kéry and 
Royle 2015, Gilbert et al. 2021b), with sev-
eral capable of estimating habitat relation-
ships used to predict population density in 
non-sampled areas (Moeller et al. 2018, 
Nakashima et al. 2020), thereby enabling 
broader monitoring across spatiotemporal 
scales. Camera traps can also potentially 
collect information on sex ratio and age 
structure of moose. Although no study in the 
literature survey used camera traps, studies 
are being implemented in the northeastern 
United States (R. J. Moll and H. Jones, pers. 
comm.) and appear in recent literature (see 
Pfeffer et al. 2018).

Likewise, UAS surveys of wildlife pop-
ulations are becoming more common due to 
technological advances related to field reli-
ability and sensor capability (Linchant et al. 
2015, Witczuk et al. 2018). Recent UAS sur-
veys for moose in Minnesota yielded high 
accuracy at a cost lower than traditional 
aerial surveys; albeit, scale (time and area) 
and flight regulations remain as issues 
(McMahon et al. 2021). Nonetheless, UAS 
can provide aerial surveys at lower cost and 
without the potential danger of traditional 
flights (Sasse 2003), and will be especially 
useful for local scale surveys (McMahon 
et al. 2021).

IPMs represent an underused approach 
to improve moose monitoring by enhancing 
analytical power, reducing bias, and increas-
ing the precision of population estimates 
(Schaub and Abadi 2011). For example, 
IPMs can be used to combine known-fate 
data from collared individuals, counts from 
aerial surveys, and occupancy data collected 
by hunter observations to estimate popula-
tion density and demographic parameters 
such as survival (Zipkin and Saunders 
2018). The ultimate goals of IPMs are usu-
ally to estimate abundance over time (i.e., 

population trends) and obtain demographic 
parameters that are traditionally difficult to 
estimate from a single data source (e.g., 
immigration). Many moose monitoring pro-
grams seek these parameters, particularly 
population change over time, to inform man-
agement. To take advantage of the power of 
IPMs, researchers and managers might first 
build a process-based model of the target 
population using standard stage- or 
age-structured population matrices (Boyce 
et al. 2012). Then, multiple, independent 
datasets could be jointly analyzed to esti-
mate parameters of that model, which would 
in turn be used to predict and project popula-
tion trends under management scenarios. 
IPMs require that some parameters be shared 
among projects with different datasets. 
Examples of such shared parameters in 
moose monitoring efforts might include sur-
vival (e.g., from capture-recapture or telem-
etry data) and recruitment rates (e.g., from 
aerial counts or harvest data; Zipkin and 
Saunders 2018). While use of IPMs has 
increased dramatically in other applications 
and taxa (Schaub and Abadi 2011, Arnold 
et al. 2018, Zipkin and Saunders 2018, 
Gilbert et al. 2021a), they are rarely used for 
moose (although see recent implementation 
by Marrotte et al. 2021). We direct readers 
interested in using IPMs for moose monitor-
ing to the comprehensive reviews and texts 
published in recent years for additional 
details regarding data requirements and ana-
lytical implementation methods (Schaub and 
Abadi 2011, Zipkin and Saunders 2018, 
Schaub and Kéry 2021).

Moose and moose managers face a myr-
iad of environmental and conservation chal-
lenges in the 21st century, and using accurate 
and reliable population information will be 
paramount in management decisions. The 
broad variation in range, habitat, environ-
mental conditions, and population density 
precludes a single survey method that can 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

44

address each jurisdictional goal or need. 
While aerial surveys are often described as 
the best method for population estimation 
and monitoring (e.g., Peters et al. 2014, 
Boyce and Corrigan 2017), we suggest that 
the “best” method is case-specific and meets 
an acceptable target precision while account-
ing for sightability at a reasonable cost. 
Fortunately, the technological and analytical 
toolboxes available to researchers and man-
agers have never been fuller, and develop-
ments continue. We emphasize the judicious 
adaptation and evaluation of new methods 
and approaches to address context-specific 
needs and objectives, and likewise encour-
age coordinated efforts across jurisdictions 
and spatial scales.

