Estimation of Wildlife Abundance Based on Track Counts and Assessment of Environmental

and Anthropogenic Factors Influencing Abundance

Remigio Rangel Nhamussua1; Juarez C. B. Pezzuti2; Daniely Felix-Silva3 & Sérgio de Azevedo4

1. INTRODUCTION

The wildlife monitoring should be a fundamental practice for any stakeholder aiming to conserve
or manage wildlife populations (Stephenson, 2019). Monitoring and assessing trends in wildlife
abundance provide crucial information for management decisions, such as protecting endangered
species, controlling invasive and pest species, and setting annual quotas for sustainable hunting.
This is vital for stakeholders like wildlife managers, researchers, and local communities
(Ahlswede et al., 2019; Keeping & Pelletier, 2014).

Reliable information on wildlife distribution and abundance is essential for conservation

planning and action, making the selection of appropriate field survey methods and robust
modeling approaches play a key role in wildlife protection and habitat management (Li et al.,
2022). Various methods are used to estimate animal abundance, such as sign counts, direct
counts along line transects, vehicle-based counts, aerial counts, and the use of camera traps. The
choice of abundance estimation methods generally depends on several factors, including costs,
the technical and scientific capacity of the people involved in the research, and the type of
species being studied (Forsyth et al., 2021).

Conventional wildlife abundance estimation methods often require data collection over long
periods, demanding substantial material, financial, and human resources. Without these
resources, robust analyses and reliable abundance estimates are not possible. For example,
transect surveys require conducting multiple repeated routes over long periods to gather
sufficient data for robust analysis (Ahmad et al., 2021).

1
UFPA-Núcleo de Altos Estudos Amazónicos. [email protected]
?
Univrsidade Lúrio-Faculdade de Ciências Agrárias
2
UFPA-Núcleo de Altos Estudos Amazônicos, [email protected]
3
Dra, Daniely Felix-Silva- REDEFAUNA. [email protected]
4
Universidade Lúrio-Faculdade de Ciências Agrárias, [email protected]
In sub-Saharan Africa, aerial surveys have been widely used to obtain abundance and density
estimates (Frederick et al. 2015). Despite a long history of using this method in research and
monitoring programs, common problems have not yet been adequately addressed, such as error
in the detection of some species, since in some cases the individual is not available to be detected
or is missed due to error in counting, and incorrect identification of the species, since some
species do not have nocturnal behaviour (Davis et al. 2022).

An alternative that is increasingly being used consists of indirect survey methods (Keeping &
Pelletier, 2014; Esbach, 2023). Given that conservation resources are limited, there is a need for
efficient means to assess wildlife abundance, making indirect methods a viable alternative for
detecting terrestrial mammals for many species (Keeping & Pelletier, 2014).

Moreover, knowledge about how animals use space within their range and the connections
between habitat patches is fundamental to the success of conservation actions (Boitani et al.,
2008). Species density and distribution can be influenced by various factors, both biotic and
abiotic (Gunda et al., 2020). Additionally, to properly manage wildlife populations, both to
ensure their conservation and to reduce conflicts with rural communities, it is crucial to
understand the relationship between wildlife and the environment, as well as the factors that
influence their distribution and abundance (Williams et al., 2017). Some studies (Tsheboeng,
2018; Stegmann et al., 2019; Kauffman, 2022) show the influence of river availability and
proximity on species distribution. On the other hand, human activities such as agriculture and
grazing (Kauffman, 2022) influence species abundance. According to Renwick et al. (2017),
biodiversity management, species protection, species reintroduction, and predicting potential
environmental impacts require detailed information based on knowledge of species distribution
and the relationships between distribution and environmental variables. These pieces of
information, according to the authors, are increasingly important for management and decision-
making. Franklin and Miller (2014) also state that species distribution maps or habitat suitability
maps are necessary for resource management and conservation planning. The present study aims
to estimate wildlife density using the indirect "footprint counting" method, as well as to assess
the effects of river distance and proximity to communities on species abundance and distribution.
2. MATERIALS AND METHODS

2.1. Study Area

The study was conducted in the Chipanje Chetu Conservation Area (ACCC), a community
conservation area adjacent to the Niassa Special Reserve, located in the administrative posts of
Matchedje and Macaloge in Sanga District, in the northern region of Niassa Province,
Mozambique (Figure 1). Covering an area of 6,500 km², it was established by the government as
a Community Conservation Area and assigned to the Chipanje Chetu Program Management
Council (PCC) to explore, use, and conserve the area's forest and wildlife resources, and to
practice ecotourism. Five communities reside in the area, totaling 6,500 inhabitants. With two
main rivers, Moola and Massinge, the predominant forest formation in Chipanje Chetu is open
miombo, characterized by the presence of various wildlife species, including mammals, birds,
reptiles, and amphibians (Sitoe et al., 2007; Lipilichi Investimentos, 2021). The ACCC has two
seasons: a hot and rainy season during which the area presents clayey, red, and well-drained
soils, offering favorable conditions for agriculture (MAE, 2014).
Figure 1. Geographical location of the Chipanje Chetu Conservation Area

