Moderate Levels of Human Disturbance Impact Persecuted Mammals in the Peruvian Amazon Rainforest

doi:10.21203/rs.3.rs-3150260/v1

Download PDF

Research Article

Moderate Levels of Human Disturbance Impact Persecuted Mammals in the Peruvian Amazon Rainforest

https://doi.org/10.21203/rs.3.rs-3150260/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Human population growth and associated land conversion in tropical regions have led to habitat fragmentation, resource extraction, and increased hunting pressure, resulting in defaunation. To assess the impacts of human disturbances on 17 mammal species spanning both predator and prey guilds, we conducted an eight-year (2015–2022) camera trap study in the unprotected region of Madre de Dios, Peru. Analyzing 14,849 mammal captures across 293 sites with occupancy models, we assessed the influence of eight ecological and anthropogenic covariates on species occupancy and detection. We hypothesized that the most persecuted species would be most negatively affected by human disturbance. Our results supported this hypothesis; persecuted species including jaguars (Panthera onca), tapirs (Tapirus terrestris), brocket deer (Mazama spp.), collared peccaries (Pecari tajacu), and agoutis (Dasyprocta variegata) exhibited significant changes in occupancy probability in relation to distance from settlements, proximity to agriculture, and land use type. Contrary to our prediction, ocelot and white-lipped peccary occupancy probability did not change with human disturbance covariates; however, there were no detections of white-lipped peccaries in the first seven years of the study. We found that distance to the river was the most common ecological covariate where species occupancy was higher closer to the river. Given the importance of rivers as access points for humans, this emphasizes the significance of protecting riverine habitats in lowland Peruvian Amazonia. Our findings highlight the importance of considering species-specific responses and the need for targeted conservation measures to mitigate the negative effects of human activities on vulnerable and highly persecuted species.

camera traps

defaunation

ecosystem health

felids

occupancy modeling

Panthera onca

Throughout the tropics, rates of human population growth have reached unprecedented levels, leading to rapid land conversion fueled by natural resource-based economies in many developing countries (Beaudrot et al., 2016). This conversion is primarily driven by extractive activities, agricultural expansion, and the construction of roads, which collectively contribute to habitat fragmentation, resource extraction, and increased hunting pressure (Espinosa et al., 2014; Gil-Sánchez et al., 2021). A significant surge in land use and land cover change has been observed in the Amazon rainforests of Brazil and Peru, particularly in proximity to newly established road networks branching from the Interoceanic Highway (IOH) (Chávez Michaelsen et al., 2013).

The impact of human activities on wildlife extends beyond habitat destruction and fragmentation. Humans can directly increase mortality rates via hunting and car strikes, for example, but they can also disrupt important cues used by species to assess predation risk, such as light and noise, affecting their overall landscape of fear (Mendes et al., 2020). Consequently, species are often forced to modify their foraging patterns and temporal activity in response to human disturbances, which can have significant implications for their fitness and overall population dynamics. Moreover, it has been observed that wildlife abundance is more closely related to patterns of hunting rather than traditional ecological variables like habitat area or forest type (Benítez-López et al., 2017); which highlights the crucial role of hunting and other human activities in shaping mammal populations. Higher landscape accessibility to hunters has been linked to lower occurrence and biomass of game species as well as apex predators. For instance, jaguar (Panthera onca) density was found to be up to 18 times higher in a more remote site compared to the most accessible site within a biosphere reserve in the rainforests of Ecuador (Espinosa et al., 2018). Additionally, a major contributor to the decline of predators is the decline of prey availability (Boron et al., 2019; Fuller & Sievert, 2001), mainly through anthropogenic habitat loss and hunting. Studies have found that the fine-scale habitat use of jaguar and puma (Puma concolor) was best explained by prey availability (Palomares et al., 2016; Santos et al., 2019); moreover, larger-bodied prey with higher fecundity determined the carrying capacity of top predators in these tropical systems (Terborgh & Estes, 2010). Prey availability as a factor of habitat quality therefore plays a critical role in the occupancy and persistence of predators. Only by incorporating the diverse geographic, ecological, and anthropogenic dynamics at play can we holistically predict trends in the spatial patterns of wildlife.

As a consequence of multifarious anthropogenic activities, defaunation—the loss of animal populations ranging from local declines to species extinctions—has become most pronounced in tropical ecosystems, leading to an alarming projected loss of biodiversity (Beaudrot et al., 2016). Notably, mammal abundance has declined by 83% across tropical regions (Benítez-López et al., 2017). These declines can be traced back to the historical impact of human hunters, which have persisted over time and resulted in population declines and near-extinctions of large-bodied mammal species, primarily in tropical ecosystems (Benítez-López et al., 2017; Jorge et al., 2013). The consequences of defaunation can have far-reaching effects on the structure and functioning of entire ecosystems, including disruptions in dispersal mutualisms and declines in total biomass (Benítez-López et al., 2017; Gil-Sánchez et al., 2021).

Despite the ecological significance of highly persecuted felid and game species, the impact of less visible activities such as hunting and wildlife extraction remain poorly studied, and the majority of population dynamics research occurs inside protected areas (Espinosa et al., 2014, 2018). Wildlife abundance tends to be higher within protected areas, highlighting the importance of these designated regions in supporting diverse and thriving populations (Benítez-López et al., 2017) but also the importance of monitoring wildlife populations in human-modified landscapes. Despite efforts to protect larger, relatively undisturbed areas, ecological processes within these protected regions are gradually eroding, often without apparent signs (Jorge et al., 2013). Assessing the intactness of natural areas becomes crucial, and indicator species play a vital role in signifying the presence or absence of these ecological interactions. By focusing on these flagship species, conservation efforts can prioritize areas for protection and implement targeted strategies to mitigate further biodiversity declines.

Indicator species are generally large carnivores and herbivores that possess unique ecological roles, are easily detectable, and their status can represent the overall health and integrity of ecosystems (Caro, 2010; Jorge et al., 2013). Large mammal assemblages have been defined as rare evolutionary and ecological phenomena and one of the most crucial layers of conservation to optimize global biodiversity protection and stabilize the climate (Dinerstein et al., 2020). Large mammals are often long-lived, slow to reproduce, and have wide-ranging habitat requirements, making them more susceptible to human disturbances (Ripple et al., 2014); therefore, the decline or disappearance of these species can signal significant ecological disruptions, including changes in trophic interactions, seed dispersal patterns, and habitat quality (D’Cruze et al., 2021; Jorge et al., 2013). For example, a decline in key seed dispersers has been shown to decrease populations of plants, insects like dung beetles, amphibians, and insectivorous birds (Jorge et al., 2013). Expectedly, species that serve as indicators for biodiversity are generally the most highly persecuted species for similar reasons (Di Minin et al., 2016; Morrison et al., 2007); they are large/ easily detected, serve as resources competitors for humans, and are profitable for their meat, parts, or as pets (D’Cruze et al., 2021; Jorge et al., 2013; Morrison et al., 2007). Since indicator species and those most highly persecuted are the first to vanish from fragmented and hunted forests (D’Cruze et al., 2021; Jorge et al., 2013), they serve as strong indicators of species richness and ecosystem health.