ACKNOWLEDGEMENTS
We thank the University of New Hampshire 
and the National Science Foundation Graduate 
Fellowship Program for supporting our 
research. We thank S. Richard for assistance 
with the literature survey. Partial funding was 
provided by the New Hampshire Agricultural 
Experiment Station. This is Scientific 
Contribution Number 2941. This work was 
supported by the USDA National Institute of 
Food and Agriculture Hatch Project 1024128. 
We thank B. Patterson, an anonymous 
reviewer, and B. McLaren for comments that 
improved the manuscript.

REFERENCES
AdAms, K. P., P. J. PeKins, K. A. GustAfson, 

and K. BontAites. 1997. Evaluation of 
infrared technology for aerial moose 
surveys in New Hampshire. Alces 33: 
129–140.

AlBertA environment and PArKs. 2016. 
Aerial ungulate surveys using distance 
sampling techniques – protocol manual. 
Alberta Environment and Parks, 
Edmonton, Alberta, Canada.

Arnold, t. W., r. G. ClArK, d. n. Koons, 
and M. sChAuB. 2018. Integrated popu-
lation models facilitate ecological 
understanding and improved manage-
ment decisions. Journal of Wildlife 
Management 82: 266–274. doi: 10.1002/
jwmg.21404

BoBeK, B., d. mertA, P. sulKoWsKi, and A. 
siutA. 2005. Moose recovery plan for 
Poland: main objectives and tasks. Alces 
41: 129–138.

BontAites, K. m., K. A. GustAfson, and R. 
mAKin. 2000. A Gasaway-type moose 
survey in New Hampshire using infrared 
thermal imagery: preliminary results. 
Alces 36: 69–75.

BormAnn, K. J., r. d. BroWn, C. derKsen, 
and T. H. PAinter. 2018. Estimating 
snow-cover trends from space. Nature 
Climate Change 8: 924–928. doi: 
10.1038/s41558-018-0318-3

Boutin, S. 1992. Predation and moose popu-
lation dynamics: a critique. Journal of 
Wildlife Management 56: 116–127. doi: 
10.2307/3808799

BoyCe, m. s., P. W. J. BAxter, and h. P. 
PossinGhAm. 2012. Managing moose 
harvests by the seat of your pants. 
Theoretical Population Biology 82: 
340–347.

BoWyer, r. t., m. C. niCholson, e. m. 
molvAr, and J. B. fAro. 1999. Moose on 
Kalgin Island: Are density-dependent 
processes related to harvest? Alces 35: 
73–89.

BoyCe, m. s., and R. CorriGAn. 2017. Moose 
survey app for population monitoring. 
Wildlife Society Bulletin 41: 125–128. 
doi: 10.1002/wsb.732

Burton, A. C., e. neilson, d. moreirA, A. 
lAdle, r. steenWeG, J. t. fisher, e. 
BAyne, and S. Boutin. 2015. Wildlife 
camera trapping: a review and recom-
mendations for linking surveys to eco-
logical processes. Journal of Applied 
Ecology 52: 675–685. doi: 10.1111/ 
1365-2664.12432



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

45

CAuGhley, G. 1974. Bias in aerial survey. 
Journal of Wildlife Management 38: 
921–933. doi: 10.2307/3800067