2.2. Data Collection

The Data collection was conducted from May to December 2022, along 20 transects of 10 km
each (Fig. 2), following sections on the main road and secondary roads, on dirt roads with good
conditions for tracking footprints. Additionally, 38 surveys were conducted off-road, known as
guided trails, and 19 surveys in community agricultural areas, totaling a distance of 366 km
crossing all types of environments. Surveys of large and medium-sized vertebrate abundance
were conducted on foot, following the procedures described by Keeping and Peletier (2014) and
Ahlswede et al. (2019), which essentially consist of walking a transect and recording all animal
tracks that cross it. For density estimates, only animal tracks that crossed the transect within the
last 24 hours were used, requiring highly refined knowledge and experience (Elbroch et al. 2010,
Liebenberg 2013). The transects were covered by a team of three researchers, invariably
including a local community member experienced in tracking animals and their footprints. Track
counts began at 6:00 AM and were conducted between 10:00 and 11:00 AM.
 Figure 2. Location of the 10 km transects

Each observation was recorded using the Cybertracker app for smartphones, with a specific
customization for the region's wildlife. All large and medium-sized vertebrate ungulates and
carnivores observed, detected by signs, sounds, and tracks were recorded.

2.3. Data Analysis

The Species density estimates were determined based on the equation used by Ahlswede et al.
(2019) and Keeping and Pelletier (2014):

D
Where D = Density (km²); X = number of track intersections along the transect;

S = length of the transect; M = average daily distance traveled by a species.

The average daily distance was determined based on Garland's (1983) equation using the
following equations:
DMDCarnívore(km/dia) =3.877M0.22±0.08
DMDOutros mammals(km/dia) =0.0875M0.22±0.08

Where: DMD = Daily movement distance; M = Body mass of the animal.

The population estimate for each species was determined by multiplying the absolute density by
the total area size (A = 6500 km²). Since the density values for some species were low, statistics
for average density in the area were computed as the number of individuals per 100 km². To
understand the variation in observations, the robustness and effectiveness of the sampling
process were analyzed using the confidence interval (Formula 3).

onde Fórmula (3)

Fórmula (4)

Where:
Var Y = variance of the sampled population; LC = confidence limit (t-value of the bilateral t-
student at 95% confidence interval with n-1 degrees of freedom); SE = standard error; Y = total
number of individuals of a given species; %LC = relative percentage of precision (Y = total
number of individuals of a given species).

The Body mass values were obtained from the Skinner & Chimimba (2005) manual and online
sources (www.ielc.libguides.com; https://eol.org; https://www.krugerpark.co.za/africa). All
calculations and density analyses were performed using the R software package (R Core Team,
2014). To compare species composition in different types of environments where data were
collected, a permutational multivariate analysis of variance (PERMANOVA, 10,000
permutations) was conducted using the “vegan” package in R (Skalski, 2018; Anderson, 2001).
The following hypotheses were tested:
Ho: Vegetation type does not influence species composition and distribution;
Ha: Vegetation type influences species composition and distribution.

Based on data obtained from different types of environments, species distribution maps were
created following the procedure described by Gunda et al. (2020). Initially, using the
geographical coordinates of each animal presence record made with Cybertracker, the points
were overlaid with the vegetation layer using QGIS 3.32 software. The vegetation cover layer of
the study area was obtained from the ARCGIS platform
(www.arcgis.com/apps/webappviewer/index.html?id=1e201cf974584b38ac5dd92b005c99ae).
After overlaying the species records, the intersection between the records and vegetation types
was analyzed to determine the number of individuals for each vegetation type and for each
species.

To assess the influence of environmental and anthropogenic factors on abundance and density,
the 10 km transects were divided into 1 km segments, according to Benchimol and Peres (2015).
From the center of each 1 km segment of the transect, the distance to the nearest river and
community was measured using QGIS 3.32 software. Considering that abundance or species
richness analyses commonly have many zeros (Da Silva et al., 2022), the relationship between
environmental predictor variables (distance from rivers) and anthropogenic predictor variables
(distance from communities) on response variables (species abundance) was evaluated using a
zero-inflated negative binomial model (Sileshi et al., 2009).