In addition to traditional bushmeat subsistence hunting, Peru has emerged as a hotspot for the wildlife trade in Latin America, with widespread exploitation and poorly regulated exportation of live wild animals and parts (D’Cruze et al., 2021). Among the most commercially valuable bushmeat species in Peru are the white-lipped peccary (Tayassu pecari), collared peccary (Pecari tajacu), brocket deer (Mazama spp.), the lowland tapir (Tapirus terrestris), paca (Cuniculus paca), and agouti (Dasyprocta spp.) (Espinosa et al., 2014). In forests without hunting, white-lipped peccaries in particular account for the highest biomass of neotropical ungulates; 200-peccary herds can have a biomass greater than 4.8 tons (Peres, 1996). The sale of wildlife is primarily driven by the demand for wild meat, followed by the pet trade, posing further threats to these species (D’Cruze et al., 2021). The declining abundance of these herbivores, particularly terrestrial seed predators and dispersers, has far-reaching repercussions, including a decline in plant and amphibian diversity, the reduction of dung beetle diversity, the decrease in insect fauna associated with litter, and the loss of insectivorous birds (Jorge et al., 2013). Wild cats such as the jaguar and the ocelot (Leopardus pardalis) are also impacted by illicit trafficking, which exploits both live animals and their body parts, including their skins. Pre-CITES (Convention on International Trade in Endangered Species of Wild Fauna and Flora) jaguars and ocelots were overexploited with 228,376 jaguar skins and 17,301 ocelot skins harvested across their ranges, which decimated populations (Mena et al., 2021). Felid declines have consequences not only for the population dynamics of their prey, but the subsequent decline or alteration in ecosystem structure and function (Mena et al., 2021).

Our study focused on understanding the occupancy and site use probability of 17 mammal species, including the felid guild and their prey, within an unprotected landscape in Peru. We included ecological and anthropogenic covariates and considered landscape accessibility to hunters by incorporating distances between camera stations and settlements and agricultural areas. Though the Las Piedras region is experiencing moderate levels of deforestation and habitat conversion compared to other areas within Madre de Dios (MDD) and Peruvian Amazonia, we hypothesized that mammal occupancy would still be shaped by anthropogenic threats and activities in our study area. We predicted that the highly persecuted species, including the two most persecuted cats (jaguar and ocelot) and the six prey species of the highest market value (lowland tapir, white-lipped peccary, brocket deer, collared peccary, paca, and agouti), will exhibit the most significant responses to human disturbances. Our goal was to assess if moderate levels of human impact is intensifying defaunation in a biodiversity hotspot, potentially affecting ecosystem interactions across human-modified landscapes within the Amazon rainforest.

Study Area

The southern part of the Peruvian Amazon lies within the department of MDD, one of the largest unprotected and intact tropical rainforests in Peru, which has the second largest extent of rainforest in the Amazon, and still consists of largely contiguous forests (Asner et al., 2010). The area contains four national protected areas of over one million hectares each, including the Manu National Park, the Alto Purus National Park, the MDD Territorial Reserve, and the Tambopata/Bahuaja Sonene National Reserve, as well as several large indigenous reserves, conservation concessions, and private protected lands (Fig. 1). The region is known for its diversity and abundance of fauna; however, it is also a mosaic of concessions with registered activities that often do not reflect actual land use and continues to face increasing pressures of logging and clear-cut agriculture. The study was carried out at the Hoja Nueva research center (-12.127, -69.476) along the Las Piedras River, 70 km northwest of Puerto Maldonado. The four study areas included approximately 550 km² of lowland rainforest representative of diverse land cover and land use (Fig. 1).

Focal Species

Our study focused on five felid species and 12 prey species, many of which are either data deficient or threatened with extinction locally and/or globally (Table 1). The five sympatric felid species in our region include the jaguar, puma, ocelot, jaguarundi (Herpailurus yagouaroundi), and the margay (Leopardus wiedii). The 12 terrestrial prey species we included in our study included the lowland tapir, giant armadillo (Priodontes maximus), giant anteater (Myrmecophaga tridactyla), white-lipped peccary, brocket deer, collared peccary, Dasypus armadillos, paca, agouti, common opossum (Didelphis marsupialis), green acouchi (Myoprocta pratti), and the Brazilian rabbit (Sylvilagus brasiliensis).

Table 1

Data for all 17 focal species including their weight range, ecological role, IUCN red list status, and Peru Red Book status. Cells shaded with orange represent the most highly persecuted species in this system that we predicted would respond significantly to human disturbances.
Species	Weight Range (kg)	Ecological Role	IUCN Status	Peru Red Book Status	Citations
Jaguar (Panthera onca)	30–120	Hypercarnivore / Apex Predator	Near Threatened	Near Threatened	(Emmons & Feer, 1999; IUCN, 2022; SERFOR, 2018)
Puma (Puma concolor)	22–80	Hypercarnivore / Apex Predator	Least Concern	Near Threatened	(IUCN, 2022; Moreno et al., 2006; SERFOR, 2018)
Ocelot (Leopardus pardalis)	8–18	Hypercarnivore / Mesopredator	Least Concern	Unlisted	(IUCN, 2022; Moreno et al., 2006)
Jaguarundi (Herpailurus yagouaroundi)	4.5–9	Hypercarnivore / Mesopredator	Least Concern	Unlisted	(IUCN, 2022; Santos et al., 2019)
Margay (Leopardus wiedii)	3–8	Hypercarnivore / Mesopredator	Least Concern	Data Deficient	(IUCN, 2022; Santos et al., 2019; SERFOR, 2018)
Lowland Tapir (Tapirus terrestris)	150–250	Herbivore / Seed Disperser	Vulnerable	Near Threatened	(IUCN, 2022; Padilla & Dowler, 1994; SERFOR, 2018)
Giant Armadillo (Priodontes maximus)	28–44	Insectivore / Ecosystem Engineer	Vulnerable	Vulnerable	(IUCN, 2022; SERFOR, 2018; Silveira et al., 2009)
Giant Anteater (Myrmecophaga tridactyla)	27–39	Insectivore	Vulnerable	Vulnerable	(IUCN, 2022; Medri & Mourão, 2005; SERFOR, 2018)
White-Lipped Peccary (Tayassu pecari)	25–35	Omnivore / Seed Predator / Ecosystem Engineer	Near Threatened	Near Threatened	(Kiltie & Terborgh, 1983)
Brocket Deer (Mazama spp.)	11.4–30	Herbivore / Seedling Predator	Data Deficient	Data Deficient	(Emmons & Feer, 1999; IUCN, 2022; Richard-Hansen et al., 1999; SERFOR, 2018)
Collared Peccary (Pecari tajacu)	15–25	Herbivore / Seed Predator / Ecosystem Engineer	Least Concern	Unlisted	(IUCN, 2022; Kiltie & Terborgh, 1983)
Armadillo (Dasypus spp.)	3–13	Insectivore / Ecosystem Engineer	Data Deficient	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)
Paca (Cuniculus paca)	5.2–9.5	Herbivore / Seed Predator	Least Concern	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)
Agouti (Dasyprocta variegata)	3–5.8	Herbivore / Seed Predator	Least Concern	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)
Common Opossum (Didelphis marsupialis)	0.6–2	Omnivore / Scavenger	Least Concern	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)
Green Acouchi (Myoprocta pratti)	0.6–1.8	Herbivore / Seed Predator	Least Concern	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)
Brazilian Rabbit (Sylvilagus brasiliensis)	0.7–1	Herbivore	Least Concern	Unlisted	(IUCN, 2022; Richard-Hansen et al., 1999)

Camera Trapping

Within each study area, we placed 1–5 camera trap arrays between 2015–2022, resulting in eight sampling grids and a total of 293 camera stations. Cameras were spaced equidistantly for each grid, with an average distance between stations of 1.22 km. We based the spacing of six of the sample grids (2015–2022) on the ocelot, which has an average home range of 38.8 km² for males and 17.4 km² for females (Crawshaw, 1995). In 2022, we used smaller camera trap spacing in the remaining two sample grids to increase the detection of the margay. All grids except the two from 2022 accommodated at least one home range of our largest focal species, the jaguar, which is known to have a home range of 128.61 ± 49.5 km² across Central and South America; however, jaguar home range size is known to be markedly reduced in lowland rainforest (Gonzalez-Borrajo et al., 2017).

