Content uploaded by Sophie Louise Gilbert
Author content
All content in this area was uploaded by Sophie Louise Gilbert on Dec 21, 2019
Content may be subject to copyright.
241
© The Ecological Society of America www.frontiersinecology.org
Unmanned aircraft systems (UAS) are emerging as
powerful tools in wildlife ecology, and can provide
novel remote- sensing data at fine spatial and temporal
scales (Anderson and Gatson 2013). Applications of
UAS technology are diverse and growing, ranging from
sampling airborne microbes, to locating wildlife poachers,
to providing data on cetacean behavior and body condi-
tion. As the technology and regulatory frameworks
improve, research applications are diversifying rapidly,
and studies incorporating this technology are likely to
proliferate in the future.
For a number of applications, UAS may increasingly
replace manned fixed- wing aircraft and helicopters,
which are popular tools for surveying animals and plants
for research, conservation, and management purposes.
While effective for covering large areas, manned aircraft
are also expensive, disturb wildlife, and are the leading
cause of work- related deaths among biologists (Sasse
2003; Wiegmann and Taneja 2003; Watts et al. 2010).
Recent technological advances in UAS, combined with
increasingly sophisticated remote- sensing equipment, are
facilitating ecological research that may be safer, more
cost- effective, and less invasive than traditional methods
(Figure 1; Anderson and Gaston 2013).
The tendency of fixed- wing airplanes and helicopters
to disturb wildlife is well- known (Andersen et al. 1989;
Bleich et al. 1994; Delaney et al. 1999; Giese and Riddle
1999; Richardson 2002). Most wildlife researchers use
small multicopter or fixed- wing UAS due to their afforda-
bility and maneuverability (Figure 1), and these small
UAS are considerably quieter than manned aircraft and
in general appear to cause minimal disturbances to wild-
life if operated responsibly (Figure 2; Sardà- Palomera
et al. 2012). However, accounts of UAS disturbing big-
horn sheep (Ovis canadensis) in National Parks within the
US have spurred a nationwide ban on the use of UAS by
the US National Park Service (NPS; NPS 2014), raising
broader concerns about wildlife photographers and
enthusiasts who widely adopt UAS without proper train-
ing or regard for potential impacts on animals.
Unmanned aircraft systems in wildlife
research: current and future applications of
a transformative technology
Katherine S Christie1*, Sophie L Gilbert1, Casey L Brown1, Michael Hatfield2, and Leanne Hanson3
Unmanned aircraft systems (UAS) – also called unmanned aerial vehicles (UAVs) or drones – are an emerging
tool that may provide a safer, more cost- effective, and quieter alternative to traditional research methods. We
review examples where UAS have been used to document wildlife abundance, behavior, and habitat, and
illustrate the strengths and weaknesses of this technology with two case studies. We summarize research on
behavioral responses of wildlife to UAS, and discuss the need to understand how recreational and commercial
applications of this technology could disturb certain species. Currently, the widespread implementation of
UAS by scientists is limited by flight range, regulatory frameworks, and a lack of validation. UAS are most
effective when used to examine smaller areas close to their launch sites, whereas manned aircraft are recom-
mended for surveying greater distances. The growing demand for UAS in research and industry is driving
rapid regulatory and technological progress, which in turn will make them more accessible and effective as
analytical tools.
Front Ecol Environ 2016; 14(5): 241–251, doi:10.1002/fee.1281
REVIEWS REVIEWS REVIEWS
1The Institute of Arctic Biology, University of Alaska
Fairbanks, Fairbanks, AK *(katiec@alaskasealife.org); 2The
Geophysical Institute, University of Alaska Fairbanks,
Fairbanks, AK; 3US Geological Survey, Fort Collins Science
Center, Fort Collins, CO
In a nutshell:
• Unmanned aircraft systems (UAS) are becoming increas-
ingly common in wildlife research and may be less expensive,
quieter, and safer than traditional manned aircraft
• Most studies we reviewed recorded minimal or no visible
behavioral responses to UAS; however, UAS are capable of
causing behavioral and physiological responses in wildlife
when observing at close range
• In some cases, UAS can replace traditional surveys of wild-
life and provide data with high levels of accuracy
• UAS are best used in studies where they can be deployed
from nearby platforms to cover small areas, and are not well-
suited for surveys of large areas
• Additional technological advances, combined with a more
streamlined regulatory process, will likely transform the way
we collect ecological information in the future
242
Unmanned aircraft systems and wildlife KS Christie et al.
www.frontiersinecology.org © The Ecological Society of America
Small UAS have the advantage of being cost- effective,
fuel- efficient, and able to access dangerous or inhospita-
ble areas. These types of UAS are either battery or fuel
powered, and – due to their limited size – have lower
power requirements than manned aircraft. As a result,
small UAS operate at a fraction of the cost of manned
aircraft but with greatly reduced flight ranges (Hodgson
et al. 2013). Nevertheless, scientists are able to conduct
repeat surveys using UAS, which allow for more accurate
population estimates (Sardà- Palomera et al. 2012).
Repeat surveys are further facilitated by the ability of
UAS to consistently follow precise, predetermined flight
paths (Watts et al. 2010). Moreover, their small size and
the absence of a human pilot and observer onboard allow
UAS to fly at low altitudes over dangerous areas such as
islands, rough waters, and regions where illegal poaching
or logging occurs (Koski et al. 2009; Sardà- Palomera et al.
2012; Mulero- Pázmány et al. 2014).
The use of remote- sensing equipment mounted on
UAS can increase the precision and accuracy of esti-
mates of wildlife population size. For instance, thermal
cameras detect animals based on their body heat and
have the advantage of identifying animals that are not
easily visible to the naked eye. UAS equipped with
thermal cameras could be extremely useful for detecting
cryptic nocturnal species such as owls and felids, and
this technology has been used successfully to detect
and estimate abundance of deer (Odocoileus virginianus
[Potvin and Breton 2005; Kissell and Nimmo 2011]
and Capreolus capreolus [Israel 2011]) and caribou
(Rangifer tarandus [Carr et al. 2010]).