Crete, m., l. rivest, h. JoliCoeur, J. 
BrAssArd, and F. messier. 1986. 
Predicting and correcting helicopter 
counts of moose with observations made 
from fixed-wing aircraft in southern 
Quebec. Journal of Applied Ecology 23: 
751–761. doi: 10.2307/2403931

deCesAre, n. J., J. r. neWBy, v. J. 
BoCCAdori, t. Chilton-rAdAndt, t. 
their, d. WAltee, K. Podruzny, and J. 
A. Gude. 2016. Calibrating minimum 
counts and catch-per-unit-effort as indi-
ces of moose population trend. Wildlife 
Society Bulletin 40: 537–547. doi: 
10.1002/wsb.678

evAns, C. d., W. A. troyer, And C. J. 
lensinK. 1966. Aerial census of moose 
by quadrat sampling units. Journal of 
Wildlife Management 30: 767–776. doi: 
10.2307/3798283

field, s. A., P. J. o’Connor, A. J. tyre, and H. 
P. PossinGhAm. 2007. Making monitoring 
meaningful. Austral Ecology 32: 485–491. 
doi: 10.1111/j.1442-9993. 2007.01715.x

fletCher, R. J., and M.-J. fortin. 2018. 
Spatial Ecology and Conservation 
Modeling: Applications with R. Springer 
Nature Switzerland, Cham, Switzerland.

fournier, A. m. v., e. r. White, and S. B. 
heArd. 2019. Site-selection bias and 
apparent population declines in long-
term studies. Conservation Biology 33: 
1370–1379. doi: 10.1111/cobi.13371

frAnzmAnn, A. W., and C. C. sChWArtz, edi-
tors. 2007. Ecology and management of 
the North American Moose. 2nd edition. 
University Press of Colorado, Boulder, 
Colorado, USA.

frid, A., and L. dill. 2002. Human-caused 
disturbance stimuli as a form of preda-
tion risk. Ecology and Society 6: 11. doi: 
10.5751/ES-00404-060111

fryxell, J. m., W. e. merCer, and R. B. 
GellAtely. 1988. Population dynamics 

of Newfoundland moose using cohort 
analysis. Journal of Wildlife 
Management 52: 14–21. doi: 
10.2307/3801050

GAsAWAy, W. C., S. D. duBois, and S. J. 
hArBo. 1985. Biases in aerial transect 
surveys for moose during May and June. 
Journal of Wildlife Management 49: 
777–784. doi: 10.2307/3801711

_____., _____, d. J. reed, and S. J. hArBo. 
1986. Estimating moose population 
parameters from aerial surveys. 
Biological Papers of the University of 
Alaska 22: 1–99.

GilBert, N. A., B. S. PeAse., C. m. AnhAlt-
dePies, J. d. J. ClAre, J. l. stenGlein, P. 
A. toWnsend, t. r. vAn deelen, and B. 
zuCKerBerG. 2021a. Integrating harvest 
and camera trap data in species distribu-
tion models. Biological Conservation 
258: 109147. doi: 10.1016/j.
biocon.2021.109147

_____, J. d. J. ClAre, J. l. stenGlein, and B. 
zuCKerBerG. 2021b. Abundance estima-
tion of unmarked animals based on cam-
era-trap data. Conservation Biology 35: 
88–100. doi: 10.1111/cobi.13517

hArris, r. B., m. AtAmiAn, h. ferGuson, 
and I. Keren. 2015. Estimating moose 
abundance and trends in northeastern 
Washington State: index counts, 
sightability models, and reducing uncer-
tainty. Alces 51: 57–69.

hAtter, i. W. 2001. An assessment of catch 
per unit effort to estimate rate of change 
in deer and moose populations. Alces: 
37: 71–77.

hinton, J. W., r. e. WheAt, P. sChuette, J. 
e. hurst, d. W. KrAmer, J. h. stiCKles, 
and J. l. frAir. 2022. Challenges and 
opportunities for estimating abundance 
of a low-density moose population. The 
Journal of Wildlife Management 86: 
e22213. doi: 10.1002/jwmg.22213

Jensen, W., r. v. reA, C. e. Penner, J. r. 
smith, e. v. BrAGinA, e. rAzenKovA, l. 
BAlCiAusKAs, h. BAo, s. BystiAnsKy, s. 



