A new arboreal suspension technique to strengthen the understanding of canopy ecological interactions and dynamics

Abstract

Recent studies in forest canopies have revealed new findings on species interactions and diversity; hitherto poorly understood. This glaring gap results from the challenges of recording arboreal and scansorial species and their interactions in the vertical strata. We present a new methodology that allows installing camera-traps at multiple forest strata to detect species and record ecological interactions. We have developed wooden support to place the environmental monitoring equipment at the forest canopy, evaluating and testing its cost-effectiveness by monitoring six fruiting plant species across a triple ecotone of the Cerrado, Amazon, and Pantanal biomes. Our structure to install camera-traps in the forest canopy had a total cost of US50.00,whereasatree−climbingkitreachesUS 50.00, whereas a tree-climbing kit reaches US 1,248.00. After 11 months of using camera-traps coupled with our wood structure, we recorded 137 mammal-plant interaction events and identified 11 mammal species. Our approach recorded 47.8% of mammal species that potentially can both occur in the area and use the forest canopy. Our findings indicate that our approach to documenting arboreal ecological dynamics is both efficient and innovative, offering researchers a cost-effective tool for future studies in several vegetation types.

Full text

Vol.:(0123456789)
Mammal Research
https://doi.org/10.1007/s13364-024-00778-7
METHODS PAPER
A new arboreal suspension technique tostrengthen
theunderstanding ofcanopy ecological interactions anddynamics
LuanGabrielAraujoGoebel1,2 · JulianoA.Bogoni1,2 · HernaniFernandesMagalhãesdeOliveira3,4 ·
CarlosdeSouzaFerreira5 · ManoeldosSantos‑Filho1,2
Received: 29 July 2024 / Accepted: 20 December 2024
© The Author(s), under exclusive licence to Mammal Research Institute Polish Academy of Sciences 2025
Abstract
Recent studies in forest canopies have revealed new findings on species interactions and diversity; hitherto poorly understood.
This glaring gap results from the challenges of recording arboreal and scansorial species and their interactions in the verti-
cal strata. We present a new methodology that allows installing camera-traps at multiple forest strata to detect species and
record ecological interactions. We have developed wooden support to place the environmental monitoring equipment at the
forest canopy, evaluating and testing its cost-effectiveness by monitoring six fruiting plant species across a triple ecotone
of the Cerrado, Amazon, and Pantanal biomes. Our structure to install camera-traps in the forest canopy had a total cost
of US$ 50.00, whereas a tree-climbing kit reaches US$ 1,248.00. After 11months of using camera-traps coupled with our
wood structure, we recorded 137 mammal-plant interaction events and identified 11 mammal species. Our approach recorded
47.8% of mammal species that potentially can both occur in the area and use the forest canopy. Our findings indicate that our
approach to documenting arboreal ecological dynamics is both efficient and innovative, offering researchers a cost-effective
tool for future studies in several vegetation types.
Keywords Animal-plant interactions· Biomonitoring tools· Canopy research methods· Ecological networks· Ecosystem
dynamics
Introduction
In the last 20years, many studies have focused on recording
and understanding the mechanisms involved in animal-plant
interaction assembly as well as arboreal forest ecosystem
dynamics (Gregory etal. 2014). The results from several
studies have revealed the crucial importance of sampling
arboreal species to increase our comprehension of pollina-
tion, frugivory, and seed dispersal networks (e.g., Andresen
etal. 2018; Moore etal. 2021), which are important eco-
system services for forest phytodynamics and regeneration
processes (Lacher etal. 2019).However, the biology and
ecology (e.g., interactions in vertical forest strata, natural
history aspects) of most arboreal, semi-arboreal, and scanso-
rial (hereafter so-called “arboreal”) species are still poorly
understood, mainly due to the challenges in recording these
species (Olson etal. 2012). Ultimately this challenge is
paramount to solve critical knowledge gaps in terms of our
comprehension about arboreal species roles in ecosystem
dynamics (Bowler etal. 2017; Zhu etal. 2021).