In addition to detecting cryptic spe-
cies, such remote- sensing approaches
minimize observer fatigue and pro-
duce a permanent record of the data,
which can be reviewed multiple
times for quality- control purposes
(Hodgson et al. 2013). Similar to
having a security camera record crim-
inal activity, a permanent recording
of an ecological or wildlife survey
provides an objective, enduring
record of the organism of interest
for future reference, data sharing, and
further analysis. However, as with
other data- intensive methods, stand-
ardized metadata and long- term
archiving will be crucial to main-
taining the usefulness of such records.
In many cases, the advantages of
using small UAS for wildlife research
are outweighed by major limitations
in flight range due to both techno-
logical and legal factors. Current
Federal Aviation Administration
(FAA) regulations as well as battery
life markedly restrict the distance
that a UAS is able to cover, particularly for multirotor
UAS (Watts et al. 2010; Anderson and Gaston 2013;
Hodgson et al. 2013; Mulero- Pázmány et al. 2014).
Medium and large UAS have greater fuel capacity, but
the cost is likely to be prohibitive for most scientific
researchers. Additionally, technological limitations,
data- processing time, and uncertainty as to how UAS
affect wildlife may also pose obstacles to their widespread
use by scientists.
In this paper, we synthesize published and original
research featuring novel applications of UAS in wildlife
research, and weigh the advantages and disadvantages of
this technology. We describe how UAS have been used
to gather information on terrestrial and marine wildlife
and their habitats; additional information on fisheries
research and monitoring for conservation and manage-
ment is provided in WebPanel 1. Further, we compare
the cost, sound production, and operating range of differ-
ent manned and unmanned vehicles using data from the
Alaska Center for Unmanned Aircraft Systems
Integration (ACUASI), the US Geological Survey
(USGS), and the National Oceanic and Atmospheric
Administration (NOAA).
JComparisons of different models of UAS and
manned aircraft
Due to their diverse sizes and payloads, UAS vary
substantially in speed, range, cost of operation, and
noise emissions. For example, at an altitude of 100 m,
Figure 1. A Ptarmigan multicopter unmanned aircraft system (UAS) equipped
with a stabilized high- resolution camera, used for USGS vessel- based surveys of
walruses.
ACUASI
243
KS Christie et al. Unmanned aircraft systems and wildlife
© The Ecological Society of America www.frontiersinecology.org
a large UAS (the US Army Shadow
fixed- wing) produced sound levels of
84 dBA (A- weighted decibels); a
medium UAS (the ScanEagle fixed-
wing) produced 66 dBA; and a small
UAS (the Aeryon Scout quadcopter
and the Raven fixed- wing) produced
55 dBA and 50 dBA, respectively
(Figure 2). By comparison, helicopter
noise levels were approximately 95
dBA, and manned fixed- wing aircraft
varied from 75–88 dBA, depending
on the aircraft. The cost to purchase
a small multicopter UAS can
range from $300 for a basic, com-
mercially available model to more
than $100,000 for a custom model
designed for research (WebTable 1).
Fixed- wing UAS costs extend
from $2000 for smaller models to
$16.9 million for large, long- range,
military- grade models. Manned air-
craft cost upwards of $60,000 to
purchase, and hourly rates vary
between $600–3000 per hour,
depending on type (helicopter or
fixed- wing) and size (WebTable 1). With the exception
of the large US Army Shadow, UAS were much slower
(50–204 kilometers per hour [kph]) than manned air-
craft (223–332 kph). UAS varied in survey time from
0.5 hours for a small multicopter UAS to 24 hours
for the large US Army Shadow UAS (WebTable 1).
Manned aircraft could remain airborne for 2–7 hours
(WebTable 1). In terms of distance, the operating range
of UAS and manned aircraft varied from 3 to 6000
km and from 439 to 2112 km, respectively. UAS could
carry payloads of 0.5 kg (small multicopter) to 907 kg
(large NASA Ikhana), whereas manned aircraft could
carry heavier payloads of 760 to 2118 kg (WebTable 1).
JTerrestrial wildlife
Researchers interested in the ecology of terrestrial ani-
mals have long relied on aerial surveys to quantify
their abundance, distribution, and habitat. Recently,
UAS have been used to carry out these key functions,
and to capture data that were previously difficult to
collect using manned aircraft. Count- based estimates
of abundance have been obtained by UAS for water-
birds at wildlife refuges (USGS 2014), white pelican
(Pelecanus erythrorhynchos) breeding colonies (USGS
2014), sandhill crane (Grus canadensis) migratory stop-
over sites (USGS 2011), snow goose (Chen caerulescens)
and Canada goose (Branta canadensis) flocks (Chabot
and Bird 2012), and greater sage- grouse (Centrocercus
urophasianus) lek sites (Hanson et al. 2014) (WebTable
2). In studies of common terns (Sterna hirundo) and
sandhill cranes, counts obtained from UAS were within
6% and 5% of counts from ground- based surveys,
respectively (USGS 2011). Another promising appli-
cation of UAS for avian ecology involves the charac-
terization of flight paths and habitat selection, as
demonstrated by Rodríguez et al. (2012), who equipped
foraging lesser kestrels (Falco naumanni) with Global
Positioning System (GPS) loggers, enabling UAS to
follow the GPS tracks in near- real time. Finally, UAS
are increasingly being relied upon to access nest sites.
Studies quantifying the reproductive success of birds
nesting above ground level are challenging, and often
involve human observers climbing to nest sites to
monitor egg and nestling survival. In a study of hooded
crows (Corvus cornix), the use of UAS resulted in lower
levels of disturbance as compared with traditional
climbing surveys (Wessensteiner et al. 2015). UAS
platforms used to study birds included both fixed- wing
and multicopter designs, and flight altitudes ranged
from 30–183 m, with variable airspeeds (range: 15–80
kph; WebTable 2). Sampling technology included still
and video imagery in visible wavelengths, as well as
two applications of infrared (IR) videography to count
sandhill cranes at roost sites overnight (L Hanson,
unpublished data) and to detect greater sage- grouse
during predawn in low light conditions (Hanson et al.
2014).
Overall, of the 13 UAS- based avian studies reviewed in
this paper (WebTable 2), seven collected behavioral
observations, with bird reactions to UAS ranging from
no response (n = 1 study; snow and Canada geese), to
Figure 2. Sound levels (dBA) of different models of UAS, fixed- wing aircraft, and
helicopters. For each type, sound levels were obtained as follows: for the Aeryon Scout
quadcopter (recorded by the authors using a Larson- Davis 831 decibel reader), for the
Raven UAS (obtained from USGS), for the Scan Eagle UAS (from Hodgson et al.