3. RESULTS

3.1. Density Estimates

Seventeen species of medium and large mammals were identified along the 10 km transects
(Table 1), including 4 species from the order Carnivora, 10 from the order Artiodactyla, and 1
each from the orders Rodentia, Perissodactyla, and Proboscidea, respectively. In the Artiodactyla
order, the highest density was recorded for the bushbuck (0.8 individuals/km²), followed by the
pala-pala (0.71 individuals/km²) and the piva (0.36 individuals/km²), with the lowest abundances
recorded for the bushpig (0.007 individuals/km²), the Imbabala (0.02 individuals/km²), and the
Gondonga (0.04 individuals/km²). For the Carnivora order, the highest density was observed for
the jackal (0.04 individuals/km²), followed by the leopard (0.03 individuals/km²), with the lowest
density recorded for the lion (0.001 individuals/km²). In trails, the highest density estimate was
for the pala-pala (0.24 individuals/km²), followed by the bushbuck (0.187 individuals/km²) and
the porcupine (0.124 individuals/km²). In agricultural fields, the highest densities were recorded
for the pala-pala (0.57 individuals/km²) followed by the eland (0.49 individuals/km²).

The correlation analysis between estimated density and body mass of the species shows that
density is inversely related to mass (P = -0.17), meaning that as the species' weight increases, its
abundance tends to decrease (Figure 3).

Figure 3. Relationship between species mass and abundance

 Table 1. Density estimates obtained through indirect records in the Chipanje Chetu Conservation Area, Mozambique.
TRANSECÇÕES MACHAMBAS TRILHAS
Density Average
Class Ordem Specie
(ind/km density Density
2
) (100km2) ICI ICS (ind/km2) Density (ind/km2)
Mammal Artiodactyla Eland 0.03 0.13 0.01 0.26 0.49 0.007
Mammal Artiodactyla Buffalo 0.27 1.30 0.82 3.42 0.03 0.015
Mammal Artiodactyla Duiker 0.8 3.93 1.90 5.95 0.36 0.187
Mammal Artiodactyla Reedbuck 0.24 1.18 0.52 1.83 _ 0.064
Mammal Artiodactyla Lichtenstein's Hartebeest 0.04 0.18 0.13 0.49 _ _
Mammal Artiodactyla Bushbuck 0.02 0.12 0.01 0.23 _ 0.081
Mammal Artiodactyla Warthog 0.13 0.63 0.22 1.03 _ 0.022
Mammal Artiodactyla Sable 0.71 3.31 1.70 4.93 0.57 0.240
Mammal Artiodactyla Waterbuck 0.36 1.65 0.40 2.90 _ 0.018
Mammal Artiodactyla Bushpig 0.007 0.01 0.01 0.04 _ 0.076
Mammal Rodentia Porcupine 0.03 0.15 0.03 0.33 _ 0.124
Mammal Carnivora Spotted Hyaena 0.022 0.11 0.01 0.21 0.01 0.027
Mammal Carnivora Lion 0.001 0.01 0.00 0.01 0 _
Mammal Carnivora Leopard 0.03 0.14 0.02 0.26 _ 0.041
Mammal Carnivora Jackal 0.04 0.21 0.00 0.43 0.056 0.079
Perissodactyl
Mammal a Zebra 0.09 0.44 0.15 1.03 _ _
Mammal Proboscidea Elephant 0.006 0.03 0.01 0.08 0.019 _
3.2. Abundance in Relation to Distance from Local Communities

To assess the effect of distance from the community on species abundance, 14 mammal species
were evaluated (Fig. 3). A significant positive relationship was observed between distance from
the community and the abundance of Warthog (SE = 0.03794, Z = 1.786) and Waterbuck (SE =
0.04355, Z = 4.371). Non-significant positive trends were found for Zebra (SE = 0.1379, Z =
0.795), Reedbuck (SE = 0.03022, Z = 1.027), Sable (SE = 0.02463, Z = 0.993), Lichtenstein's
Hartebeest (SE = 0.09, Z = 0.341), and Bushbuck (SE = 0.07148, Z = 0.281). These positive
relationships indicate that the abundance of these species tends to increase as the distance from
the community increases. A significant negative relationship was found for the following
species: Buffalo (SE = 0.08949, Z = -1.662), Duiker (SE = 0.02071, Z = -0.498), and Leopard
(SE = 0.04258, Z = -2.290). For Eland, Jackal, Porcupine, and Spotted Hyena, non-significant
negative trends were observed.
Figure 4. Estimated abundance in relation to distance from local communities in the Chipanje
Chetu Community Conservation Area.