One of the four study areas surveyed consistently was the Las Piedras Conservation Corridor (LPCC) in the years 2018, 2019, 2021, and twice in 2022 for a total of 167 camera stations (see Area A in Fig. 1). The three remaining study areas were unprotected and included a variety of human disturbances typical to the Las Piedras region. Firstly, a network of logging roads between the IOH and the Las Piedras River was surveyed in 2015 (30 stations; Area B). A remote indigenous settlement of subsistence hunters was surveyed in 2017 (46 stations; Area C). Lastly, a cacao agricultural association south of the LPCC was surveyed in 2021 (50 stations; Area D). All areas were sampled for approximately 90 days in the dry or dry-wet transition seasons between the months of July and December, minimizing the possibility that camera traps would be lost in the wet season due to failure from weather. The two macro-habitat types that were surveyed included terra firme (upland) forest and floodplain forest; the latter is a forest that is seasonally inundated during the wet season (November – April).

First, existing trail and road networks were located and mapped in each sample area, followed by overlaying a grid to select camera trap sites after selecting the first camera site at random. Where accessible, wildlife trails were used at grid points if a road or trail could not be found within 100 m. For example, the grids in 2015 (logging road networks) and 2017 (remote indigenous settlement) were more elongated with larger gaps due to a lack of human trails and roads, so more cameras were placed on wildlife trails in these areas. Throughout the study areas, 153 camera stations were placed on human trails or roads, and 140 were placed on wildlife trails.

We used Browning Strike Force trail cameras (BTC-5 and BTC-6 models), which trigger with movement at 20 m with an open view. Camera stations were made up of one or two cameras placed facing a trail, road, or wildlife trail, approximately 40–50 cm off the ground, set to capture videos.

Covariates

To investigate the effects of both ‘bottom-up’ ecological variables and human disturbance on mammal occupancy, we incorporated a suite of eight ecological and anthropogenic covariates (see Table 2 for details). These covariates were selected to capture various aspects of the landscape, including macrohabitat (floodplain or terra firme forest), indices representing large prey availability and small prey availability, enhanced vegetation index (EVI), distance from the Las Piedras River, distance from the nearest settlement, proximity to agriculture, and land use. The indices of large and small prey were specifically used as proxies for prey availability and habitat quality and were considered when modeling predator occupancy. Additionally, to account for landscape accessibility to hunters, we incorporated covariates mirrored from the work of Espinosa et al. (2018), such as distances between survey stations and both settlements agricultural areas. We included trail type (roads, human trails, or wildlife trails) as a covariate on detection probability (Harmsen et al., 2010).

Table 2

Descriptions of the covariates placed on detection and occupancy for the 17 mammal species across the four study areas, including two covariates on detection and eight covariates on occupancy.
	Covariate	Description
Detection	Trail Type	Trails were divided into two types based on their size and use by humans and wildlife: human trails/roads and wildlife trails.
	Trap Nights Operable	The number of nights operable for each camera site.
Occupancy	Macrohabitat	Macrohabitat refers to the general biome of each site and is defined by one of two categories: floodplain forest or terra firme forest. Floodplain forest represents seasonally inundated forests, usually in closer proximity to rivers and streams. Terra firme forest represents upland forest macrohabitat. A Digital Elevation Model (DEM) was obtained from the local forestry department in Puerto Maldonado in 2015 to identify these two macrohabitats using elevation and a detailed map of the transition zones in each survey area.
	Small Prey Index	The index was calculated as the number of captures of prey < 5 kilos divided by the total number of trap nights per site.
	Large Prey Index	The index was calculated as the number of captures of prey > 5 kilos divided by the total number of trap nights per site.
	Enhanced Vegetation Index	Enhanced Vegetation Index (EVI) was obtained from the United States Geological Survey’s EarthExplorer website (earthexplorer.usgs.gov) using the MOD13A1 Version 6 product (Didan, 2015). Imagery was downloaded for the week before each sample grid was surveyed. An EVI scale of 500m was used to quantify vegetation greenness around each site.
	Distance from River	The Las Piedras River was tracked via handheld GPS three times to get an accurate river layer. The distance between each camera trap station and the river was then measured using the “near” tool in ArcMap 10.4’s Proximity toolset (ESRI, Redlands, USA).
	Distance from Settlement	GPS coordinates were taken in the center of both Lucerna and Puerto Nuevo. Distance between each camera trap station and the nearest settlement was measured using the “near” tool in ArcMap 10.4’s Proximity toolset (ESRI, Redlands, USA).
	Proximity to Agriculture	All surrounding farmland was mapped manually using a handheld GPS unit whenever active farms were located within or near the study area. Proximity to agriculture was recorded as a category: whether a camera trap station was located within 500 meters of a farm greater than 4 hectares in size or not.
	Land-Use	We created an index of human land-use activity categorized as Low or High. Low land-use areas included conservation lands protected by conservation and eco-tourism NGOs within the LPCC. High-land use included settlements, agriculture areas, and Brazil nut harvesting concessions outside of the LPCC, including a number of sites where vehicles, hunting, and timber extraction were recorded.

We examined correlations between all continuous variables, and none were high enough to preclude independent inclusion in the subsequent occupancy analyses (all r > |0.7|). All continuous covariates (5 total) were standardized to have a mean of 0 and a standard deviation of 1 for analysis, and all categorical variables were listed as factors. The most common classification within each of the three categorical variables (e.g., “conservation” within land use) was set to be the default category (i.e., included in the intercept).

Data Analysis

All data analyses were carried out using R (R Core Team, 2022) in RStudio (RStudio Team, 2020). The package ‘camtrapR’ (Niedballa et al., 2016) was used to create encounter histories for each species based on collected camera trap footage and the given sample occasions indicated for each focal species (Table 2). We selected the length of the sampling occasions (i.e., 5 or 10 days) based on the overall number of captures per species. Data from all eight camera grids were combined such that each camera station was considered a unique site in the model, resulting in 293 sites.