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

46

Csányi, z. ChovAnovA, G. done, K. 
hACKländer, m. heuriCh, G. JiAnG, A. 
KAzArez, J. Pusenius, e. J. solBerG, r, 
veeroJA, and F. Widemo. 2020. A review 
of circumpolar moose populations with 
emphasis on Eurasian Moose distribu-
tions and densities. Alces 56: 63–78.

Joly, K., P. A. duffy, and t. s. ruPP. 2012. 
Simulating the effects of climate change 
on fire regimes in Arctic biomes: impli-
cations for caribou and moose habitat. 
Ecosphere 3: art36. doi: 10.1890/
ES12-00012.1

Jones, h., P. J. PeKins, l. e. KAntAr, m. 
o’neil, and D. ellinGWood. 2017. 
Fecundity and summer calf survival of 
moose during 3 successive years of 
winter tick epizootics. Alces 53: 
85–98.

KAntAr, l. e., and r. e. CumBerlAnd. 2013. 
Using a double-count aerial survey to 
estimate moose. Alces 49: 29–37.

Kellie, K. A., K. e. Colson, and J. h. 
reynold. 2019. Challenges to monitor-
ing moose in Alaska. Alaska Department 
of Fish and Game, Juneau, Alaska, 
USA.

Kéry, m., B. GArdner, and C. monnerAt. 
2010. Predicting species distributions 
from checklist data using site-occupancy 
models. Journal of Biogeography 37: 
1851–1862. doi: 10.1111/j.1365-2699. 
2010.02345.x

Kéry, M., and J. A. royle. 2015. Applied 
Hierarchical Modeling in Ecology: 
Analysis of Distribution, Abundance 
and Species Richness in R and BUGS. 
Volume 1. Prelude and Static Models. 
Elsevier, San Diego, California, USA.

KrAusmAn, P. r. 2002. Introduction to 
Wildlife Management. Pearson, Upper 
Saddle River, New Jersey, USA.

KreBs, C. J. 2006. Mammals. Pages 351–
369 in W. Sutherland, editor. Ecological 
Census Techniques: A Handbook. 2nd 
edition. Cambridge University Press, 
Cambridge, England.

Krester, h., m. Glennon, A. WhitelAW, A. 
hurt, K. PilGrim, and m. sChWArtz. 
2016. Scat-detection dogs survey low 
density moose in New York. Alces 52: 
55–66.

lArson, l. r., r. C. stedmAn, d. J. deCKer, 
W. f. siemer, and m. s. BAumer. 
2014. Exploring the social habitat 
for hunting: toward a comprehensive 
framework for understanding hunter 
recruitment and retention. Human 
Dimensions of Wildlife 19: 105–122. 
doi: 10.1080/10871209.2014.850126

leoPold, A. 1933. Game Management. 
Charles Scribner’s Sons, New York, 
New York, USA.

linChAnt, J., J. lisein, J. semeKi, P. leJeune, 
and C. vermeulen. 2015. Are unmanned 
aircraft systems (UASs) the future of 
wildlife monitoring? A review of accom-
plishments and challenges. Mammal 
Review 45: 239–252. doi: 10.1111/
mam.12046

lyons, J. E., M. C. runGe, h. P. lAsKoWsKi, 
and W. L. KendAll. 2008. Monitoring in 
the context of structured decision-mak-
ing and adaptive management. Journal 
of Wildlife Management 72: 1683–1692. 
doi: 10.2193/2008-141

mACKenzie, d. i., J. d. niChols, G. B. 
lAChmAn, s. droeGe, J. A. royle, and 
C. A. lAnGtimm. 2002. Estimating site 
occupancy rates when detection proba-
bilities are less than one. Ecology 83: 
2248–2255. doi: 10.1890/0012-9658 
(2002)083[2248:ESORWD]2.0.CO;2

månsson, J., C. e. hAuser, h. Andrén, and 
h. P. PossinGhAm. 2011. Survey method 
choice for wildlife management: the 
case of moose Alces alces in Sweden. 
Wildlife Biology 17: 176–190. doi: 
10.2981/10-052

mArrotte, r. r., B. r. PAtterson, and J. 
m. northruP. 2021. Harvest and densi-
ty-dependent predation drive long-term 
population decline in a northern ungu-
late. Ecological Applications e2629. 