3.3. Abundance in Relation to Distance from the River

To assess the effect of distance from the river on species abundance, 14 mammal species were
evaluated (Fig. 4). A significant positive relationship was observed between distance from the
river and the abundance of Porcupine (SE = 0.1607, Z = 1.898). Non-significant positive trends
were found for Common Duiker (SE = 0.04874, Z = 0.685), Jackal (SE = 0.10397, Z = 1.270),
Lichtenstein's Hartebeest (SE = 0.212, Z = 0.125), and Leopard (SE = 0.09831, Z = 0.296).
Significant negative relationships were identified for the following species: Buffalo (SE = 0.42,
Z = -3.152), Reedbuck (SE = 0.0819, Z = -2.079), Sable (SE = 0.06094, Z = -2.963), Spotted
Hyena (SE = 0.1555, Z = -2.440), Waterbuck (SE = 0.1250, Z = -4.528), Bushbuck (SE = 0.256,
Z = -1.704). For Zebra, Eland, Warthog, and Bushbuck, non-significant negative trends were
observed.
Figure 5. Estimated abundance in relation to distance from the river in the Chipanje Chetu
Community Conservation Area.
3.4. Species Composition and Distribution Across Different Vegetation Types in the

Chipanje Chetu Conservation Area

Seventeen mammal species were identified across all surveyed environments. The

PERMANOVA test showed that the type of environment influences species composition and

distribution in the Chipanje Chetu Conservation Area (R² = 0.13, P-Value = 0.0003). The

PERMANOVA results indicate that species composition in deciduous forests was significantly

different when compared to open semi-deciduous forests (R² = 0.08, Adjusted P = 0.028).

Additionally, species composition in deciduous forests was significantly different when

compared to evergreen forests (R² = 0.09, Adjusted P = 0.014). Deciduous and open semi-

deciduous forests exhibited greater mammal richness and abundance compared to other

environments (Fig. 5). In deciduous forests, there were more sightings of sable, Bushbuck, and

Common Duiker. The highest buffalo sightings were observed in open semi-deciduous forests.

The majority of carnivore records, including Jackal, Leopard, Lion, and Spotted Hyaena, were

also observed in open semi-deciduous and deciduous forests, respectively. Notably, there was

also a higher abundance of sable and Elands in agricultural fields. Based on the geographic

coordinates of each animal sighting, distribution maps were created according to vegetation type

(Figures 7, 8, and 9).

Figure 6. Species Composition in Different Environments Surveyed.

AA - Shrublands; AC - Cultivated Areas; FASD - Open Semi-Deciduous Forests; FD -

Deciduous Forests; FSV - Evergreen Forest.

Figure 7. Spatial Distribution of Carnivores in Different Vegetation Types in the Chipanje Chetu

Community Conservation Area.

Figure 8: Spatial Distribution of Medium-Sized Herbivores in Different Vegetation Types in the

Chipanje Chetu Community Conservation Area.

Figure 9: Spatial Distribution of Large-Sized Herbivores in Different Vegetation Types in the

Chipanje Chetu Community Conservation Area.

4. DISCUSSION

Studies using tracks have been effectively utilized for estimating wildlife density in various
conservation areas (Keeping et al., 2014; Ahlswede et al., 2019; Esbach, 2023). The density
estimates obtained in this study closely align with the results from aerial surveys conducted by
Craig and Gibson (2018) in Chipanje Chetu, which recorded higher abundances of species like
impala and bushbuck, and lower abundances of species like bushpig and nyala. Furthermore, this
method allowed for the detection of several species that are typically not observed in aerial
surveys due to their smaller size or preference for denser vegetation. Using footprint
identification, we were able to estimate the densities of some nocturnal species that occur in low
densities, such as the fox, lion, hyena, and leopard, which were not recorded by the aerial survey
method used by Craig and Gibson in 2018. Additionally, species like the bush duiker and
porcupine were not identified in the aerial census. The absence of carnivores in the 2018 census
was expected due to the nature of the method employed. Davis et al. (2022) emphasize that
although aerial methods have a long history in wildlife research and monitoring programs,
common issues such as detection errors for certain species remain unresolved, especially in
estimating the population sizes of elusive or nocturnal animals like carnivores. Fryxell et al.
(2014) note that many terrestrial mammals exhibit characteristics that can be challenging to
identify or comprehend at first glance and generally avoid being seen, making reliable estimates
difficult to produce. Keeping et al. (2014) recommend the use of indirect methods, such as track
counts, for identifying species that are difficult to detect using direct methods.