We used the R package ‘unmarked’ (Fiske & Chandler, 2011) to fit single-season single-species occupancy models (MacKenzie, 2006) for each of the 17 species. We first fit a global model to each species, including all ten covariates for each felid species and eight covariates (all but the two prey indices) for the other mammals. The function “dredge” within the package ‘MuMIn’ (Bartoń, 2009) was used to fit all combinations of covariates from the chosen global model per species. The top model was selected based on the lowest AIC (Akaike information criterion) value (Akaike, 1973), and results were presented from that model for each species. Owing to variations in detection probability and sample sizes, not all covariates could be used for all species. If the global model for a species resulted in a Hessian error (i.e., an indicator of overfitting), we then fit the ecological covariates separately from the anthropogenic covariates. Subsequently, we moved forward with a reduced global model that included any covariates that featured in the top models from those two pools of covariates. We found that four species required a reduced global model: jaguarundi, margay, giant armadillo, and white-lipped peccary. The presence of anthropogenic covariates in the top models of persecuted species (and their absence for species that are not hunted) would support our predictions. We considered parameter estimates with a p-value of less than 0.1 to be significant.

Following model selection, the mean probabilities of both detection and occupancy were then calculated using the ‘predict’ function within ‘unmarked’ and averaged across all 293 camera sites. Due to the large home range and dispersal of jaguars and pumas relative to the camera spacing, which violates the assumption of site independence, we interpreted jaguar and puma results as the probability of site use rather than occupancy.

Out of a total of 14,849 detections of these 17 mammals, the most highly photo-captured mammals were brocket deer (3016 captures) and agouti (2998 captures) whereas the least captured mammals were white-lipped peccary (34 captures) and jaguarundi (35 captures) (Table 3). Mean probability of detection ranged between 0.055 and 0.397, and the mean probability of occupancy ranged between 0.182 and 0.931. Distance from the river, land use, and macro-habitat were the most frequent covariates amongst the top models. Both trail type and trap nights operable were common in the top models, explaining the detection probability of 71% of studied species (Table 3).

Table 3

Single-species occupancy results for the 17 focal species including total captures from camera traps, individual sample occasions used in detection histories, naïve occupancy, covariates for detection and occupancy in each top model, and the resulting probabilities of both detection and occupancy driven by the model covariates. Cells shaded with orange represent the species whose occupancy probability was significantly lower in areas with higher human disturbance.
Species	Total Captures	Sample Occasion (Days)	Naïve Occupancy	Top Model		Mean Probability of Detection	Mean Probability of Occupancy
Species	Total Captures	Sample Occasion (Days)	Naïve Occupancy	Detection	Occupancy	Mean Probability of Detection	Mean Probability of Occupancy
Jaguar (Panthera onca)	231	5	0.358	Trail + Operable	Habitat + Agriculture + Land Use	0.065	0.611
Puma (Puma concolor)	471	5	0.461	Trail	Prey + River	0.130	0.638
Ocelot (Leopardus pardalis)	1181	5	0.700	Trail	Habitat + Agriculture	0.211	0.824
Jaguarundi (Herpailurus yagouaroundi)	35	10	0.085	-	Prey + Small Prey	0.080	0.228
Margay (Leopardus wiedii)	139	5	0.273	-	EVI + River	0.063	0.441
Lowland Tapir (Tapirus terrestris)	1412	5	0.761	Trail + Operable	EVI + Habitat + Agriculture	0.257	0.787
Giant Armadillo (Priodontes maximus)	92	10	0.208	-	River + Settlement	0.059	0.600
Giant Anteater (Myrmecophaga tridactyla)	117	10	0.249	-	EVI + River + Settlement	0.087	0.532
White-Lipped Peccary (Tayassu pecari)	34	10	0.072	Trail + Operable	-	0.055	0.278
Brocket Deer (Mazama spp.)	3016	5	0.918	Trail	Habitat + Land Use	0.395	0.931
Collared Peccary (Pecari tajacu)	1726	5	0.826	Trail + Operable	River + Settlement + Land Use	0.346	0.848
Armadillo (Dasypus spp.)	439	5	0.375	-	EVI + Habitat + River	0.144	0.430
Paca (Cuniculus paca)	1348	5	0.594	Trail	-	0.274	0.613
Agouti (Dasyprocta variegata)	2998	5	0.812	Trail	Settlement + Land Use	0.397	0.825
Common Opossum (Didelphis marsupialis)	945	5	0.549	Trail	Habitat + River + Agriculture + Land Use	0.218	0.586
Green Acouchi (Myoprocta pratti)	446	5	0.294	Operable	Land Use	0.158	0.337
Brazilian Rabbit (Sylvilagus brasiliensis)	219	5	0.116	Trail	EVI + River + Land Use	0.148	0.182

We found substantial support for our prediction that the most persecuted species would respond to anthropogenic pressures; five of the eight species expected to be most sensitive to human disturbance experienced significant changes in occupancy probability in response to a covariate that reflected either direct or indirect human impact including the jaguar, lowland tapir, brocket deer, collared peccary, and the agouti. Of the nine species that are less persecuted within the study area, only two responded negatively to any of the covariates reflecting human disturbance. Overall, we observed that covariates characterizing habitat area/ forest type and resources, such as macro-habitat, EVI, distance to river, and the two prey indices, affected the occupancy probability of 65% of our focal species. Supporting our hypothesis, covariates reflective of deforestation and human impacts described the occupancy patterns of 47% of our focal species.

No consistent patterns emerged across the felid guild (Table 4). Jaguars and ocelots had proximity to agriculture in their top models, supporting our prediction that they would have a more heightened response to human disturbance than pumas, jaguarundis, and margays. While jaguar site use probability was lower in areas closer to agriculture, results for ocelots were not significant. Additionally, jaguar site use probability increased in mixed-use areas. Ocelot occupancy probability increased in floodplain forest compared to terra firme forest, and puma probability of use decreased with distance from the river. Interestingly, jaguarundi probability of occupancy increased in areas where both the small and large prey indices were higher, and margay occupancy increased at sites with a high EVI.

Table 4

Resulting estimates and standard errors from the top models of the five felid species, organized by the relative size of the species. Cells with a dash represent covariates that were not present in the top model. Orange shaded cells represent a persecuted species with a significant negative response to human disturbance. Detection intercept includes off-trail (wildlife trails) for trail type. Occupancy intercept includes floodplain for macro-habitat, > 500m for proximity to agriculture, and low land use (conservation corridor). All estimates are shown on the logit-scale. Significance: * p < 0.10.
Species	Detection Intercept	Trail (On)	Trap Nights Operable	Occupancy Intercept	Terra Firme Macro-Habitat	Big Prey Index	Small Prey Index	EVI	Distance from River	Agriculture < 500 m	Mixed Land Use
Jaguar (Panthera onca)	-337(0.18)*	0.95(0.19)*	0.26(0.11)*	0.499(0.41)	-0.77(0.50)	-	-	-	-	-2.23(1.33)*	2.62(1.3)*
Puma (Puma concolor)	-3.09(0.17)*	1.76(0.18)*	-	0.58(0.18)*	-	0.26(0.18)	-	-	-0.33(0.17)*	-	-
Ocelot (Leopardus pardalis)	-2.26(0.10)*	1.50(0.11)*	-	2.56(0.46)*	-1.32(0.49)***	-	-	-	-	-0.82(0.53)	-
Jaguarundi (Herpailurus yagouaroundi)	-2.44(0.34)*	-	-	-1.61(0.40)*	-	0.50(0.28)*	0.45(0.24)*	-	-	-	-
Margay (Leopardus wiedii)	-2.69(0.14)*	-	-	-0.25(0.23)	-	-	-	0.31(0.18)*	0.26(0.17)	-	-

The four prey species that are both highly persecuted and had a significant response to human impact variables included tapir, brocket deer, collared peccary, and agouti, and land use was the most common anthropogenic variable in prey top models (Table 5). The occupancy probability of acouchi, brocket deer, collared peccary, and common opossum significantly decreased in mixed-use areas outside of the protected conservation corridor; only the occupancy probability of Brazilian rabbits increased in this matrix habitat compared with low land use by humans. One species, the common opossum, had a higher probability of occupancy near agricultural areas, whilst the occupancy of the lowland tapir decreased significantly at sites within 500 m of agriculture. Only agouti had a significant increase in occupancy in response to an increasing distance from settlements. The occupancy probability of the following five species significantly decreased with distance from the river: giant anteater, Brazilian rabbit, collared peccary, common opossum, and Dasypus armadillos. While the occupancy probability of brocket deer and Dasypus armadillos increased in terra firme (upland) forest, the lowland tapir had higher occupancy in seasonally flooded forest. Lastly, increased greenness (EVI) had a significant positive effect on the occupancy probability of the giant anteater and lowland tapir.