Communicated by: Dries Kuijper
* Luan Gabriel Araujo Goebel
lg.araujogoebel@gmail.com
1 Programa de Pós-Graduação Stricto Sensu em Ciências
Ambientais, Centro de Pesquisa em Limnologia,
Biodiversidade e Etnobiologia doPantanal, Universidade
doEstado de Mato Grosso, Cáceres, MatoGrosso, Brazil
2 Laboratório de Mastozoologia, Universidade doEstado de
Mato Grosso, Cáceres, MatoGrosso, Brazil
3 Departamento de Ecologia, Universidade de Brasília,
Brasília,DistritoFederal, Brazil
4 Departamento de Zoologia, Universidade Federal doParaná,
Paraná, Brazil
5 Estação Ecológica da Serra das Araras, Instituto Chico
Mendes de Conservação da Biodiversidade, MatoGrosso,
Brazil

Mammal Research

Mammal Research
Addressing all these issues is especially important given
that many animal species and their interactions are segre-
gated across forest strata, only occurring at specific forest
heights (Thiel etal. 2021). Moreover, ca. 40% of the world's
species occur in forest canopies (Ozanne etal. 2003). The
urgency to fill these gaps in knowledge is further exacerbated
by the high sensitivity of arboreal species to anthropogenic
factors (e.g., habitat loss and fragmentation) in a rapidly
changing world (Lowman etal. 2013; Nakamura etal. 2017).
For instance, habitat loss had a rampant intensification in
tropical regions worldwide, given that both humid- and dry-
tropics have experienced a staggering loss, amounting to ca.
475,000 km2 of forest cover loss (~ 50%) in only half-decade
(2000 to 2005; Hansen etal. 2010), widely penalizing forest-
dependent species in the Neotropical forests (Püttker etal.
2020).
The installation of biomonitoring equipment (e.g., cam-
era-traps and autonomous recording units) at higher forest
strata is usually done with the use of vertical tree-climbing
techniques (Haysom etal. 2021; Kaizer etal. 2022). The use
of tree-climbing techniques can both: (1) double the costs
associated with any ecological study using arboreal equip-
ment — given that this method requires complex logistics
to transport the equipment and human-resource climbing
training (Haysom etal. 2021) — and (2) increase of potential
risks to people associated with climbing trees, especially
in remote tracks of forest remnants (Haysom etal. 2021).
Furthermore, even after all these efforts, this method still
requires several hours whereby the equipment is placed in
the correct position for the arboreal biomonitoring (Houle
etal. 2004; Bowler etal. 2017). An alternative that has been
used to deal with the problems of accessing forest canopies
has been the use of folding ladders, but this technique also
represents many difficulties in terms of fieldwork logistics
and transport, failing to access the forest canopy embracing
tall trees (40–50m) (Zhu etal. 2021).
Improving the accessibility of researchers and their equip-
ment to monitor forest canopies is crucial, especially against
the pervasive deforestation of tropical forests worldwide.
The use of safe- and low-cost strategies to access forest
canopies to install camera traps, audio recording systems,
and weather stations, can benefit the understating of several
ecological features in the forest canopy. In this study, we
propose a new low-cost and safe methodology to set cam-
era-traps and other types of equipment (e.g., audio record-
ers) on tree canopies. Our goal was to depict an effective
technique in replacement of tree-climbing (an expansive
and risky technique) to install any recording equipment in
the forest canopy, being therefore widely applied for any
major forest biomes. Furthermore, we offer new directions
and insights for future studies involving camera traps and
other techniques. Our primary hypothesis was that our new
technique would be responsible for recording an important
parcel of all those mammals able to occur in the tallest tree
canopies in the region (N = 23) according to IUCN distribu-
tion range maps (IUCN 2024). Whereas the estimated cost
of camera-trap installation in the forest canopy based on
our new technique represents less than 5% in comparison to
the estimates involving the traditional tree-climbing method
(e.g., Gregory etal. (2014) and its logistics features.
Methods
Study area
Our study was performed in the Serra das Araras Ecological
Station (EESA; 15°38′32.0″S; 57°11′27.3″W), Mato Grosso
State, Brazil (FigureS1). This fully protected area (PA) is
located in a triple-ecotone zone, represented primarily by the
Cerrado biome and embodied by both the Amazon and Pan-
tanal biomes (Brazil 2016). We sampled a landscape mosaic
under the influence of the rivers Camarinha and Salobinha
composed by both semideciduous seasonal forests — widely
dominated by the presence of Attalea speciosa Mart. Ex
Spreng — riparian forest (“Mata-de-galeria”) and savannah
vegetation (Cerrado sensu stricto) (Brazil 2016).