2013), for the Shadow UAS (from US Army 2004), and for all manned aircraft
(from FAA 1988). We adjusted sound levels at different altitudes, to a common
altitude of 100 m using this equation: 20 × log(altitude of measurement ÷ desired
altitude) + measured sound level. Sound levels were recorded in A- weighted decibels,
which reduce the decibel values of sounds at very low frequencies.
244
Unmanned aircraft systems and wildlife KS Christie et al.
www.frontiersinecology.org © The Ecological Society of America
minimal to no response (n = 5 studies; greater sage-
grouse, sandhill cranes, black- headed gulls, common
terns, mallard ducks, and flamingos), to moderate
response (n = 1 study; hooded crows). Most bird responses
to UAS appear to be transient, but more research is
needed to explicitly test the reactions of different species
to this technology, and to quantify non- visible but possi-
bly important stress responses. In addition, behavioral
responses in some studies varied by UAS flight character-
istics or time of day. For example, in one study testing
behavioral responses of waterbirds, UAS were able to
approach to within 4 m of birds without disturbing them
on 80% of flights, and birds reacted more strongly to
UAS approaching vertically than horizontally (Vas et al.
2015). In another study, migrating sandhill cranes were
not disturbed if flights occurred while the birds were
roosting but were temporarily disturbed if UAS
approached while the birds were loafing and feeding (L
Hanson, unpublished data).
Terrestrial mammals have also been effectively sur-
veyed with UAS. Abundance and distribution surveys
have been conducted for elk (Cervus elaphus; USGS
2014), deer (Cervus spp, Dama dama; Israel 2011;
Barasona et al. 2014), orangutans (Pongo pygmaeus; Koh
and Wich 2012; Van Andel et al. 2015), elephants
(Loxodonta africana; Vermeulen et al. 2013), and rhinoc-
eros (Diceros bicornis and Ceratotherium simum; Mulero-
Pázmány et al. 2014) (WebTable 2). Fixed- wing UAS
platform designs predominated, likely because of the
extensive home ranges associated with these large mam-
malian species. Altitude of surveys varied considerably
(30–183 m), and sampling included digital and IR wave-
lengths in both video and still format (WebTable 2).
Similar to birds, mammals appeared to show minimal
behavioral responses to UAS, although experiments are
needed to explicitly test this for a variety of species. Of
the six studies that directly surveyed terrestrial mammals
(rather than focusing on habitat), three included behav-
ioral responses in their results (WebTable 2), and these
ranged from no response (n = 2; elephants and rhinos) to
moderate to high response (n = 1; black bears [Ursus
americanus]). Importantly, black bears often exhibited
minimal visible response to UAS, but still had elevated
heart rates, indicating that physiological stress responses
may occur without a visible behavioral cue (Ditmer et al.
2015). Given the interest within the wildlife research
community to develop alternative survey methods for
ungulates, we tested for UAS- induced behavioral changes
in captive caribou and semi- domesticated reindeer (both
Rangifer tarandus; hereafter “caribou”). Animals were
exposed to an Aeryon Scout quadcopter flying at an alti-
tude of 60 m (WebTable 2). Overflights lasted approxi-
mately 30 seconds to 2 minutes, and each animal was
exposed to a maximum of two overflights. Scan samples
(15 total) conducted on 33 caribou before and during
UAS flights indicated that the animals did not change
their activity patterns when exposed to a UAS flying
overhead (tall < 1.2, Pall > 0.25; Figure 3). However,
because these caribou were in a captive setting, and had
been exposed to anthropogenic noise, they may have
been less sensitive to UAS noise as compared with wild
caribou. Nevertheless, our data suggest that caribou in
the wild would either (1) not respond behaviorally to
UAS overflights, or (2) respond initially but habituate to
such flights.
JMarine wildlife
Marine species are notoriously difficult to study, and
manned aircraft have played a key role in investigations
of their distributions, movements, abundance, and body
condition. Advances in UAS technology have made
it possible to successfully survey marine mammals at
primary feeding, birthing, and haul- out areas up to
150 km from shore (Koski et al. 2009). Similar to the
constraints of manned aerial surveys, detection of mam-
mals by UAS surveys is strongly dependent on wave
conditions and the color of the animal, and is max-
imized by high- resolution imagery (Koski et al. 2009).
Dugongs (Dugong dugon), sperm whales (Physeter mac-
rocephalus), killer whales (Orcinus orca), and bowhead
whales (Balaena mysticetus) have been successfully sur-
veyed using fixed- wing and multicopter UAS fitted with
high- resolution digital cameras (WebTable 2; Hodgson
et al. 2013; NOAA 2014a; Durban et al. 2015; Koski
et al. 2015). In addition to counting marine mammals,
useful information on body condition, age, and sex
can be obtained. For instance, NOAA scientists used
a multicopter equipped with a digital camera to pho-
tograph and later identify individual resident killer
whales, while simultaneously quantifying their body size
and diagnosing pregnancies (Figure 4a; NOAA 2014b).
The multicopter hovered 30 m above the whales with-
out disturbing them, and provided greater resolution
for measuring body condition and length as compared
with traditional helicopter surveys (Fearnbach et al.
2011; Durban et al. 2015). In addition, a fixed- wing
UAS was used successfully to photograph and later
identify individual bowhead whales without causing any
observable disturbance to the animals (Koski et al.
2015). Likewise, in photo- identification studies, the
slower speed of the UAS as compared with manned
aircraft facilitates the capture of high- resolution images
with less blur (Koski et al. 2015). Finally, Bevan et al.
(2015) were able to locate and monitor hatchling and
adult sea turtles with the aid of a GoPro camera
mounted on a small quadcopter UAS operating at
heights of 30–50 m, and recorded no behavioral responses
to the UAS among the turtles.