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

47

doi: 10.22541/au.162210989.948595 
93/v1

mCmAhon, m. C., m. A. ditmer, e. J. isAAC, 
s. A. moore, and J. d. forester. 2021. 
Evaluating unmanned aerial systems for 
the detection and monitoring of moose 
in northeastern Minnesota. Wildlife 
Society Bulletin: 45: 312–324. doi: 
10.1002/wsb.1167

messier, F. 1994. Ungulate population mod-
els with predation: a case study with the 
North American moose. Ecology 75: 
478–488. doi: 10.2307/1939551

miChAud, J. s., n. C. CooPs, m. e. AndreW, 
m. A. Wulder, G. s. BroWn, and G. J. 
m. riCKBeil. 2014. Estimating moose 
(Alces alces) occurrence and abundance 
from remotely derived environmental 
indicators. Remote Sensing of 
Environment 152: 190–201. doi: 
10.1016/j.rse.2014.06.005

moeller, A. K., P. m. luKACs, and J. s. 
horne. 2018. Three novel methods to 
estimate abundance of unmarked ani-
mals using remote cameras. Ecosphere 
9: e02331. doi: 10.1002/ecs2.2331

montGomery, r. A., K. m. redillA, r. J. 
moll, B. vAn moorter, C. m. 
rolAndsen, J. J. millsPAuGh, and e. J. 
solBerG. 2019. Movement modeling 
reveals the complex nature of the 
response of moose to ambient tempera-
tures during summer. Journal of 
Mammalogy 100: 169–177. doi: 
10.1093/jmammal/gyy185

murrAy, d. l., e. W. Cox, W. B. BAllArd, 
h. A. WhitlAW, m. s. lenArz, t. W. 
Custer, t. BArtnett, and t. K. fuller. 
2006. Pathogens, nutritional deficiency, 
and climate influences on a declining 
moose population. Wildlife Monographs 
166: 1–30. doi: 10.2193/0084-0173 
(2006)166[1:PNDACI]2.0.CO;2

musAnte, A. r., P. J. PeKins, and D. L. 
sCArPitti. 2010. Characteristics and 
dynamics of a regional moose Alces 
alces population in the northeastern 

United States. Wildlife Biology 16: 
185–204. doi: 10.2981/09-014

nAKAshimA, y., s. honGo, and e. f. AKomo-
oKoue. 2020. Landscape-scale estima-
tion of forest ungulate density and 
biomass using camera traps: applying 
the REST model. Biological 
Conservation 241: 108381. doi: 
10.1016/j.biocon.2019.108381

nyGrén, t., and M. Pesonen. 1993. The 
moose population (Alces alces L.) and 
methods of moose management in 
Finland, 1975–89. Finnish Game 
Research: 45: 45–53.

oyster, J. h., i. n. Keren, s. J. K. hAnsen, 
and R. B. hArris. 2018. Hierarchical 
mark-recapture distance sampling to 
estimate moose abundance. Journal of 
Wildlife Management 82: 1668–1679. 
doi: 10.1002/jwmg.21541

PeeK, J. A. 1974. Initial response of moose to 
a forest fire in northeastern Minnesota. 
The American Midland Naturalist 91: 
435–438. doi: 10.2307/ 2424334

PeKins, P. J. 2020. Metabolic and popula-
tion effects of winter tick infestations 
on moose: unique evolutionary cir-
cumstances? Frontiers in Ecology and 
Evolution 8: 1–13. doi: 10.3389/
fevo.2020.00176

Peters, W. m. heBBleWhite, K. G. smith, s. 
m. WeBB, n. WeBB, m. russell, C. 
stAmBAuGh, and r. B. Anderson. 2014. 
Contrasting aerial moose population 
estimation methods and evaluating 
sightability in west-central Alberta, 
Canada. Wildlife Society Bulletin 38: 
639–649. doi: 10.1002/wsb.433