Sampling design
From September 2019 to July 2020, using 29 camera-traps
installed at canopy heights ranging from 10-m to 30-m, we
continuously monitored fruit consumption performed by
mammals in six plant species. In doing so, we installed one
camera-trap per individual tree (TableS1), therefore, sam-
pling four individuals of Hymenaea courbaril L. (Fabaceae)
(46days each plant – totalling 184 camera-trap-days [CTD]),
five individuals of Genipa americana L. (Rubiaceae) (355
CTD), three individuals of Pouteria ramiflora (Mart.) Radlk.
(Sapotaceae) (192 CTD), nine individuals of Cordiera mac-
rophylla (K. Schum.) Kuntze (Rubiaceae) (270 CTD), five
individuals of Dipteryx alata Vogel (Fabaceae) (705 CTD),
and three individuals of Diospyros hispida ADC. (Eben-
aceae) (192 CTD). Amounting to 416days of sampling
and 1,898 camera-trap-days of effort — once the study
Fig. 1 A Guide for fabrication and installing our new approach for
arboreal suspension method associated with different tools to monitor
ecological dynamics at different forest strata, Bplace the wood sup-
port together with the camera trap on a branch of the fruiting tree,
Ctie the ropes of the wood support on stable points, Drecord of the
White-eared Opossum (Didelphis marsupialis) with a fruit from Dip-
teryx alata, Erecord from Azara’s Capuchin (Sapajus cay) manipu-
lating a fruit from Dipteryx alata, and (F) record from a squirrel
(Guerlinguetus sp.). L.G.A. Goebel created all illustrations used in
this figure with a paid license (invoice number 20171) on the Mind
the Graph website (https:// mindt hegra ph. com/). The photograph is
credited to L.G.A. Goebel
◂

Mammal Research
was dependent on trees phenology of fructification — this
camera-trapping design was employed in tree individuals at
least 200-m apart (Fig.1) to avoid spatial autocorrelation
and allow replication between sampling trees. The camera-
traps remained active for 24h per day, configured to perform
video records of 10-s after motion detection, with intervals
of five seconds between any videos (Goebel etal. 2023).
Construction andinstallation ofthecamera‑trap
structure intheforest canopy
To ecological interactions and environmental variables in
the tree canopy, we developed an entirely new hand-made
wood support in which the camera-traps and other equip-
ment could be placed, representing a stable instrument for
forest canopy monitoring (Fig.1B). The wooden support
contains two rods that weigh about 700g: one vertical
(40 × 4 × 1-cm) and one horizontal (30 × 4 × 1-cm), which
were nailed in a cross shape with nails (see Fig.1A). Among
the holes in the structure, the upper one serves to suspend
the camera in the canopy. The second lower holes allow the
structure to be rotated by 360° and inclined by up to 120°.
Moreover, this part of our wood support permits interesting
stability, therefore avoiding unwanted photos with shots trig-
gered by camera-traps movement.
To install the structure in the tree canopy, we previously
chose the branches with the highest number of fruits, pre-
sumably all those with the largest species detectability.
Afterward, we cast a sinker tied to a fishing line using a
slingshot for the system (camera-trap and wood support)
to be installed (Fig.1A and Fig.2). Subsequently, 2.5-mm
ropes were inserted into each hole, ropes were used to sus-
pend the structure, and directed to branches. These ropes
belonging to the three holes of the structure were tied to
surrounding trees with approximately five centimeters of
diameter at breast height (DBH), to give more stability and
avoid unwanted shots. For the entire realization of this sam-
ple, we used 10 wooden structures, composed — beyond
the wood — by 20 nails (to the structure confection), about
150-m of nylon line, 300-of 2.5-mm braided rope, a sling-
shot, and 15 fishing sinkers, with a final total cost of US$
50.00. Installing the structure takes between 30 and 40min,
while changing cards and batteries takes around 10min (just
climb down the top rope).