Given their low noise production and ability to access
remote, dangerous locations, UAS may be particularly
useful for surveying marine wildlife at haul- out sites
and breeding colonies. There is much concern about the
tendency of hauled- out marine mammals to stampede or
245
KS Christie et al. Unmanned aircraft systems and wildlife
© The Ecological Society of America www.frontiersinecology.org
otherwise move into the water when disturbed by a
low- flying fixed- wing aircraft (Born et al. 1999; Udevitz
et al. 2013). Walruses (Odobenus rosmarus divergens) are
particularly prone to stampede, and large numbers have
been killed at haul- out sites during stampede events in
recent years (Udevitz et al. 2013). UAS produce substan-
tially less sound (50 dBA for a small fixed- wing UAS at
an altitude of 100 m) than manned aircraft (75 dBA at
the same altitude; Figure 2), and may therefore circum-
vent this problem (Hodgson et al. 2013). Moreland et al.
(2015) surveyed spotted (Phoca largha) and ribbon
(Histriophoca fasciata) seals at 122 m using digital single-
lens reflex cameras, and found little to no behavioral
response. In contrast, Pomeroy et al. (2015) reported that
gray (Halichoerus grypus) and harbor (Phoca vitulina) seals
exhibited varied, moderate responses to UAS depending
on season, reproductive status, and UAS survey heights
(ranging from 5–250 m). Steller sea lions (Eumetopias
jubatus) were surveyed and photographed at 45 m with a
multicopter in the outer Aleutian Islands, and observed
behavioral responses were negligible or absent (Table 1).
In a study of foraging behavior, sea otters (Enhydra lutris)
were surveyed using a small multicopter, with little or no
discernable behavioral response to the UAS (Figure 4b;
UAF 2015). In a study by Goebel et al. (2015), penguins
(Pygoscelis papua and Pygoscelis antarctica) and their
chicks were identified and counted using high- resolution
georeferenced mosaic images taken by a multicopter
UAS flying at 60- m altitude; UAS- derived counts were
within 5% of traditional ground- based surveys, and the
penguins were not disturbed. Overall, of the 13 studies
that we found on marine wildlife, seven examined behav-
ioral reactions, which ranged from no response (n = 4;
killer whales, bowhead whales, leopard [Hydrurga lep-
tonyx] and fur [Arctocephalus gazella] seals, penguins) to
minimal to no response (n = 2; Steller sea lion, ribbon
and spotted seals), to moderate response (n = 1; gray and
harbor seals).
JSpatial ecology
Advances in technology – such as higher payload
capacity of small UAS and miniaturization of mul-
tispectral and hyperspectral sensors in conjunction
with improved computer- processing capabilities – have
allowed practitioners to monitor habitat for fish and
wildlife species. Data collected through traditional
remote- sensing techniques (eg manned aircraft or
satellite) are often too coarse in resolution to suit
fine- scale ecological studies (Wulder et al. 2004).
Commercially operated satellite sensors can now
ACUASI
ACUASI
(a) (b)
Figure 3. Infrared images taken with an Aeryon Scout UAS of
caribou (a) in an open field and (b) in a patch of forest. Also
shown are observations of caribou behavior before and during
UAS flights (c). A total of 15 scan samples were collected from
33 captive animals at the Large Animal Research Station at the
University of Alaska Fairbanks. Error bars denote standard
errors. No statistically significant difference in behavior was
detected between UAS and no- UAS time periods.
(c)
246
Unmanned aircraft systems and wildlife KS Christie et al.
www.frontiersinecology.org © The Ecological Society of America
produce data at finer resolutions; however, operational
constraints include prohibitively high costs associated
with acquiring images, cloud contamination of regions
of interest, and the inability to repeat measurements
over required timescales (Loarie et al. 2007). It has
been argued that UAS equipped with remote- sensing
payloads (eg RGB cameras, color IR sensors, and light-
weight thermal systems) may help to resolve issues
associated with spatial ecology (Anderson and Gaston
2013). Small UAS can hover at lower altitudes to
capture fine- scale habitat metrics such as forest canopy
gaps and understory plant diversity (Getzin et al. 2012).
Additionally, the flexible maneuverability of UAS may
circumvent issues associated with cloud contamination
and minimize time between site revisits (Herwitz et al.
2004).
Fine- resolution remotely sensed data obtained via
UAS have been used to quantify habitat characteris-
tics in a number of studies. For example, habitats of
wetland birds, including the US federally listed Yuma
clapper rail (Rallus longirostris yumanensis) and south-
western willow flycatcher (Empidonax traillii extimus),
were mapped using color IR to capture NDVI
(normalized- difference vegetation index) and subse-
quently classify vegetation (Figure 5; USGS 2014).
Sage- brush habitat for another endangered species, the
pygmy rabbit (Brachylagus idahoensis), was also success-
fully mapped with a UAS using visible- spectrum digi-
tal still photography (WebTable 2; Breckenridge et al.
2011; Levy 2011). Imagery obtained via UAS has also
been applied to delineate localized cover types, esti-
mate percentage of bare ground (Breckenridge et al.
2011), catalog forest composition (Dunford et al.
2009), and calculate leaf area index and chlorophyll
content (Figure 5; Berni et al. 2009; McGwire et al.
2013). Furthermore, UAS equipped with IR or high-
resolution cameras have been used to monitor the dis-
tributions of invasive species (Zaman et al. 2011; Wan
et al. 2014), produce vegetation maps (Laliberte et al.
2011), and identify forest canopy mortality (Dunford
et al. 2009).
JCase studies
To illustrate the advantages and disadvantages of UAS
technology in specific settings, we present two case
studies. The first study documents the use of an APH-
22 hexacopter to survey sea lion haul- outs and rook-
eries in the outer Aleutian Islands in Alaska (Table 1).
The UAS was used to survey areas that were oth-
erwise inaccessible by traditional survey methods (Twin
Otter fixed- wing plane) due to inclement weather,
remoteness, and a lack of suitable landing sites, and
resulted in the most comprehensive survey of Steller
sea lions in the Aleutians since the 1970s (K Sweeney,
pers comm). Researchers launched the UAS from a
vessel that was close (<1 km) to rookeries, and were
able to capture high- resolution imagery of individual
animals while causing minimal disturbance (Table 1).
After an initial investment of $25,000 to purchase
the UAS, the cost of the UAS program consisted
primarily of operating the research vessel, and was
less than the cost of operating the Twin Otter, given
the multiple projects that shared the expense of run-
ning the vessel (Table 1). Although the UAS could
survey rookeries that were inaccessible by the Twin
Otter, the major disadvantage of the UAS was its
dependence on a research vessel and its limited range:
only 30 sites (400 km of coastline) were surveyed as
compared to 201 sites (2500 km of coastline) surveyed
by the Twin Otter over a similar time period.