Table 5

Coefficient estimates and standard errors from the top models of the 12 prey species, organized by the relative size of the species. Dashed cells represent covariates that were not present in the top model. Orange shaded cells represent a persecuted species with a significant negative response to human disturbance. Detection intercept is for wildlife trails for the trail type. Occupancy intercept includes floodplain for macro-habitat, > 500m for proximity to agriculture, and low land use (conservation corridor). All estimates are shown on the logit-scale. Significance: * p < 0.10.
Species	Detection Intercept	Trail (On)	Trap Nights Operable	Occupancy Intercept	Terra Firme Macro-Habitat	EVI	Distance from River	Dist. from Comm.	Agriculture < 500 m	Mixed Land Use
Lowland Tapir (Tapirus terrestris)	-1.20(0.06)*	0.25(0.08)*	-0.18(0.05)*	2.47(0.38)*	-1.49(0.41)*	0.35(0.16)*	-	-	-1.02(0.44)*	-
Giant Armadillo (Priodontes maximus)	-3.05(0.20)	-	-	2.37(2.18)	-	-	-3.13(2.24)	2.07(1.57)	-	-
Giant Anteater (Myrmecophaga tridactyla)	-2.35(0.19)*	-	-	0.16(0.40)	-	0.81(0.27)*	-0.73(0.34)*	0.48(0.31)	-	-
White-Lipped Peccary (Tayassu pecari)	-2.43(0.45)*	-1.81(0.59)*	-0.56(0.29)*	-0.96(0.53)*	-	-	-	-	-	-
Brocket Deer (Mazama spp.)	-0.28(0.05)*	-0.29(0.07)*	-	3.39(0.70)*	1.27(0.53)*	-	-	-	-	-2.08(0.73)*
Collared Peccary (Pecari tajacu)	-0.71(0.06)*	0.13(0.08)	-0.16(0.05)*	2.25(0.30)*	-	-	-0.57(0.24)*	0.42(0.26)	-	-0.83(0.38)*
Armadillo(Dasypus spp.)	-1.79(0.08)*	-	-	-0.73(0.26)*	0.71(0.35)*	0.22(0.14)	-0.58(0.21)*	-	-	-
Paca (Cuniculus paca)	-0.82(0.07)*	-0.30(0.09)*	-	0.46(0.12)*	-	-	-	-	-	-
Agouti (Dasyprocta variegata)	-0.65(0.06)*	0.43(0.07)*	-	1.38(0.21)*	-	-	-	0.33(0.19)*	-	0.49(0.33)
Common Opossum (Didelphis marsupialis)	-1.41(0.09)*	0.25(0.10)*	-	0.44(0.28)	0.49(0.33)	-	-0.37(0.18)*	-	0.82(0.44)*	-1.0(0.29)*
Green Acouchi (Myoprocta pratti)	-1.69(0.11)*	0.23(0.11)*	-	-0.31(0.18)*	-	-	-	-	-	-0.86(0.29)*
Brazilian Rabbit (Sylvilagus brasiliensis)	-3.56(0.42)*	2.50(0.43)*	-	-2.61(0.40)*	-	-0.35(0.23)	-1.12(0.40)*	-	-	1.40(0.45)*

In terms of detection probability, we found that three prey species – the lowland tapir and both peccary species – had a negative relationship with the number of trap nights. However, more trap nights significantly increased the detection probability of the jaguar and the green acouchi. Although brocket deer, paca, and white-lipped peccary had higher detection probability on wildlife trails, the following mammals generally had significantly higher detection on human trails and roads than on wildlife trails: lowland tapir, common opossum, agouti, Brazilian rabbit, jaguar, puma, and ocelot.

Our study yielded new insights into the influence of various ecological and anthropogenic variables on a community of felid predators and their prey in the Amazon rainforest. Responses to human impacts were strongest for the most highly persecuted mammals – both predators and game species – but were also very species-specific within the predator and prey guilds. In addition, mammals with less hunting pressure did not spatially avoid the more disturbed sites within our study area.

We did not observe any anthropogenic effects on one of the most persecuted species, white-lipped peccary (hereafter “white-lips”), because we did not capture them for seven years consecutively. Their absence is concerning given their ecological role, vulnerable status according to the IUCN, and near-threatened designation in Peru by SERFOR (IUCN, 2022; SERFOR, 2018). In contrast to the usual abundance of white-lips compared to collared peccary in Amazonian ecosystems, our study observed an abundance of collared peccary and no white-lips until 2022. White-lips typically dominate over collared peccaries, leading to lower occupancy probability for the latter at sites occupied by white-lips (Ferreguetti et al., 2017). They serve as ecosystem engineers, seed predators/ dispersers, and important prey for top predators like puma and jaguar (Ferreguetti et al., 2017; Kiltie & Terborgh, 1983). The disappearance of white-lips from large regions like MDD for over ten years initially led to the hypothesis of migration; however, a 15-year study in northern Brazil debunked this, showing that population declines were caused by disease outbreaks rather than migration (Fragoso, 2004). Disease outbreaks resulted in high mortality, with hunting further reducing the remaining individuals, while areas without hunting allowed for repopulation (Fragoso, 2004). Recolonization of herds from distant locations explains the 10–15-year process observed across the Amazon rainforest. The development of remote areas in the Las Piedras region, including road construction, settlement formation, and agriculture, has led to increased habitat fragmentation, hunting, and disease, likely contributing to the disappearance of white-lips. Our camera traps did not capture white-lips from 2014 to 2021, suggesting an epidemic after urbanization and the introduction of agriculture/livestock in the early 2000s. However, small groups were observed in person between 2017 and 2020 (D. J. Singer, personal communication, May 20, 2023), indicating that a few groups remained to repopulate the area; we then obtained our 35 records of white-lips on camera within the confines of the protected corridor in 2022. Encouragingly, in May 2023, our team witnessed a group of over 100 individuals, indicating population stabilization within the conservation corridor but unknown status outside of this privately protected area.