Data analysis
We used the bipartite R-package to visualize the networks
of interactions between frugivorous mammals and the fruit-
ing trees. We therefore constructed adjacency matrices of
visits and frugivory events between mammals and plants
using bipartite graphs (Dormann etal. 2008). To do so, we
considered both frugivory (i.e., fruit removal or eating) and
visit events (Dormann etal. 2008) (Fig.3). We considered
independent observations every time that an individual
of any species was captured by the camera-trap — either
on a frugivory or visit event — but did not return for at
least 30-s (Goebel etal. 2023). Moreover, to evaluate sam-
pling efficiency to detect species with arboreal habits, we
created a rarefaction curve using the iNEXT R-package R
software (Hsieh etal. 2016) (see Supplementary Material).
We extrapolated the rarefaction curves based on threefold
the minimum abundance of any tree sample. Interpolation-
extrapolation via the rarefaction approach generates confi-
dence intervals of 95% hindering statistical comparisons
between samples (Colwell etal. 2012; Hsieh etal. 2016).
All data analyses were performed in R v. 4.1.0 (R Develop-
mentCore Team 2020).
In order to compare the mammal species recorded with
our new technique to use camera-traps in the forest canopy
in relation to the potential local fauna, we obtained the mam-
mal fauna with putative occurrence within the EESA region
based on the IUCN species range polygons (IUCN 2024),
and then classified this assemblage to rendering a subset
of species able to use the canopy. Finally, to compare our
technique with the traditional tree-climbing technique to
install camera-traps in the forest canopy, we used an online
search in the Google search engine and obtained — using the
first search result — the mean price (N = 10) of a complete
kit of tree-climbing (ignoring any importation and delivery
costs). As the studies only indicate equipment costs and do
not detail labour costs, we assumed a daily cost of a tree-
climbing professional near US$ 60.00, disregarding human
resources training which costs $686 per person (Haysom
etal. 2021).
Results
Overall, our technique recorded ~ 50 blank events during
the installation and camera-trap revisions. Thus, during the
sampling period using our new methodology to suspend and
monitor arboreal fauna in the forest canopy, we recorded 137
ecological interactions, being 26 (19.0%) frugivory events,
and 111 (81%) visits records (Fig.3). We have identified 11
mammal species, belonging to six orders and eight families
interacting with the six plant species sampled (TableS2).
Of these species, Azaras's capuchin(Sapajus cay) (Fig.1E),
wolly mouse opossum (Marmosa demerarae), and kinkajou
(Potos flavus) had the highest visitation rate. The asymp-
tote of the curve indicated that new frugivore species and
frugivory events would be recorded with additional sampling
(FigureS2).
Our extraction of mammal species that range in the
EESA area and arguably use the forest canopy resulted in
23 species, belonging to Primates, Didelphimorphia, Pilosa,

Mammal Research
Carnivora, and Rodentia orders. Therefore, our approach
recorded 47.8% of the mammal species that potentially can
both occur in the area and use the forest canopy. As afore-
mentioned, our wood structure to install camera-traps in
the forest canopy had a total cost of US$ 50.00, whereas
our search online to acquire a tree-climbing kit reached
US$ 1,248.00 (± 614.17; ranging from 599.00 to 2,491.00)
amounting to a total cost, considering the climbing daily,
of US$1,308.00. Thus, our approach using 10 structures
decreased the fieldwork cost by 96.2%. Even considering
only the climbing daily — stipulated in US$ 60.00 — each
wood structure to sample forest canopy represents only
8.33% of the professional climbing daily, being widely
re-installed according to sampling requirements. Further-
more, the time for installation (40min) and maintenance
(10min) is reduced compared to other methods.
Discussion
The cost-effectiveness of biodiversity research depends
on taxa, equipment evolved, and labor costs (Moore etal.