In our second case study, researchers compared surveys
of roosting sandhill cranes – at Monte Vista National
Wildlife Refuge in Colorado – using a fixed- wing Raven
RQ- 11A UAS versus relying on ground- based surveys
(Table 2). The UAS surveyed the same 38- ha area as
Figure 4. Photographs taken from UAS demonstrating (a) a group of killer whales (platform: APH- 22 hexacopter, study by
NOAA) and (b) a foraging sea otter (platform: Aeryon Scout, study by the University of Alaska Coastal Marine Institute).
(a) (b)
NOAA
ACUASI
247
KS Christie et al. Unmanned aircraft systems and wildlife
© The Ecological Society of America www.frontiersinecology.org
ground surveys, with no observable effect on crane
behavior and a difference of 4.6% in accuracy between
survey methods. UAS surveys used half the number of
observers and therefore cost half as much to conduct. The
Raven was owned by the US military, so a UAS purchase
price was not factored into the cost estimates.
Disadvantages of the UAS included the inability to fly
during high winds or heavy rains, the requirement to fly
within line- of- sight, and a lengthy flight approval process
(Table 2).
JLimitations of UAS technology
Major limitations to the widespread adoption of UAS
include difficulties in obtaining permits for use, limited
survey range, and data- processing time. Many of the
small, battery- powered multicopters that are favored
due to their low costs, energy efficiency, and ease
of operation must be recharged or have batteries
replaced approximately every 20 minutes, thereby
restricting survey range. Larger UAS are capable of
longer flights and greater payloads but are prohibi-
tively expensive for most researchers (WebTable 1).
Furthermore, many small UAS cannot currently be
flown safely during severe weather conditions
(Weissensteiner et al. 2015).
Lengthy, complex permitting processes required by
national aviation authorities constrain UAS- based
ecological research in the US (Vincent et al. 2015) and
elsewhere around the world. In addition, many govern-
ment permitting frameworks require that the UAS be
operated within line- of- sight only (Watts et al. 2010;
Anderson and Gaston 2013; Hodgson et al. 2013; Hanson
et al. 2014; Mulero- Pázmány et al. 2014). This restriction
reduces the survey range of a UAS beyond the limitations
imposed by fuel or battery capacity (Table 1; Mulero-
Pázmány et al. 2014). Nevertheless, some restrictions are
necessary to ensure privacy, minimize the chance of air-
craft collisions, and avoid harassing wildlife. Already,
UAS have been used irresponsibly by civilians to
approach wild animals, necessitating the introduction of
strong regulations to protect wildlife from harassment in
Table 1. Case study 1: comparison of traditional and UAS surveys of Steller sea lions in Alaska
Manned aerial surveys UAS surveys
Purpose of surveys Estimate the abundance of Steller sea lions in the
inner Aleutians
Estimate the abundance of Steller sea lions
in the outer Aleutians
Cost per day $4700 per day including fuel and pilot, or $400 per
site
$3000 per day based on the cost of vessel
support, or $1700 per site
Type of aircraft NOAA Twin Otter APH- 22 hexacopter
Distance/area surveyed 2500 km of coastline, including the Gulf of Alaska
and part of Aleutians; 210 sites surveyed
400 km of coastline along the western
Aleutian chain, 30 sites surveyed; maximum
distance from the vessel was 634 m, longest
flight was 16 minutes
% animals detected 100% of hauled- out animals 100% of hauled- out animals
Data collected Quantitative imagery, animal counts Quantitative imagery, animal counts,
individual identification
Number of personnel 6 2
Observed effect on animal Slight and variable, 5% of adults moved toward
water
Very low to none, 0.3% of adults moved
toward water
Advantages (1) surveyed up to 50 sites per day
(2) high-quality images
(3) cost per site low
(1) surveyed remote sites with no airfields
(2) extremely low disturbance
(3) very high-quality images (flew at altitude
of 45 m)
(4) less subject to flight restrictions due to
weather
(5) biologists can double as pilots
Disadvantages (1) requires good weather at primary and alternate
airfields (minimum of 750-ft ceilings)
(2) relatively noisy
(3) may only fly on half (or less) of days available
(4) requires a runway for takeoff/landing
(5) imagery has lower resolution (flight altitude:
150–305 m)
(6) requires flight crew of 3 plus 3 observers
(1) can survey only a few (1–3) sites per day
(2) requires costly vessel for use as transport
(3) cannot fly in high winds (wind speed must
be less than 25 knots on the ground)
(4) must stay within line-of-sight and 0.8 km
of observer
Notes: Surveys were conducted by the National Oceanic and Atmospheric Administration.
248
Unmanned aircraft systems and wildlife KS Christie et al.
www.frontiersinecology.org © The Ecological Society of America
US National Parks. The NPS has issued a park- wide ban
on the use of UAS – with an exception for scientific
research conducted by park employees – due to visitor
complaints of UAS disturbing wildlife and creating
unwanted noise (NPS 2014). UAS operators have sug-
gested regulations that limit operations to areas of low
human density and aircraft traffic, or the designation of
UAS corridors (Mulero- Pázmány et al. 2014).
Another current limitation of UAS involves the pro-
cessing of large amounts of data generated by surveys.
Digital photos, video, and other remote- sensing data
often require a substantial time investment for data
organization and processing; however, automated pro-
grams have been developed to improve the efficiency of
this procedure (Groom et al. 2011; Dehvari and Heck
2012). In addition, the effectiveness of UAS in replicat-
ing results obtained by traditional surveys of wildlife is
currently being debated, and the accuracy and precision
of UAS- derived population estimates is being tested.
Promisingly, recent research on the effectiveness of UAS
in estimating wildlife population parameters (eg Koski
et al. 2009; USGS 2011; Martin et al. 2012; Goebel et al.
2015) shows that this technique can indeed be highly
accurate.
A key area for future research will be testing the effect
of UAS on behavioral and physiological responses of
different species of wildlife. Among the studies we ana-
lyzed, we found that UAS disturbed wildlife less than
traditional methods when direct comparisons were made,
although behavioral and physiological responses to UAS
occurred in some situations. Of particular concern are
species adapted to avian predators, as well as birds of
prey themselves, some of which have
been known to attack airborne UAS.