After white-lips, the most sought after bushmeat species include agouti, collared peccary, brocket deer, and lowland tapir, which all had significant responses to our anthropogenic covariates: distance from the closest settlement, proximity to agriculture, and land use type. Although agoutis are thought to be more adaptable to human-modified areas (Naughton-Treves et al., 2003), their negative relationship with settlements indicates a direct response to human presence and likely hunting pressure. Brocket deer and collared peccary both decreased in occupancy probability in mixed-use areas, which aligns with results from previous studies, confirming their heightened sensitivity to human disturbance (Nagy-Reis et al., 2017). Tapirs likely responded to agriculture in particular for several reasons: the presence of hunting, a decrease in food availability, and reduced access to water. Given that permanent households are prominent among the cacao and papaya farms and these households are isolated from both the closest settlement and the city, families rely upon the combination of their local crops and bushmeat to subsist. Tapirs are hunted with relative ease due to their heaviness and a method called “spotlighting” where their large eyeshine is spotted using flashlights at night. Additionally, clearcutting for agriculture has drastic, immediate effects on the floristic composition and vegetation structure at the forest edge and well into the remaining forest interior (Williams-Linera, 1990), decreasing interior species of the herbaceous layer that are vital to the tapir’s diet. Tapirs are semiaquatic and need consistent access to water bodies for various reasons including thermoregulation, to facilitate movement, and to escape predators (Burs et al., 2023). Drier and hotter forest edges cause increased forest desiccation and the disappearance of water bodies; canopy water loss has been known to extend over one kilometer into the forest in a landscape that has only been moderately fragmented (Briant et al., 2010). Matrix habitat or mixed-land use areas likely did not affect tapir occupancy because food and water availability remain high in these areas compared to landscapes of agriculture and forest fragments.

Due to their large home ranges and requirement for substantial contiguous areas to survive, jaguars are especially vulnerable to fragmentation and human development (Michalski & Peres, 2005; Zeilhofer et al., 2014). Our research shows that jaguars in the Las Piedras are using areas of mixed land-use outside of protected areas, but specifically avoiding cacao farms within the agricultural association that borders the LPCC. The agricultural association is predominantly made up of 5–20-hectare areas clearcut for cacao production, but oftentimes there are permanent households where domestic animals such as dogs, pigs, and chickens reside without fencing. One of several reasons that jaguar site use decreases in proximity to agriculture could be the higher likelihood of retaliation for livestock kills in these areas (Amador et al., 2013). Additionally, cacao farms likely do not provide sufficient vegetation cover and have lower occupancy of the jaguar’s main prey species, the lowland tapir. Mixed-use areas, which we defined as matrix habitat outside the protected LPCC, increased the site use probability of jaguars in our study area. Vehicles, hunting, and timber extraction were recorded at these sites, which were in closer proximity to human settlements, logging camps, and Brazil nut concessions. An increase in jaguar site use probability in these areas could be due to several factors: the extensive daily distance traveled by jaguars (McBride & Thompson, 2018), the use of habitual travel routes and paths of least resistance (Karanth & Nichols, 1998), their documented tolerance of human disturbance (e.g. ability to become more nocturnal in human-occupied areas) (Foster et al., 2010; Monette et al., 2020), and the relatively low level of human impact in these areas comparatively. Although these sites were outside of the protected corridor, the forest is generally intact and connected, and human activity is limited; therefore, jaguars still use these areas for navigation, at least to get from one zone of their range to another.

While most of our persecuted predator and game species showed a significant response to human impacts, the ocelot and the paca did not. Due to the extent to which ocelots are hunted (Valsecchi et al., 2023), usually in retaliation for livestock killings (Mena et al., 2021), we expected similar results to the jaguar with either distance from settlements or agriculture proving significant in affecting ocelot occupancy. Our results could indicate that ocelots are more capable of adapting to human disturbance; however, it could also be due to a lack of data from agricultural areas, as sites within 500 m of cacao or papaya agriculture only represented 12% of our total sites (in part due to camera theft near agricultural locations). This reduced sample size is likely causing us to miss patterns in species occupancy in relation to proximity to agriculture for not just ocelots but all of the species. Although the paca did not have a negative response to any human impact covariates, their significantly lower detection probability on human trails could represent an avoidance of humans at finer scales in response to hunting or could indicate they prefer smaller mammal trails. Additionally, paca are known to be a riverine species, and past studies have deployed camera traps within 100 m of a water source to effectively study their ecology (Figueroa-de-León et al., 2017). Therefore, the design of our study may be impeding our ability to detect patterns in the occupancy of paca given their habitat requirements and space use.

The Las Piedras River was a significant predictor of mammal occupancy across our study area. Distance from the closest river and vicinity to fresh water have been recorded to increase species occupancy in similar systems, within both the predator and prey guilds (Boron et al., 2019; Dias et al., 2019; Rich et al., 2017). The river and floodplain forest are crucial as mammal habitat; however, riverine habitat is also used by humans for the establishment of settlements and agriculture. The combination of abundant natural resources, transportation opportunities, and fertile soils make riverine habitats a primary choice for colonization in the Amazon rainforest (Lentz, 2000; Moran, 2007). Consequently, subsistence hunting is concentrated around farms and settlements and near the margins of rivers; distance from rivers has been used as a covariate to predict hunting intensity given it is a common access point for hunters (Espinosa et al., 2018; Whitworth et al., 2019). Riverine concessions in MDD should therefore receive prioritized protection in the face of further colonization due to the importance of rivers for mammal habitat.

Only two of the nine relatively non-persecuted mammals spatially avoided disturbed sites including the common opossum and green acouchi; in addition, the common opossum and Brazilian rabbit benefited from matrix habitat in our study area. The common opossum, although not persecuted for its meat, parts, or as a pet, has been known to be frequently killed by humans and dogs across its geographic range, which could explain its avoidance of areas with high land use (Vaughan & Hawkins, 1999). Common opossums also had a positive relationship with agricultural sites, likely due to the abundance of food (fruit, vegetables, and small domestic animals/ eggs), so their response to human disturbance varied. The green acouchi also avoided areas of high land use, which could be in response to human pressures like hunting; however, it could also be attributable to changes in community structure and interspecific competition. Populations of generalist rodents likely surged in matrix habitat (Suzán et al., 2008), increasing competition for food resources in human-modified areas, and pushing out the acouchi. The Brazilian rabbit increased in occupancy in areas of high land use and avoided intact forest within the private protected area; although there is little research on this particular species, rabbits in general are known to better cope with human disturbance and benefit from human infrastructure and resources (Bock et al., 2009).

By employing occupancy models and analyzing various ecological and anthropogenic covariates, our research sheds light on the impacts of human activities and their consequences for wildlife populations. Additionally, the methods developed in our study provide a means to detect the often subtle and overlooked process of defaunation, even in rapidly changing ecosystems like those in South America, by incorporating potential hunting access points (proximity to agriculture and settlements) into occupancy models. Our findings highlight the urgency of addressing deforestation and defaunation to prevent further ecosystem degradation. In some cases, like white-lipped peccary, we had no detections in areas with moderate levels human impact, indicating that even some amount of habitat destruction and hunting are fostering defaunation and the loss of ecological processes within the Las Piedras region. However, it is important to note that there are still opportunities for action and conservation interventions to mitigate defaunation in many areas as found that protected areas had higher occupancy probabilities for a number of prey species in our study. Protected areas, especially those bordering or utilized by human communities, can act important refugia for wildlife species from habitat loss and hunting (Boron et al., 2022; Lyra-Jorge et al., 2008). By identifying key indicator species and understanding their responses to human pressures, targeted conservation efforts can be implemented to ensure the long-term stability and functionality of these vital rainforest ecosystems across the globe.