2021). While camera-trapping has proven to be a remark-
ably effective non-invasive method for wildlife monitor-
ing, its application to the study of arboreal species remains
relatively limited and depends on high-complex techniques
Fig. 2 Step-by-step procedures showing how our suspension tech-
nique can be used to steadily place a wood support in association with
different tools (a camera-trap highlighted in our example) to monitor
canopy dynamics. A-B: A sinker should be tied to an end of a nylon
line and catapulted over a tree branch, where the wood support will
be positioned over the next steps; C-D: the sinker will be removed
and the nylon line will be replaced by a 2.5mm rope (red circle); D-
E–F: Ropes will be tied to each of the three holes of the wood sup-
port and to a tree with at least 30 cm of diameter at breast height
(DBH) to lift up the camera trap after it is configured and activated
(to give the wood support more stability); To uninstall the camera-
trap, the first rope who was used to suspend it should be untied from
the tree in order to take it safely to the ground. Afterwards, the other
ropes can be untied, and the equipment completely removed. All
illustrations used in this figure were created by L.G.A. Goebel with a
paid license (invoice number 20171) on the Mind the Graph website
(https:// mindt hegra ph. com/)

Mammal Research
(e.g., Gregory etal. 2014). Camera-trap per se reaches a
cost of ca. US$ 360.00 per unit in Brazil, representing
an expensive equipment for many tropical countries. Our
technique to monitor the forest canopy using camera-traps
can reduce the installation costs from 91.7% to 96.2% in
comparison with the tree-climbing costs of suspending
camera-traps in the forest canopy. Our results revealed
that ~ 50% of mammal fauna able to occupy the canopy
in the region can be recorded with this technique, whose
results can be even better with an increase in sampling
effort. Thus, our results indicate a successfully applied
new methodology to set camera-traps for recording arbo-
real faunas and their ecological features.
Moreover, considering the blank events (~ 50 to install
and uninstall) our new methodology also had considerable
success, given that camera traps produce many blank images
in any vertical strata (Wei etal. 2020). In our study, the tall-
est trees (e.g., H. courbaril with individuals up to 30m tall)
recorded the vast majority of blank events. In terms of cost-
effectiveness, we have detected one-third of the non-flying
mammal species previously reported for the Serra das Araras
Ecological Station (Santos-Filho 2000). This reinforces the
importance of the use of camera-traps in superior strata of
forests; not only to record the mammal species occurring
in the sampling sites but also their ecological interactions.
Hence, considering 147 camera-trap-days along 29 trees
Fig. 3 Interaction networks
between frugivorous mammals
and fruiting trees in the Serra
das Araras Ecological Station,
located in a transitional area
of the Cerrado biome with
the Amazon and Pantanal
biomes, where (A) represents
the frugivory events and (B)
represent the visiting events.
Fernanda D. Abra (ViaFAUNA)
kindly provided the mammal
species drawing used in the fig-
ure and the tree was illustrated
by J.A. Bogoni

Mammal Research
monitored we were able to record 137 ecological interac-
tions, representing a ratio of 0.94 interaction per sampling
day. Disregarding the cost of the camera-traps, each mam-
mal record using our approach reached a cost of US$ 0.36
whereas the use of the tree-climbing technique would cost
US$ 9.11 per record, an increase of 96.0% per record.
Other sampling methods to capture information on for-
est canopy faunas and their dynamics consist of direct
observations, interviews, bibliographic reviews, and census
(Fominka etal. 2020; Fournier etal. 2023). All these meth-
odologies have different limitations, which hinder the ability
of researchers to properly understand canopy complexity in
comparison to some advantages derived from camera-trap-
ping (Quintero etal. 2022). Even though camera-traps can
have a high cost, this equipment — in comparison to other
methods, such as direct observations and censuses — per-
mits substantial increments in species detectability, decreas-
ing the in loco sampling effort in terms of hours spent in
the fieldwork to detect and record ecological interactions of
secretive and arboreal species (Quintero etal. 2022).
Some advantages of using our new sampling method
are both the low-cost approach and the fast speed of setting
up the entire equipment. To install the cameras following
the methodology described here, it takes about 40min (see
methods) while climbing to the canopy and positioning the
equipment takes between 1.5 and 6h per tree (Haysom etal.