Lower densities of carnivores compared to other orders were expected based on the trophic
structure of the ecosystem, with lower abundance and biomass at higher trophic levels. The
results are also consistent with those of Keeping et al. (2018) in their study conducted in the
Kgalagadi Transfrontier Park in Botswana, using the same methodology as the present study, and
Ahlswede et al. (2019), where the authors found lower densities of carnivores compared to other
orders. On the other hand, there was a higher presence of impala and eland in the cultivated
fields, which can be attributed to the greater presence of agricultural crops, particularly maize
and beans, in the study area, a fact confirmed by the local communities during the research.
According to the residents of Chipanje Chetu, species such as impala, eland, and monkeys
damage their crops, especially maize and beans. For example, elands require a large area to
graze, and their diet can be supplemented with maize and beans (Pappas, 2002). Animals like
elephants, monkeys, impalas, bushpigs, and warthogs cause significant damage to agricultural
crops, urban areas, and rural properties, resulting in agricultural and economic losses (NCP
Annual Report, 2021).

The results show higher abundances of spotted hyena, buffalo, eland, jackal, leopard, porcupine,
and duiker near communities. The greater abundances near communities can be explained by the
fact that some herbivore species are benefiting from the presence of agricultural crops produced
by communities for their sustenance, which end up attracting these animals. During our surveys,
we recorded higher occurrences of elands in community fields, which can be attributed to the
presence of bean crops, a preferred food source for this species according to local communities.
According to Mekonen (2020), humans and wildlife have been in conflict because agricultural
crops often offer a nutritious and abundant food source for wildlife. The NCP (2021) reported
that in the Niassa National Reserve, wildlife causes significant damage to agricultural crops in
local communities, resulting in agricultural losses. Additionally, the increased abundance near
communities can be explained by the reduction in poaching due to the increase in surveillance
activities in the study area, a fact noted by Kablan et al. (2019) in their study conducted in Tai
National Park in Ivory Coast, where the authors concluded that surveillance activities contributed
to the reduction of poaching and the consequent increase in animal abundance. With the
reduction in illegal hunting, these animals may feel safer moving closer to communities, where
they can find food and shelter.

The results also show a trend of increasing carnivore abundance near communities, which can be
explained by the presence of domestic animals such as chickens, goats, and dogs, which attract
wild animals, especially predators like jackals, leopards, and spotted hyenas. During this
research, some families reported incidents of chickens and goats being attacked by jackals and
spotted hyenas in their enclosures. This finding aligns with a study conducted in Bale Mountains
National Park in Ethiopia by Mekonen (2020), in which the author found that animals such as
jackals, leopards, and spotted hyenas attack domestic animals raised by communities,
particularly goats, dogs, and sheep. In South Africa, specifically in Limpopo, leopards and
jackals are the main predators of livestock (Turpie & Akinyemi, 2018). Torres et al. conducted a
global review on human-wildlife conflicts and found that leopards and spotted hyenas are among
the predators that attack domestic animals in various African countries.

There was an observed increase in the abundance of zebra, sable, and waterbuck with increasing
distance from communities, which can be explained by the fact that these animals are among the
species targeted in sport hunting conducted by the company. According to Parsons et al. (2022),
the most hunted species are relatively less abundant and less associated with humans and
modified landscapes. Hunting activities can contribute to changes in the behavior of some
species, and animals typically avoid areas where hunting occurs, thus altering their distribution
and demographics (Verdade, 1996).

Regarding the distance to rivers, all studied species, except for duiker, jackal, leopard, and
porcupine, exhibited a negative linear relationship; that is, as the distance from the river
increases, the abundance of these species tends to decrease. This relationship can be explained by
several factors, such as the availability of resources along the river, as areas closer to the water
are rich in food. These findings align with those of Mworia (2008), where the distribution of
large mammals in a semi-arid area in southwestern Kenya was influenced by proximity to water
sources during both the dry and rainy seasons. In another study conducted by Gunda et al. (2020)
in the Kijereshi Game Reserve, southwest of Serengeti National Park, a large number of large
mammals were observed near water bodies. A similar result was obtained by Sianga et al. (2017)
in their evaluation of habitat selection by buffalo in northern Botswana, where these species were
found to concentrate in areas close to water. Some species, such as the waterbuck, have a
preference for habitats near water, which may explain their higher abundance in these areas, in
line with the findings of Tolcha et al. (2019), who observed a greater presence of waterbuck in
Maze National Park, Ethiopia.