Pfeffer, s. e., r. sPitzer, A. m. Allen, t. 
r. hofmeester, G. eriCsson, f. Widemo, 
n. J. sinGh, and J. P. G. m. CromsiGt. 
2018. Pictures or pellets? Comparing 
camera trapping and dung counts as 
methods for estimating population den-
sities of ungulates. Remote Sensing in 
Ecology and Conservation 4: 173–183. 
doi: 10.1002/rse2.67



MONITORING MOOSE POPULATIONS – MOLL ET AL. ALCES VOL. 58, 2022

48

PriAdKA, P., G. s. BroWn, B. r. PAtterson, 
and f. f. mAllory. 2020. Sex and 
age-specific differences in the perfor-
mance of harvest indices as proxies of 
population abundance under selective 
harvesting. Wildlife Biology 2020: 
1–11.

remPel, r. s., P. C. elKie, A. r. rodGers, 
and M. J. GluCK. 1997. Timber-
management and natural-disturbance 
effects on moose habitat: landscape 
evaluation. The Journal of Wildlife 
Management 61: 517–524. doi: 
10.2307/3802610

rotA, C. t., r. J. fletCher, r. m. dorAzio, and 
m. G. Betts. 2009. Occupancy estimation 
and the closure assumption. Journal of 
Applied Ecology 46: 1173–1181. doi: 
10.1111/j.1365-2664.\2009.01734.x

sAmPle, C., r. v. reA, and G. hesse. 2020. 
Tracking moose- and deer-vehicle colli-
sions using GPS and landmark inven-
tory systems in British Columbia. Alces 
56: 49–61.

sAsse, d. B. 2003. Job-related mortality of 
wildlife workers in the United States, 
1937–2000. Wildlife Society Bulletin 
31: 1015–1020.

sChAuB, m., and F. ABAdi. 2011. Integrated 
population models: a novel analysis 
framework for deeper insights into pop-
ulation dynamics. Journal of Ornithology 
152: S227–S237. doi: 10.1007/s10336- 
010-0632-7

_____, and M. Kéry. 2021. Integrated 
 population models: theory and ecolog-
ical applications with R and 
JAGS. Academic Press, Cambridge, 
Massachusetts, USA.

silverBerG, J. K., P. J. PeKins, and r. A. 
roBertson. 2003. Moose responses to 
wildlife viewing and traffic stimuli. 
Alces 39: 153–160.

silvy, N. J., editor. 2012. The Wildlife 
Techniques Manual: Management. The 
Johns Hopkins University Press, 
Baltimore, Maryland, USA.

sinClAir, A. R. E., J. M. fryxell, and G. 
CAuGhley. 2006. Wildlife Ecology, 
Conservation, and Management. 2nd 
edition. Blackwell Publishing, Malden, 
Massachusetts, USA.

sinGh, n. J., and e. J. milner-GullAnd. 
2011. Monitoring ungulates in Central 
Asia: current constraints and future 
potential. Oryx 45: 38–49. doi: 10.1017/
S0030605310000839

sKAlsKi, J. R., K. E. rydinG, and J. J. 
millsPAuGh. 2005. Wildlife 
Demography: Analysis of Sex, Age, and 
Count Data. Elsevier Academic Press, 
New York, New York, USA.