2021). Our methodology does not require the presence of
robust trees that support the weight of a climbing profes-
sional to install the camera-traps, deprecating the traditional
method for setting of cameras-traps in the canopy. Further-
more, our method also enables us to use guidance ropes to
orientate camera-traps to specific areas of interest in the tree
canopy, such as specific branches with fruits or animal path-
ways. Whereas traditional options to install arboreal cam-
era-traps are more challenging given they require climbing
skills, special equipment, and are difficult to apply in remote
wildlands (Bowler etal. 2017). Moreover, forests with tall
trees, such as the Amazon Forest (where trees can reach
over 50-m), can pose an important additional challenge to
place canopy camera-traps with methods that require climb-
ing skills and expensive equipment. Finally, it can also be
potentially important in projects involving citizen science,
since the local citizens do not need to have or develop tree-
climbing skills to set the cameras-traps and other biomoni-
toring tools.
Even though other methods have been proposed for lift-
ing camera-traps to study canopy ecology, such as the Orion
Camera System (OCS) (Méndez-Carvajal 2014), our method
represents a fine-tuned advance considering the adjustments
of the position of the camera-traps in the canopy, given that
camera-traps can rely on three ropes attached to different
parts of the wood structure. Thus, we are able not only to
rotate the camera-trap, such as what is done in the OCS,
but also to adjust the equipment attached to different suit-
able angles to better frame the targeted object in the forest
canopy.
One of the limitations related to our method remains
the time and expertise to get the fishing line over a spe-
cific branch of interest while using the slingshot. Untrained
researchers trying to get our equipment set might need some
time learning to use slingshots to successfully hit specific
branches. For instance, installing cameras in higher strata,
where canopy density is greater, could pose significant chal-
lenges, making placing cameras in specific locations dif-
ficult. Additionally, remotely adjusting the camera to focus
on target branches with fruit can be particularly challenging
when relying solely on manipulating ropes from the ground.
However, this is also a limitation of the climbing techniques
used to place camera-traps and other tools in the canopy.
Thus, our method still remains considerably advantageous
in relation to other high-cost-effectiveness forms of access-
ing the canopy.
In terms of the wood structure here depicted, we also
suggest that other studies in different areas should tie the
ropes to trees that have at least 30-cm of DBH. It is impor-
tant to highlight that the amount of rope used in this pro-
cess depends on the height of the trees chosen for the study.
To install the structure in the canopy, two researchers are
needed: one who can hit the target branch and an assistant
for the installation of equipment. We suggest that research-
ers use helmets and gloves as safety equipment, avoiding
accidents with sinkers and fishing lines. Despite our method
providing comprehensive advances to solve the challenges
associated with placing arboreal camera-traps, some details
still need to be observed to increase the success in recording
animal species and their interactions in the canopy. One of
them is to focus the camera-traps on branches and tree parts
that are steadier, to avoid camera-traps being triggered by
moving objects in the background, resulting in a large num-
ber of photographs without animal records, which has been
listed as one of the important potential problems for arboreal
camera-traps (Gregory etal. 2014; Hongo etal. 2020).
Accessing the limits of forest strata in order to collect
ecological data is extremely important, as tropical canopies
were considered one of the world’s last ecological frontiers
(Erwin 1983; Nakamura etal. 2017). Despite many efforts
to understand canopy biodiversity, many gaps still persist
(see Rowcliffe 2017; Moore etal. 2021). The challenge of
recording ecological interactions in the tree canopy (Jordano
2016) depends on enhancing the methodologies to suspend
camera-traps and other tools, especially considering that this
modern equipment only recently began to be used for this
purpose (Rowcliffe 2017). Thus, there are many fields for
improvements aiming to increase our detection ability of
elusive and arboreal species (Moore etal. 2021; Zhu etal.
2021; Kaizer etal. 2022).

Mammal Research
In future studies using camera-traps, we strongly advo-
cate that sampling arboreal animals needs to be included
given that this parcel of local biotas is essential to improve
our understanding of community structuring and ecosystem
processes. Furthermore, this recommendation is particularly
important considering the susceptibility of these groups
of species to habitat change, rapidly overwhelming in the
Anthropocene era. We also suggest a higher sampling effort
in order to properly sample arboreal species, especially to
record mammal-plant interactions. Finally, we emphasize
that the information here synthesized should supply a new
approach for the use of different biomonitoring techniques
and stimulate further discussions about the use of robust
tools that allow for a better comprehension of community
ecology and ecosystem functioning. Therefore, our tech-
nique depicts a new methodology with a potential for broad
use in ecology, and our primary hypothesis posed here was
corroborated.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 1007/ s13364- 024- 00778-7.