On the other hand, duiker, jackal, and leopard exhibited a non-significant positive linear
relationship, while porcupine showed a significant positive linear relationship; as the distance
from the river increases, the abundance of these species also increases. According to Urstengurg
(2008), the duiker is a water-independent species and is rarely attracted to water points, which
may explain the positive relationship between the distance from the river and the abundance of
this species. This observation is consistent with the study by Yao et al. (2017) on the distribution
of duiker in a forest reserve in Ivory Coast, where the authors found few individuals in areas
close to the river. A study conducted by Band (1996) showed that porcupines use sites with high
levels of tree cover and a large number of dead trees to protect themselves from predators.
According to Kebede (2017), the jackal is a species that tolerates dry habitats and is a generalist,
meaning it occurs in different types of habitats, mainly in areas associated with grazing activities,
agricultural lands with human populations, and savannas. The fact that it is a generalist species
may explain its lower abundance near rivers. On the other hand, studies conducted by Mondal et
al. (2012) in India and Mann et al. (2020) showed that the probability of leopard presence
increased with decreasing distance to water, indicating that water was considered a factor
influencing leopard distribution. This contrasts with our results, which show an increase in
leopard abundance with increasing distance from water, possibly due to the time of year when
the study was conducted.

In the present study, it was found that species composition and distribution are influenced by the
type of environment. There was greater species richness and abundance in deciduous and semi-
deciduous forests, which can be explained by the fact that most of the herbivores found in the
study area feed mainly on a variety of grasses, which are more common in open deciduous forest
environments where they can receive sunlight, allowing for the growth and development of
grasses. According to Ribeiro et al. (2001), in Mozambique, miombo is the predominant forest
type, and the understory of this vegetation formation consists of shrubs, regenerating trees, and
grasses. The forest cover map of Mozambique, developed by FNDS in 2016, as well as the
National Forest Inventory conducted in 2017, shows that open deciduous forests are the most
predominant in Mozambique compared to other vegetation formations. Given the dependent
relationship between herbivores and carnivores, with carnivores relying exclusively on
herbivores for their diet, the results also show high records of carnivores in the vegetation
formations where herbivores were most commonly recorded. This fact is supported by the study
by del Rio et al. (2001), in which the authors state that herbivores reduce plant biomass, and in
turn, herbivore biomass is controlled by the presence of carnivores.

5. CONCLUSION

The research demonstrated that the estimation of wildlife population density using indirect
methods (footprint counting) recorded a population growth of wildlife in the Chipanje Chetu
community conservation area. The use of indirect methods provides a viable and cost-effective
alternative for monitoring and collecting wildlife data in conservation units. This methodology
allowed for density estimates of 17 species in Chipanje Chetu, including species that were not
identified by other methods used in the same study area. Given the skills of local trackers and the
type of substrate present, we believe that Chipanje Chetu has the conditions to apply the FMP
formula for density estimates, creating an opportunity for the involvement and empowerment of
local communities. The research showed that factors such as human presence, the distribution
and availability of rivers, and vegetation type influence species abundance and distribution
within the study area. For example, there was a trend of increasing carnivore abundance near
communities due to livestock presence. Additionally, a higher species composition was observed
in deciduous and semi-deciduous forests.
ACKNOWLEDGMENTS

The authors would like to thank the communities residing in the Chipanje Chetu Conservation
Area, Block L5 South of the Niassa Special Reserve. We are certain that without their
collaboration and willingness to share their knowledge, this research would not have been
possible. We also extend our gratitude to the management and staff of Lipilichi Investimentos
and Mariri Investimentos for their logistical and personnel support. A special thank you goes to
the Lipilichi enforcement team who taught and guided us during the field surveys. We are
grateful to the local and administrative authorities, specifically the traditional leaders of each
community, the Head of Post, and the Administration of the Sanga District, ANAC, and the
Federal University of Pará, particularly the Pro-Rectorate of Research for the grant awarded.

6. REFERENCES

Ahmad, A., Gary, D., Sinta, R., Putra, S.W., Sagita, N., Adirahmanta, S. N., e Miller A. E., 2021.
Leveraging local knowledge to estimate wildlife densities in bornean tropical rainforests.
Wildlife Biology. doi: 10.2981/wlb.00771

Ahlswede, S., Fabiano, E.C., Keeping, D. e Birkhofer, K., 2019. Using the Formozov–
Malyshev–Pereleshin formula to convert mammal spoor counts into density estimates for long‐
term community‐level monitoring. African Journal of Ecology, 57(2).