Funding

This work received research and financial support from USA 501c3 nonprofit Friends of Hoja Nueva and Peruvian nonprofit Hoja Nueva, and grant support from the University of Washington Student Technology Fund (2014-027), and Panthera’s Small Cat Action Fund.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Samantha Zwicker. The first draft of the manuscript was written by Samantha Zwicker and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Akaike H (1973) Information Theory and an Extension of the Maximum Likelihood Principle. Selected Papers of Hirotugu Akaike. Springer, New York, pp 199–213
Amador S, Naranjo E, Jimenez-Ferrer G (2013) Wildlife predation on livestock and poultry: Implications for predator conservation in the rainforest of south-east Mexico. Oryx, 47. https://doi.org/10.1017/S0030605311001359
Asner GP, Powell GVN, Mascaro J, Knapp DE, Clark JK, Jacobson J, Kennedy-Bowdoin T, Balaji A, Paez-Acosta G, Victoria E, Secada L, Valqui M, Hughes RF (2010) High-resolution forest carbon stocks and emissions in the Amazon. Proceedings of the National Academy of Sciences, 107(38), 16738–16742. https://doi.org/10.1073/pnas.1004875107
Bartoń K (2009) MuMIn: Multi-model inference
Beaudrot L, Ahumada JA, O’Brien T, Alvarez-Loayza P, Boekee K, Campos-Arceiz A, Eichberg D, Espinosa S, Fegraus E, Fletcher C, Gajapersad K, Hallam C, Hurtado J, Jansen PA, Kumar A, Larney E, Lima MGM, Mahony C, Martin EH, …, Andelman SJ (2016) Standardized Assessment of Biodiversity Trends in Tropical Forest Protected Areas: The End Is Not in Sight. PLoS Biol 14(1):e1002357. https://doi.org/10.1371/journal.pbio.1002357
Benítez-López A, Alkemade R, Schipper AM, Ingram DJ, Verweij PA, Eikelboom JAJ, Huijbregts MAJ (2017) The impact of hunting on tropical mammal and bird populations. Science 356(6334):180–183. https://doi.org/10.1126/science.aaj1891
Bock C, Jones Z, Bock J (2009) Abundance of cottontails (Sylvilagus) in an exurbanizing southwestern savanna. Southwest Nat 51:352–357. https://doi.org/10.1894/0038-4909(2006)51[352:AOCSIA]2.0.CO;2
Boron V, Bardales R, Hyde M, Jaimes-Rodriguez L, Stasiukynas D, Barragan J, Passos Viana DF, Payán E (2022) The role of unprotected and privately protected areas for ocelot conservation: Densities in Colombia and Brazil. J Mammal 103(3):639–647. https://doi.org/10.1093/jmammal/gyab149
Boron V, Deere NJ, Xofis P, Link A, Quiñones-Guerrero A, Payan E, Tzanopoulos J (2019) Richness, diversity, and factors influencing occupancy of mammal communities across human-modified landscapes in Colombia. Biol Conserv 232:108–116. https://doi.org/10.1016/j.biocon.2019.01.030
Briant G, Gond V, Laurance SGW (2010) Habitat fragmentation and the desiccation of forest canopies: A case study from eastern Amazonia. Biol Conserv 143(11):2763–2769. https://doi.org/10.1016/j.biocon.2010.07.024
Burs K, Möcklinghoff L, Marques MI, Schuchmann K-L (2023) Spatial and Temporal Adaptations of Lowland Tapirs (Tapirus terrestris) to Environmental and Anthropogenic Impacts. Life 13(1). Article 1. https://doi.org/10.3390/life13010066
Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press
Chávez Michaelsen A, Huamani Briceño L, Menis F, Bejar Chura R, Tito NValera, Perz F, Brown S, Domínguez IF, Del Aguila S, Pinedo Mora R, Alarcón Aguirre G (2013) Regional Deforestation Trends within Local Realities: Land-Cover Change in Southeastern Peru 1996–2011. Land 2(2) Article 2. https://doi.org/10.3390/land2020131
Crawshaw PG (1995) Comparative ecology of ocelot (Felis pardalis) and jaguar (Panthera onca) in a protected subtropical forest in Brazil and Argentina [ProQuest Dissertations Publishing]. https://search.proquest.com/docview/304194774?pq-origsite=primo
D’Cruze N, Galarza FER, Broche O, Bizri E, Megson HR, Elwin S, Machado A, Norrey FC, Coulthard J, E., Megson D (2021) Characterizing trade at the largest wildlife market of Amazonian Peru. Global Ecol Conserv 28:e01631. https://doi.org/10.1016/j.gecco.2021.e01631
Di Minin E, Slotow R, Hunter LTB, Montesino Pouzols F, Toivonen T, Verburg PH, Leader-Williams N, Petracca L, Moilanen A (2016) Global priorities for national carnivore conservation under land use change. Sci Rep 6(1). Article 1. https://doi.org/10.1038/srep23814
Dias DdeM, Massara L, de Campos R, C. B., Henrique Guimarães Rodrigues F (2019) Human activities influence the occupancy probability of mammalian carnivores in the Brazilian Caatinga. Biotropica 51(2):253–265. https://doi.org/10.1111/btp.12628
Dinerstein E, Joshi AR, Vynne C, Lee ATL, Pharand-Deschênes F, França M, Fernando S, Birch T, Burkart K, Asner GP, Olson D (2020) A “Global Safety Net” to reverse biodiversity loss and stabilize Earth’s climate. Sci Adv 6(36):eabb2824. https://doi.org/10.1126/sciadv.abb2824
Emmons L, Feer F (1999) Neotropical Rainforest Mammals: A Field Guide. Bibliovault OAI Repository, the University of Chicago Press, 80. https://doi.org/10.2307/1383232
Espinosa S, Branch LC, Cueva R (2014) Road Development and the Geography of Hunting by an Amazonian Indigenous Group: Consequences for Wildlife Conservation. PLoS ONE 9(12):e114916. https://doi.org/10.1371/journal.pone.0114916
Espinosa S, Celis G, Branch LC (2018) When roads appear jaguars decline: Increased access to an Amazonian wilderness area reduces potential for jaguar conservation. PLoS ONE 13(1):e0189740. https://doi.org/10.1371/journal.pone.0189740
Ferreguetti ÁC, Tomás WM, Bergallo HG (2017) Density, occupancy, and detectability of lowland tapirs, Tapirus terrestris, in Vale Natural Reserve, southeastern Brazil. J Mammal 98(1):114–123. https://doi.org/10.1093/jmammal/gyw118
Figueroa-de-León A, Naranjo EJ, Perales H, Santos-Moreno A, Lorenzo C, Figueroa-de-León A, Naranjo EJ, Perales H, Santos-Moreno A, Lorenzo C (2017) Abundance, density and habitat use of lowland paca (Cuniculus paca, Rodentia: Cuniculidae) in the Lacandon Rainforest, Chiapas, Mexico. Therya 8(3):199–208. https://doi.org/10.12933/therya-17-460
Fiske I, Chandler R (2011) unmarked: An R Package for Fitting Hierarchical Models of Wildlife Occurrence and Abundance. J Stat Softw 43(10):1–23