Acknowledgements We would like to thank the State University of
Mato Grosso. We would also like to thank the staff of the Serra das
Araras Ecological Station and the Chico Mendes Institute for Biodiver-
sity Conservation (ICMBio) for the support provided in carrying out
the research. We thank two anonymous reviewers for their important
contributions to our manuscript.
Author contributions LGAG and MSF conceived the ideas; LGAG,
CSF, and MSF defined the sampling design; LGAG and CSF collected
the data; LGAG, JAB, and HFMO analysed the data; LGAG wrote the
first draft of the manuscript. All authors contributed critically to the
drafts and gave final approval for publication.
Funding This study was partially funded by the Coordenação de Aper-
feiçoamento de Pessoal de Nível Superior—Brasil (CAPES) and Con-
selho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
through scholarships to LGAG and HFMO.
Data availability The data used to support the findings of this study are
available from the corresponding author upon request.
Declarations
Conflict of interest The authors declare no competing interests.
References
Andresen E, Arroyo-Rodríguez V, Ramos-Robles M (2018) Primate
seed dispersal: old and new challenges. Int J Primatol 39:443–465
Bowler MT, Tobler MW, Endress BA, Gilmore MP, Anderson MJ
(2017) Estimating mammalian species richness and occupancy
in tropical forest canopies with arboreal camera traps. Remote
Sens Ecol Conserv 3:146–157
Brazil (2016) Plano de manejo Estação Ecológica da Serra das Araras.
Ministério do Meio Ambiente, Brasília
Colwell RK, Chao A, Gotelli NJ, Lin SY, Mao CX, Chaz-
don R, Longino JT (2012) Models and estimators linking
individual-based and sample-based rarefaction, extrapolation
and comparison of assemblages. J Plant Ecol 5:3–21
Dormann CF, Gruber B, Fründ J (2008) Introducing the bipartite
package: Analysing ecological networks. Interaction 1:8–11
Erwin TL (1983) Tropical forest canopies: the last biotic frontier.
Bull Ecol Soc Am 29:14–20
Fominka NT, Oliveira HFM, Camargo NF, Jost Robinson CA, Fokam
EB (2020) Altitudinal and seasonal variation in the structure
of nocturnal primate assemblages on Mount Cameroon. Int J
Primatol 41:714–731
Fournier CS, Graefen M, McPhee S, Amboko J, Noonburg EG,
Ingram V, Hart TB, Hart JA, Detwiler KM (2023) Impact of
Hunting on the Lesula Monkey (Cercopithecus lomamiensis) in
the Lomami River Basin, Democratic Republic of the Congo.
Int J Primatol 44:282–306
Goebel LGA, Vitorino BD, Frota AVB, dos Santos-Filho M (2023)
Body mass determines the role of mammal species in a frugi-
vore-large fruit interaction network in a Neotropical savanna.