Though quieter than traditional
fixed- wing or rotor aircraft at compa-
rable distances, UAS often approach
animals more closely and therefore
may have a greater impact in certain
scenarios. Moreover, animals may
become more disturbed by the sudden
occurrence of noise from a UAS than
by a slowly approaching manned
aircraft, presenting another topic
that warrants further investigation.
J
Conclusions and future
directions
The burgeoning application of UAS
in ecological and wildlife studies
demonstrates that a growing number
of scientists are embracing this novel
technology to meet their needs. This
technology has been used successfully
to address a broad diversity of eco-
logical research and management
problems, and can be a cost- effective, safe, relatively
quiet, and effective alternative to traditional survey
techniques. Despite the advantages of conducting field
research with UAS, major obstacles to their widespread
adoption by ecologists include regulatory limitations,
data- processing time, and fuel capacity or battery life.
At this time, UAS are best suited to situations where
they can be launched from platforms or areas that are
relatively close to the target, and are not suited for
surveys of large areas.
The future utility of UAS for ecologists is expected to
be determined by the regulatory framework of the avia-
tion administrations within each country in which they
are operated, rather than by technological limitations
(Vincent et al. 2015). In most countries, UAS must be
operated within line- of- sight and lengthy permitting pro-
cesses are necessary (Anderson and Gaston 2013;
Linchant et al. 2015). However, regulations are changing;
in February 2015, the FAA released proposed guidelines
for civilian use of UAS in the US, requiring a single, cer-
tified operator rather than a trained UAS pilot (FAA
2015). In addition, administrations are considering regu-
lations that include UAS flights outside of line- of- sight,
an important step if they are to become widely used
by ecologists. Commercial pressure on aviation adminis-
trations is intense, and the UAS industry is expected
to expand to over 100,000 jobs by 2025, with an eco-
nomic impact of $82 billion (Jenkins and Vasigh 2013;
Tast 2015).
UAS will likely become increasingly popular in eco-
logical research, as technological improvements allow
long- distance, highly accurate flight trajectories and
Figure 5. A color infrared (IR) image of wetland vegetation used to identify habitat
for US federally listed wetland birds. Color IR images were used to generate NDVI
values, which in turn were used to classify habitat types.
USGS
249
KS Christie et al. Unmanned aircraft systems and wildlife
© The Ecological Society of America www.frontiersinecology.org
diverse payloads, including various forms of remote-
sensing equipment. Future applications for wildlife
ecology include detecting and monitoring nests, dens,
predator kill sites, and birth and mortality sites, as well
as relocating radio- or GPS- tagged wildlife. In addition,
UAS could potentially be harnessed to immobilize large-
bodied wildlife using UAS- fired tranquilizers, collect
biological samples (such as hair, breath, blood, scat,
and saliva), track the illegal trade in wildlife products
throughout the supply chain, and reduce wildlife–human
conflict through negative conditioning of so- called “prob-
lem” animals. If the above- described limitations are
overcome through further technological advances, a more
streamlined permitting process, and continued research
of the effects on wildlife, UAS have the potential to
transform the way we collect ecological information.
JAcknowledgements
We thank members of ACUASI for their expertise
and willingness to collaborate on this research. NOAA
personnel (J Durban, E Moreland, K Sweeney, and L
Fritz) and USGS personnel (J Sloan and M Bauer)
provided valuable information about their UAS research,
and NPS scientist D Betchkal provided the decibel
reader and advice on sound measurement. Any use of
trade, product, or firm names is for descriptive purposes
only and does not imply endorsement by the US
Government.
JReferences
Andersen DE, Rongstad OJ, and Mytton WR. 1989. Response
of nesting red- tailed hawks to helicopter overflights. Condor
91: 296–99.
Anderson K and Gaston KJ. 2013. Lightweight unmanned aerial
vehicles will revolutionize spatial ecology. Front Ecol Environ
11: 138–46.
Barasona JA, Mulero-Pázmány M, Acevedo P, et al. 2014.
Unmanned aircraft systems for studying spatial abundance
of ungulates: relevance to spatial epidemiology. PLoS ONE
9: e115608.
Berni JA, Zarco-Tejada PJ, Suarez L, and Fereres E. 2009.
Thermal and narrowband multispectral remote sensing for
vegetation monitoring from an unmanned aerial vehicle.
IEEE T GeoSci Remote; doi:10.1109/TGRS.2008.2010457.
Bevan E, Wibbels T, Najera BMZ, et al. 2015. Unmanned aerial
vehicles (UAVs) for monitoring sea turtles in near- shore
waters. Marine Turtle Newsletter 145: 19–22.
Bleich VC, Bowyer RT, Pauli AM, et al. 1994. Mountain
sheep (Ovis canadensis) and helicopter surveys: ramifications
for the conservation of large mammals. Biol Conserv 70:
1–7.
Born EW, Riget FF, Dietz R, et al. 1999. Escape responses of
hauled out ringed seals (Phoca hispida) to aircraft disturbance.
Polar Biol 21: 171–78.
Breckenridge RP, Dakins M, Bunting S, et al. 2011. Comparison
of unmanned aerial vehicle platforms for assessing vegetation
cover in sagebrush steppe ecosystems. Rangeland Ecol Manag
64: 521–32.
Carr NL, Rodgers AR, Kingston SR, et al. 2010. Comparative
woodland caribou population surveys in Slate Islands
Provincial Park, Ontario. Rangifer 20: 205–17.
Chabot D and Bird DM. 2012. Evaluation of an off- the- shelf
unmanned aircraft system for surveying flocks of geese.
Waterbirds 35: 170–74.