Anderson, M.J. 2001 Permutation tests for univariate or multivariate analysis of variance and
regression Canadian Journal of Fisheries and Aquatic Sciences 58(3). DOI: 10.1139/f01-004

Band, Anne-Louise, "Population characteristics and habitat use of porcupines in a partially

burned landscape in northwestern Wyoming" (1996). Graduate Student Theses, Dissertations, &
Professional Papers. 6531. https://scholarworks.umt.edu/etd/6531

Benchimol M, Peres CA (2015) Widespread Forest Vertebrate Extinctions Induced by a Mega

Hydroelectric Dam in Lowland Amazonia. PLoS ONE 10(7): e0129818.
doi:10.1371/journal.pone.0129818
Boitani, L., Sinibaldi, I., Corsi, F. et al. (2008). Distribution of medium- to large-sized African
mammals based on habitat suitability models. Biodivers Conserv 17, 605–621.
https://doi.org/10.1007/s10531-007-9285-0

Davis, K. L., Silverman, E.D., Sussman, A.L., Wilson, R. R., e Zipkin, E. F., 2022. Errors in
aerial survey count data: Identifying pitfalls and solutions. doi: 10.1002/ece3.8733

Da Silva FR, Gonçalves-Souza T, Paterno GB, Provete DB, Vancine MH. 2022. Análises
ecológicas no R. Nupeea : Recife, PE, Canal 6 : São Paulo. 640 p. ISBN 978-85-7917-564-0.

del Rio, C.M; Dugelby, B; Foreman, D., Miller, B., Noss, Reed., Phillips.,Mike. 2001. The
importance of large carnivores to healthy ecosystems. Endangered Species Update.

Deng GT, Demisse BT. Feeding Preference and Habitat Association of Defassa Waterbuck
(Kobus ellipsiprymnus defassa) in Nechisar National Park, Southern Ethiopia.
ScientificWorldJournal. 2021 Oct 18;2021:7498662. doi: 10.1155/2021/7498662. PMID:
34707466; PMCID: PMC8545546.

Esbach, M.S. 2023. Estimating mammal density from track counts collected by Indigenous
Amazonian. High Meadows Environmental Institute, Princeton University, Princeton, NJ, USA.

Furstenburg, D. Focus on the Common / Grey Duiker (Sylvicapra grimmia). 2008. Afri Wil
Services.

Forsyth D. M.; Comte S.; Davis N.E.; Bengsen A.J.; Côté S. D.; Hewitt D. G.; Morellet N.;
Mysterud A. 2021. Methodology matters when estimating deer abundance: a global systematic
review andre commendations for improvements. https://doi.org/10.1002/jwmg.22207

Frederick H; Kohi E; Jay Lorenzo E; Coyote M; Schmitt T; Larsen K; 2015. Improving Flight Accuracy for
Aerial Wildlife Surveys in Sub-Saharan Africa. London, United Kingdom. ACM 978-1-4503-3490-7/15/11.
DOI: http://dx.doi.org/10.1145/2830629.2835221

Franklin, J; Miller, J. A. 2014. Mapping species distributions: Spatial inference and prediction.
Cambridge University Press. 978-0-521-70002-3

Kauffman JB, Coleman G, Otting N, Lytjen D, Nagy D, Beschta RL (2022) Riparian vegetation
composition and diversity shows resilience following cessation of livestock grazing in
northeastern Oregon, USA. PLoS ONE 17(1): e0250136.
https://doi.org/10.1371/journal.pone.0250136

Keeping, D., Pelletier, R., (2014). Animal Density and Track Counts: Understanding the Nature
of Observations Based on Animal Movements. PLoS ONE 9(5): e96598.
doi:10.1371/journal.pone.0096598

Kebede, Y., (2017). Distribution, Ecology and Status of Golden Jackal (canis aureus) in Africa.
Journal of Natural Sciences Research. ISSN 2224-3186 (Paper) ISSN 2225-0921. Vol.7.

Liebenberg, L. 2021. Assessing and Certifying Indigenous Tracking Expertise and Skills.
UNESCO Local and Indigenous Knowledge Systems (LINKS) section within the UNESCO
Natural Sciences Sector.. Available at
https://en.unesco.org/sites/default/files/links_indigenous_tracking_20210913.pdf

Li, X.; Li, N.; Li, B.; Sun, Y.; Gao, E. AbundanceR: A NovelMethod for Estimating Wildlife
Abundance Based on DistanceSampling and Species Distribution Models. Land 2022, 11, 660.
DOI: 10.3390/land11050660.

Lipilichi Investments. 2021. Chipanje Chetu Program: Stakeholder Review. Niassa Province,
Mozambique. 17p.