solBerG, E. J., B.-E. sAethert, o. strAnd, 
and A. loison. 1999. Dynamics of a 
harvested moose population in a vari-
able environment. Journal of Animal 
Ecology 68: 186–204. doi: 
10.1046/j.1365-2656.1999.00275.x

steenWeG, r., m. heBBleWhite, R. KAys, J. 
AhumAdA, J. t. fisher, C. Burton, s. e. 
toWnsend, C. CArBone, J. m. roWCliffe, 
J. WhittinGton, J. Brodie, J. A. royle, 
A. sWitAlsKi, A. P. ClevenGer, n. heim, 
and l. n. riCh. 2017. Scaling-up camera 
traps: monitoring the planet’s biodiver-
sity with networks of remote sensors. 
Frontiers in Ecology and the 
Environment 15: 26–34. doi: 10.1002/
fee.1448

steidl, r. J., J. P. hAyes, and E. sChAuBer. 
1997. Statistical power analysis in wild-
life research. The Journal of Wildlife 
Management 61: 270–279. doi: 
10.2307/3802582

stePhens, P. A., O. Y. zAumyslovA, D. G. 
miquelle, A. i. myslenKov, And G. d. 
hAyWArd. 2006. Estimating population 
density from indirect sign: track counts 
and the Formozov-Malyshev-Pereleshin 
formula. Animal Conservation 9: 339–
348. doi: 10.1111/ j.1469-1795.2006. 
00044.x

storAAs, t., h. Gundersen, H. henriKsen, 
and H. AndreAssen. 2001. The economic 



ALCES VOL. 58, 2022 MONITORING MOOSE POPULATIONS – MOLL ET AL.

49

value of moose – a review. Alces 37: 
9–107.

tAPe, K. d., d. d. Gustine, r. W. ruess, l. 
G. AdAms, and J. A. ClArK. 2016. Range 
expansion of moose in arctic Alaska 
linked to warming and increased shrub 
habitat. PLoS One 11: 1–12. doi: 
10.1371/journal.pone.0160049

timmermAnn, H. R. 1993. Use of aerial sur-
veys for estimating and monitoring 
moose populations – a review. Alces 29: 
35–46.

_____, and M. E. Buss. 2007. Population 
and harvest management. Pages 559–
616 in A. W. Franzmann and C. C. 
Schwartz, editors. Ecology and 
Management of the North American 
Moose. 2nd edition. University Press of 
Colorado, Boulder, Colorado, USA.

_____, and A. R. rodGers. 2017. The status 
and management of moose in North 
America – circa 2015. Alces 53: 1–22.

tyre, A. J., B. tenhumBerG, S. A. field, D. 
nieJAlKe, K. PArris, and H. P. 
PossinGhAm. 2003. Improving precision 
and reducing bias in biological surveys: 
estimating false-negative error rates. 
Ecological Applications 13: 1790–1801. 
doi: 10.1890/02-5078

ueno, m., e. J. solBerG, H. iiJimA, C. m. 
rolAndsen, and L. E. GAnGsei. 2014. 
Performance of hunting statistics as 
spatiotemporal density indices of 

moose (Alces alces) in Norway. 
Ecosphere 5: art13. doi: 10.1890/
ES13-00083.1

vAn BAllenBerGhe, v., and W. B. BAllArd. 
2007. Population dynamics. Pages 223–
246 in A. W. Franzmann and C. C. 
Schwartz, editors. Ecology and 
Management of the North American 
Moose. 2nd edition. University Press of 
Colorado, Boulder, Colorado, USA.

ver hoef, J. m. 2008. Spatial methods for 
plot-based sampling of wildlife popula-
tions. Environmental and Ecological 
Statistics 15: 3–13. doi: 10.1007/
s10651-007-0035-y

White, e. R. 2019. Minimum time required 
to detect population trends: the need for 
long-term monitoring programs. 
BioScience 69: 26–39. doi: 10.7287/
peerj.preprints.3168v4

WitCzuK, J., S. PAGACz, A. zmArz, and M. 
CyPel. 2018. Exploring the feasibility of 
unmanned aerial vehicles and thermal 
imaging for ungulate surveys in forests – 
preliminary results. International Journal 
of Remote Sensing 39: 5504–5521. doi: 
10.1080/01431161.2017.1390621

ziPKin, e. f., and S. P. sAunders. 2018. 
Synthesizing multiple data types for bio-
logical conservation using integrated 
population models. Biological 
Conservation 217: 240–250. doi: 
10.1016/j.biocon.2017.10.017