Table 2. Case study 2: comparison of ground- based population abundance estimates with small UAS surveys
of sandhill cranes in Colorado
Ground- based surveys UAS surveys
Purpose of surveys Estimate the one- night abundance of roosting
sandhill cranes at Monte Vista National Wildlife
Refuge, Colorado
Estimate the one- night abundance of roosting
sandhill cranes at Monte Vista National
Wildlife Refuge, Colorado
Cost per day $840.00 $400.00
Type of aircraft N/A – optical count from ground Raven RQ- 11A
Distance/area surveyed 37.75 hectares (93.28 acres) 37.75 hectares (93.28 acres)
% animals detected Overestimated number by 4.6% 100%
Data collected Animal counts Quantitative imagery, animal counts
Number of personnel 6 3
Observed effect on animal No discernible effect No discernible effect
Advantages (1) low technology
(2) easy to gather data – no computer required
(3) can collect data in all types of weather
(1) low disturbance
(2) imagery to verify counts
(3) biologist can be UAS pilot
Disadvantages (1) difficult to partition individual observer
viewshed – can lead to overestimate of
abundance
(2) view can be obstructed by vegetation and
lighting
(3) count cannot be validated with imagery
(1) cannot fly in high winds (ground wind
speed must be less than 25 knots)
(2) flights delayed by rain
(3) must stay within line-of-sight
(4) may take long time for flight approvals
Notes: Surveys were conducted by the US Fish and Wildlife Service and the US Geological Survey.
250
Unmanned aircraft systems and wildlife KS Christie et al.
www.frontiersinecology.org © The Ecological Society of America
Dehvari A and Heck RJ. 2012. Removing non- ground points
from automated photo- based DEM and evaluation of its
accuracy with LiDAR DEM. Comput Geosci 43: 108–17.
Delaney DK, Grubb TG, Beier P, et al. 1999. Effects of helicopter
noise on Mexican spotted owls. J Wildlife Manage 63: 60–76.
Ditmer MA, Vincent JB, Werden LK, et al. 2015. Bears show
a physiological but limited behavioral response to unmanned
aerial vehicles. Curr Biol 25: 2278–83.
Dunford R, Michel K, Gagnage M, et al. 2009. Potential and
constraints of unmanned aerial vehicle technology for the
characterization of Mediterranean riparian forest. Int J Remote
Sens 30: 4915–35.
Durban JW, Fearnbach H, Barrett-Lennard LG, et al. 2015.
Photogrammetry of killer whales using a small hexacopter
launched at sea. J Unmanned Veh Syst 3: 1–5.
FAA (Federal Aviation Administration). 1988. Appendices 7
and 11 noise levels for US certificated and foreign aircraft.
www.faa.gov/about/office_org/headquarters_offices/apl/noise_
emissions/aircraft_noise_levels. Viewed 12 Jan 2015.
FAA (Federal Aviation Administration). 2015. Overview of
small UAS notice of proposed rulemaking. www.faa.gov/
regulations_policies/rulemaking/media/021515_sUAS_
Summary.pdf. Viewed 24 Mar 2016.
Fearnbach H, Durban JW, Ellifrit DK, and Balcomb KC. 2011.
Size and long- term growth trends of endangered fish- eating
killer whales. Endangered Spec Res 13: 173–80.
Getzin S, Wiegand K, and Schoning I. 2012. Assessing biodi-
versity in forests using very high- resolution images and
unmanned aerial vehicles. Methods Ecol Evol 3: 397–404.
Giese M and Riddle M. 1999. Disturbance of emperor penguin
Aptenodytes forsteri chicks by helicopters. Polar Biol 22:
366–71.
Goebel ME, Perryman WL, Hinke JT, et al. 2015. A small
unmanned aerial system for estimating abundance of and size
of Antarctic predators. Pol Biol; doi:10.1007/s00300- 014- 1625- 4.
Groom G, Petersen IK, Anderson MD, and Fox AD. 2011.
Using object- based analysis of image data to count birds:
mapping of lesser flamingos at Kamfers Dam, Northern Cape,
South Africa. Int J Remote Sens 32: 4611–39.
Hanson L, Holmquist-Johnson CL, and Cowardin ML. 2014.
Evaluation of the Raven sUAS to detect and monitor greater
sage-grouse leks within the Middle Park population: US
Geological Survey Open-File Report 2014–1205. Reston,
VA: USGS. http://pubs.usgs.gov/of/2014/1205/pdf/ofr2014-
1205.pdf. Viewed 24 Mar 2016.
Herwitz SR, Johnson LF, Dunagan SE, et al. 2004. Imaging from
an unmanned aerial vehicle: agricultural surveillance and
decision support. Comput Electron Agr 44: 49–61.
Hodgson A, Kelly N, and Peel D. 2013. Unmanned aerial
vehicles (UAVs) for surveying marine fauna: a dugong case
study. PLoS ONE 8: e79556.
Israel M. 2011. A UAS- based roe deer fawn detection system.
Int Arch Photogramme Remote Sens Spatial Inf Sci 38: 1.
Jenkins D and Vasigh B. 2013. The economic impact of unmanned
aircraft systems integration in the United States. Association
for Unmanned Vehicle Systems International Economic Report.
Arlington, VA: AUVSI. http://robohub.org/_uploads/AUVSI_
New_Economic_Report_2013_Full.pdf. Viewed 11 Nov 2015.
Kissell RE and Nimmo SK. 2011. A technique to estimate
white- tailed deer Odocoileus virginianus density using vertical-
looking infrared imagery. Wildlife Biol 17: 85–92.
Koh LP and Wich SA. 2012. Dawn of drone ecology: low- cost
autonomous aerial vehicles for conservation. Trop Cons Sci
5: 121–32.
Koski WR, Allen T, Ireland D, et al. 2009. Evaluation of an
unmanned airborne system for monitoring marine mammals.
Aquat Mamm 35: 347–57.
Koski WR, Gamage G, Davis AR, et al. 2015. Evaluation of
UAS for photographic re- identification of bowhead whales,
Balaena mysticetus. J Unmanned Veh Syst 3: 22–29.
Laliberte AS, Winters C, and Rango A. 2011. UAS remote
sensing missions for rangeland applications. Geocarto Int 26:
141–56.
Levy S. 2011. Ravens hunting rabbits: unmanned aircraft help
protect pygmy rabbits. Unmanned Syst 29: 34–35.
Linchant J, Lisein J, Semeki J, et al. 2015. Are unmanned
aircraft systems (UASs) the future of wildlife monitoring?
A review of accomplishments and challenges. Mammal Rev
45: 239–52.
Loarie S, Joppa L, and Pimm S. 2007. Satellites miss environ-
mental priorities. Trends Ecol Evol 22: 630–32.