Mekonen, S. Coexistence between human and wildlife: the nature, causes and mitigations of
human wildlife conflict around Bale Mountains National Park, Southeast Ethiopia. BMC
Ecol 20, 51 (2020). https://doi.org/10.1186/s12898-020-00319-1

Mann, G. K. H., O’Riain, M. J., & Parker, D. M. (2020). A leopard’s favourite spots: Habitat
preference and population density of leopards in a semi-arid biodiversity hotspot. Journal of Arid
Environments, 181, 104218. doi:10.1016/j.jaridenv.2020.104218

Mondal, K., Sankar, K., Qureshi, Q., 2012. Factors influencing the distribution of leopard in a
semiarid landscape of Western India. DOI: 10.1007/s13364-012-0109-6

MWORIA JK, KINYAMARIO JI, GITHAIGA JM. Influence of cultivation, settlements and
water sources on wildlife distribution and habitat selection in south-east Kajiado,
Kenya. Environmental Conservation. 2008;35(2):117-124. doi:10.1017/S0376892908004670
Parsons, A. W; Wikelski, M; Wolff, B. K. V; Dodel, J. & Kays, R. (2022). Intensive hunting
changes human-wildlife relationships. doi: 10.7717/peerj.14159

Pappas, L. A. Mammalian Species. Taurotragus oryx. 2002. American Society of Mammalogists

Renwick AR, Robinson CJ, Garnett ST, Leiper I, Possingham HP, Carwardine J (2017) Mapping
Indigenous land management for threatened species conservation: An Australian case-study.
PLoS ONE 12(3): e0173876. https://doi.org/10.1371/journal.pone.0173876.

Sileshi, G., Hailu, G., & Nyadzi, G. I. (2009). Traditional occupancy–abundance models are
inadequate for zero-inflated ecological count data. Ecological Modelling, 220(15), 1764–
1775. doi:10.1016/j.ecolmodel.

Skalski JR, Richins SM, Townsend RL (2018) A statistical test and sample size recommendations for
comparing community composition following PCA. PLoS ONE 13(10): e0206033.
https://doi.org/10.1371/journal.pone.0206033

Stephenson, P. J. 2019. Integrating remote sensing into wildlife 1021 monitoring for
conservation. Environmental Conservation. doi.org/10.1017/S0376892919000092

Stegmann LF, Leitão RP, Zuanon J, Magnusson WE (2019) Distance to large rivers affects fish
diversity patterns in highly dynamic streams of Central Amazonia. PLoS ONE 14(10):
e0223880. https://doi.org/10.1371/journal. pone.0223880

Skinner, J.D. e Chimimba, C.T. 2005. Mammals of the Southern African Subregion (3rd edn).
Cambridge University Press, Cape Town.

Sitoe, A., e Maússe, S. S. (s/d.). Construindo Parcerias Florestais: potencial das reservas
florestais na redução do desmatamento com participação das comunidades locais.

Sianga, K; Richard W.S. Fynn; Mpaphi C. Bonyongo, (2017). Seasonal habitat selection by
African buffalo Syncerus caffer in the Savuti-Mababe-Linyanti ecosystem of northern Botswana.
http://dx.doi.org/10.4102/koedoe.v59i2.1382
Turpie, J.K. & Babatopie, A. 2018. The socio-economic impacts of livestock predation and its
prevention In South Africa. In: Livestock predation and its management in South Africa: a
scientific assessment (Eds Kerley, G.I.H., Wilson, S.L. & Balfour, D.). Centre for African
Conservation Ecology, Nelson Mandela University, Port Elizabeth, 53-81. Tolcha, A. Ayechew,
B. Shibru, S. (2019). feeding ecology of swayne’s hartebeest (alcelaphus buselaphus swaynei,
sclater, 1892) in maze national park. GSJ: Volume 7, Issue 12.ISSN 2320-9186.
Tsheboeng, G. Spatial variation of the influence of distance from surface water on riparian plant
communities in the Okavango Delta, Botswana. Ecol Process 7, 32 (2018).
https://doi.org/10.1186/s13717-018-0140-x

Verdade LM. The influence of hunting pressure on the social behavior of vertebrates. Rev Bras
Biol. 1996 Feb;56(1):1-13. PMID: 8731558.

Yao, K. A., Bitty, E. A., Kassé, K. B., Kouakou, Y. C., Yaokokoré-Beibro, K. H., Bi, S. G., &
McGraw, W. S. (2017). Distribution and relative abundance of forest duikers in Dassioko Sud
Forest Reserve (coastal Côte d’Ivoire). Wildlife Research, 44(8), 660. doi:10.1071/wr17068

Williams HF, Bartholomew DC,Amakobe B, Githiru M. Environmental factors affecting

thedistribution of African elephants in the Kasigau wildlifecorridor, SE Kenya.Afr J Ecol.
2018;56:244–253.https://doi.org/10.1111/aje.12442WILLIAMSET AL.253.

YVES A.K; ABDOULAYE D; ROGER M. G . C ; EMMANUELLE N; HJALMAR S. K.K; &

CHRISTOPHE B; (2019). Effects of anti-poaching patrols on the distribution of large mammals
in Taï National Park, Côte d'Ivoire. DOI: https://doi.org/10.1017/S0030605317001272