J Trop Ecol 39:e12
Gregory T, Carrasco Rueda F, Deichmann J, Kolowski J, Alonso A
(2014) Arboreal camera trapping: taking a proven method to
new heights. Methods Ecol Evol 5:443–451
Hansen MC, Stehman SV, Potapov PV (2010) Quantification
of global gross forest cover loss. Proc Natl Acad Sci USA
107:8650–8655
Haysom JK, Deere NJ, Wearn OR, Mahyudin A, Jami JB, Reynolds
G, Struebig MJ (2021) Life in the Canopy: Using camera-traps
to inventory arboreal rainforest mammals in Borneo. Front For
Glob Change 4:673071
Hongo S, Dzefack ZSC, Vernyuy LN, Munami S, Nakashima Y, Djiéto-
Lordon C, Yasouka H (2020) Use of multi-layer camera trapping
to inventory mammals in rainforests in southeast Cameroon. Afr
Study Monogr Suppl Issue 60:21–37
Houle A, Chapman CA, Vickery WL (2004) Tree climbing strategies
for primate ecological studies. Int J Primatol 25:237–260
Hsieh TC, Ma KH, Chao A (2016) iNEXT: an R package for rarefac-
tion and extrapolation of species diversity (Hill numbers). Meth-
ods Ecol Evol 7:1451–1456
IUCN (2024) The IUCN Red List of Threatened Species. Version
2024–2. Available at https:// www. iucnr edlist. org. Accessed on
20 November 2023
Jordano P (2016) Sampling networks of ecological interactions. Funct
Ecol 30:1883–1893
Kaizer MC, Alvim TH, Novaes CL, Mcdevitt AD, Young RJ (2022)
Snapshot of the Atlantic Forest canopy: surveying arboreal mam-
mals in a biodiversity hotspot. Oryx 56:825–836
Lacher TE Jr, Davidson AD, Fleming TH, Gómez-Ruiz EP, McCracken
GF, Owen-Smith N, Peres CA, Vander Wall SB (2019) The func-
tional roles of mammals in ecosystems. J Mammal 100:942–964
Lowman M, Devy S, Ganesh T (2013) Treetops at Risk: challenges of
global canopy ecology and conservation. Springer, New York,
NY, USA
Méndez-Carvajal PG (2014) The orion camera system, a new method
for deploying camera traps in tree canopy to study arboreal pri-
mates and others mammals: a case study in Panama. Mesoameri-
cana 18:9–23
Moore JF, Soanes K, Balbuena D, Beirne C, Bowler M, Carrasco-
Rueda F, Cheyne SM, Coutant O, Forget PM, Haysom JK, Houli-
han PR, Olson ER, Lindshield S, Martin J, Tobler M, Whitworth
A, Gregory T (2021) The potential and practice of arboreal cam-
era trapping. Methods Ecol Evol 12:1768–1779
Nakamura A, Kitching RL, Cao M, Creedy TJ, Fayle TM, Freiberg M,
Hewitt CN, Itioka T, Pinkoh L, Keping M, Malhi Y, Mitchell A,
Novotny V, Ozanne CMP, Ashton LA (2017) Forests and their
canopies: achievements and horizons in canopy science. Trends
Ecol Evol 32:438–451

Mammal Research
Olson ER, Marsh RA, Bovard BN, Randrianarimanana HL, Ravaloha-
rimanitra M, Ratsimbazafy JH, King T (2012) Arboreal camera
trapping for the Critically Endangered greater bamboo lemur Pro-
lemur simus. Oryx 46:593–597
Ozanne CM, Anhuf D, Boulter SL, Keller M, Kitching RL, Körner C,
Meinzer FC, Mitchell AW, Nakashizuka T, Dias PS, Stork NE
(2003) Biodiversity meets the atmosphere: a global view of forest
canopies. Science 301:183–186
Püttker T, Crouzeilles R, Almeida-Gomes M etal (2020) Indirect
effects of habitat loss via habitat fragmentation: A cross-taxa
analysis of forest-dependent species. Biol Conserv 241:108368
Quintero E, Isla J, Jordano P (2022) Methodological overview and
data-merging approaches in the study of plant–frugivore interac-
tions. Oikos 2022:1–18
R Development Core Team (2019) R: A language and environment
for statistical computing. R Foundation for Statistical Computing.
Available at https:// www.r- proje ct. org/. Accessed on 20 December
2023
Rowcliffe JM (2017) Key frontiers in camera trapping research. Remote
Sens Ecol Conserv 3:107–108
Santos-Filho M (2000) Uso de hábitats por mamíferos não-voadores na
Estação Ecológica Serra das Araras, Mato Grosso, Brasil. Thesis,
Instituto Nacional de Pesquisas da Amazônia
Thiel S, Tschapka M, Heymann EW, Heer K (2021) Vertical stratifica-
tion of seed-dispersing vertebrate communities and their interac-
tions with plants in tropical forests. Biol Rev 96:454–469
Wei W, Luo G, Ran J, Li J (2020) Zilong: A tool to identify empty
images in camera-trap data. Ecol Inf 55:101021
Zhu C, Li W, Wang D, Ding P, Si X (2021) Plant–frugivore interac-
tions revealed by arboreal camera trapping. Front Ecol Environ
19:149–151
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds
exclusive rights to this article under a publishing agreement with the
author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of
such publishing agreement and applicable law.