Martin J, Edwards HH, Burgess MA, et al. 2012. Estimating
distribution of hidden objects with drones: from tennis balls
to manatees. PLoS ONE 7: e38882.
McGwire KC, Weltz MA, Finzel JA, et al. 2013. Multiscale
assessment of green leaf cover in a semi- arid rangeland with
a small unmanned aerial vehicle. Int J Remote Sens 34:
1615–32.
Moreland EE, Cameron MF, Angliss RP, et al. 2015. Evaluation
of a ship- based unoccupied aircraft system (UAS) for surveys
of spotted and ribbon seals in the Bering Sea pack ice. J
Unmanned Veh Syst 3: 114–22.
Mulero-Pázmány M, Stolper R, van Essen LD, et al. 2014.
Remotely piloted aircraft systems as a rhinoceros anti- poaching
tool in Africa. PLoS ONE 9: e83873.
NOAA (National Oceanic and Atmospheric Administration).
2014a. Spying on sperm whales. www.nmfs.noaa.gov/sto-
ries/2013/03/3_07_13sperm_whales.html. Viewed 12 Jan 2015.
NOAA (National Oceanic and Atmospheric Administration).
2014b. Unmanned aerial vehicle offers a new view of killer
whales. www.fisheries.noaa.gov/podcasts/2014/10/aerial_
vehicle_killer_whale.html#.VLQdESsVhik. Viewed 12 Jan
2015.
NPS (US National Park Service). 2014. Unmanned aircraft to
be prohibited in America’s National Parks. Press release.
bit.ly/1UGBW2Z. Viewed 12 Jan 2015.
Pomeroy P, O’Connor L, and Davies P. 2015. Assessing use of
and reaction to unmanned aerial systems in gray and harbor
seals during breeding and molt in the UK. J Unmanned Veh
Syst 3: 102–13.
Potvin F and Breton L. 2005. Testing two aerial survey tech-
niques on deer in fenced enclosures – visual double counts
and thermal infrared sensing. Wildlife Soc B 33: 317–25.
Richardson WJ. 2002. Aircraft sound and disturbance to bowhead
and beluga whales during spring migration in the Alaskan
Beaufort Sea. Mar Mamm Sci 18: 309–35.
Rodríguez A, Negro JJ, Mulero M, et al. 2012. The eye in the
sky: combining use of unmanned aerial systems and GPS
data loggers for ecological research and conservation of small
birds. PLoS ONE 7: e50336.
Sardà-Palomera F, Bota G, Vinolo C, et al. 2012. Fine- scale
bird monitoring from light unmanned aircraft systems. Ibis
154: 177–83.
Sasse DB. 2003. Job- related mortality of wildlife workers in the
United States, 1937–2000. Wildlife Soc B 31: 1015–20.
Tast RW. 2015. Unmanned aerial systems: domestic statutory
issues. Nebraska Law Rev 93: 773–806.
UAF (University of Alaska Fairbanks). 2015. Researchers study
sea otters with unmanned aircraft. http://news.uaf.edu/
researchers-study-sea-otters-unmanned-aircraft. Viewed 8 Jan
2016.
Udevitz MS, Taylor RL, Garlich-Miller JL, et al. 2013. Potential
population- level effects of increased haulout- related mortal-
ity of Pacific walrus calves. Polar Biol 36: 291–98.
251
KS Christie et al. Unmanned aircraft systems and wildlife
© The Ecological Society of America www.frontiersinecology.org
US Army. 2004. Transformation of US Army Alaska Final
Environmental Impact Statement, Appendix F, Environmental
Consequences. www.wainwright.army.mil/env/NEPA/NEPA_
Transformation_EIS-ROD.html. Viewed 27 Mar 2016.
USGS (US Geological Survey). 2011. Cranes and drones: strange
airfellows? www.fort.usgs.gov/science-feature/127. Viewed 12
Jun 2014.
USGS (US Geological Survey). 2014. US Geological Survey
National Unmanned Aircraft Systems Project website. http://
rmgsc.cr.usgs.gov/UAS. Viewed 12 Jun 2014.
Van Andel AC, Wich SA, Boesch C, et al. 2015. Locating
chimpanzee nests and identifying fruiting trees with an
unmanned aerial vehicle. Am J Primatol; http://doi.
org/10.1002/ajp.22446.
Vas E, Lescroel A, Duriez O, et al. 2015. Approaching birds
with drones: first experiments and ethical guidelines. Biol
Lett 11: 20140754.
Vermeulen C, Lejeune P, Lisein J, et al. 2013. Unmanned aerial
survey of elephants. PLoS ONE 8: e54700.
Vincent JB, Werden LK, and Ditmer MA. 2015. Barriers to
adding UAVs to the ecologist’s toolbox. Front Ecol Environ
13: 74–75.
Wan H, Wang Q, Jiang D, et al. 2014. Monitoring the invasion
of Spartina alterniflora using very high resolution unmanned
aerial vehicles imagery in Beihai, Guangxi (China). Sci World
J; http://dx.doi.org/10.1155/2014/638296.
Watts AC, Perry JH, Smith SE, et al. 2010. Small unmanned
aircraft systems for low- altitude aerial surveys. J Wildlife
Manage 74: 1614–19.
Weissensteiner MH, Poelstra JW, and Wolf JBW. 2015. Low-
budget ready- to- fly unmanned aerial vehicles: an effective
tool for evaluating the nesting status of canopy- breeding
bird species. J Avian Biol 46: 425–30.
Wiegmann DA and Taneja N. 2003. Analysis of injuries among
pilots involved in fatal general aviation airplane accidents.
Accident Annal Prev 35: 571–77.
Wulder MA, Hall RJ, Coops NC, and Franklin SE. 2004. High
spatial resolution remotely sensed data for ecosystem char-
acterization. BioScience 54: 511–21.
Zaman B, Jensen AM, and McKee M. 2011. Use of high-res-
olution multispectral imagery acquired with an autonomous
unmanned aerial vehicle to quantify the spread of an inva-
sive wetlands species. Geoscience and Remote Sensing
Symposium 2011 IEEE International; 24–29 Jul 2011;
Vancouver, Canada. doi:10.1109/IGARSS.2011.6049252.
JSupporting Information
Additional, web-only material may be found in the
online version of this article at http://onlinelibrary.
wiley.com/doi/10.1002/fee.1281/suppinfo