Foster RJ, Harmsen BJ, Doncaster CP (2010) Habitat Use by Sympatric Jaguars and Pumas Across a Gradient of Human Disturbance in Belize. Biotropica 42(6):724–731. https://doi.org/10.1111/j.1744-7429.2010.00641.x
Fragoso JMV (2004) 18. A Long-Term Study of White-Lipped Peccary (Tayassu pecari) Population Fluctuations in Northern Amazonia: Anthropogenic vs. “Natural” Causes. 18. A Long-Term Study of White-Lipped Peccary (Tayassu pecari) Population Fluctuations in Northern Amazonia: Anthropogenic vs. “Natural” Causes. Columbia University Press, pp 286–296. https://doi.org/10.7312/silv12782-018
Fuller T, Sievert P (2001) Carnivore demography and the consequences of changes in prey availability (pp. 163–178)
Gil-Sánchez JM, Jiménez J, Salvador J, Sánchez-Cerdá M, Espinosa S (2021) Structure and inter-specific relationships of a felid community of the upper Amazonian basin under different scenarios of human impact. Mammalian Biology 101(5):639–652. https://doi.org/10.1007/s42991-021-00149-8
Gonzalez-Borrajo N, López-Bao JV, Palomares F (2017) Spatial ecology of jaguars, pumas, and ocelots: A review of the state of knowledge. Mammal Rev 47(1):62–75. https://doi.org/10.1111/mam.12081
Harmsen BJ, Foster RJ, Silver S, Ostro L, Doncaster CP (2010) Differential Use of Trails by Forest Mammals and the Implications for Camera-Trap Studies: A Case Study from Belize. Biotropica 42(1):126–133. https://doi.org/10.1111/j.1744-7429.2009.00544.x
IUCN (2022) The IUCN Red List of Threatened Species. https://www.iucnredlist.org
Jorge MLSP, Galetti M, Ribeiro MC, Ferraz KMPMB (2013) Mammal defaunation as surrogate of trophic cascades in a biodiversity hotspot. Biol Conserv 163:49–57. https://doi.org/10.1016/j.biocon.2013.04.018
Karanth KU, Nichols JD (1998) Estimation of Tiger Densities in India Using Photographic Captures and Recaptures. Ecology 79(8):2852–2862. https://doi.org/10.1890/0012-9658(1998)079[2852:EOTDII]2.0.CO;2
Kiltie RA, Terborgh J (1983) Observations on the Behavior of Rain Forest Peccaries in Perú: Why do White-lipped Peccaries Form Herds? Z Für Tierpsychologie 62(3):241–255. https://doi.org/10.1111/j.1439-0310.1983.tb02154.x
Lentz DL (ed) (2000) Imperfect Balance: Landscape Transformations in the Precolumbian Americas (First Edition). Columbia University Press
Lyra-Jorge MC, Ciocheti G, Pivello VR (2008) Carnivore mammals in a fragmented landscape in northeast of São Paulo State, Brazil. Biodivers Conserv 17(7):1573–1580. https://doi.org/10.1007/s10531-008-9366-8
MacKenzie DI (2006) Occupancy estimation and modeling: Inferring patterns and dynamics of species occurrence. Elsevier
McBride RT, Thompson JJ (2018) Space use and movement of jaguar (Panthera onca) in western Paraguay. Mammalia 82(6):540–549. https://doi.org/10.1515/mammalia-2017-0040
Medri ÍM, Mourão G (2005) Home range of giant anteaters (Myrmecophaga tridactyla) in the Pantanal wetland, Brazil. J Zool 266(4):365–375. https://doi.org/10.1017/S0952836905007004
Mena JL, Vento R, Martínez JL, Gallegos A (2021) Retrospective and current trend of wild-cat trade in Peru. Conserv Sci Pract 3(12):e558. https://doi.org/10.1111/csp2.558
Mendes CP, Carreira D, Pedrosa F, Beca G, Lautenschlager L, Akkawi P, Bercê W, Ferraz KMP, M. B., Galetti M (2020) Landscape of human fear in Neotropical rainforest mammals. Biol Conserv 241:108257. https://doi.org/10.1016/j.biocon.2019.108257
Michalski F, Peres CA (2005) Anthropogenic determinants of primate and carnivore local extinctions in a fragmented forest landscape of southern Amazonia. Biol Conserv 124(3):383–396. https://doi.org/10.1016/j.biocon.2005.01.045
Monette V, Kelly M, Buchholz R (2020) Human disturbance and the activity patterns and temporal overlap of tapirs and jaguars in reserves of NW Belize. Biotropica, 52. https://doi.org/10.1111/btp.12834
Moran EF (2007) Human Adaptability: An Introduction to Ecological Anthropology (3rd edition). Routledge
Moreno RS, Kays RW, Samudio R Jr (2006) Competitive Release in Diets of Ocelot (Leopardus pardalis) and Puma (Puma concolor) after Jaguar (Panthera onca) Decline. J Mammal 87(4):808–816. https://doi.org/10.1644/05-MAMM-A-360R2.1
Morrison JC, Sechrest W, Dinerstein E, Wilcove DS, Lamoreux JF (2007) Persistence of Large Mammal Faunas as Indicators of Global Human Impacts. J Mammal 88(6):1363–1380. https://doi.org/10.1644/06-MAMM-A-124R2.1
Nagy-Reis MB, Estevo CA, Setz EZF, Ribeiro MC, Chiarello AG, Nichols JD (2017) Relative importance of anthropogenic landscape characteristics for Neotropical frugivores at multiple scales. Anim Conserv 20(6):520–531. https://doi.org/10.1111/acv.12346
Naughton-Treves L, Mena JL, Treves A, Alvarez N, Radeloff VC (2003) Wildlife Survival Beyond Park Boundaries: The Impact of Slash-and-Burn Agriculture and Hunting on Mammals in Tambopata, Peru. Conserv Biol 17(4):1106–1117. https://doi.org/10.1046/j.1523-1739.2003.02045.x
Niedballa J, Sollmann R, Courtiol A, Wilting A (2016) camtrapR: An R package for efficient camera trap data management. Methods Ecol Evol 7(12):1457–1462. https://doi.org/10.1111/2041-210X.12600
Padilla M, Dowler RC (1994) Tapirus terrestris. Mammalian Species 481:1–8. https://doi.org/10.2307/3504109
Palomares F, Fernández N, Roques S, Chávez C, Silveira L, Keller C, Adrados B (2016) Fine-Scale Habitat Segregation between Two Ecologically Similar Top Predators. PLoS ONE 11(5):e0155626. https://doi.org/10.1371/journal.pone.0155626
Peres CA (1996) Population status of white-lipped Tayassu pecari and collared peccaries T. tajacu in hunted and unhunted Amazonian forests. Biol Conserv 77(2):115–123. https://doi.org/10.1016/0006-3207(96)00010-9
R Core Team (2022) R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Rich LN, Davis CL, Farris ZJ, Miller DAW, Tucker JM, Hamel S, Farhadinia MS, Steenweg R, Di Bitetti MS, Thapa K, Kane MD, Sunarto S, Robinson NP, Paviolo A, Cruz P, Martins Q, Gholikhani N, Taktehrani A, Whittington J, …, Kelly MJ (2017) Assessing global patterns in mammalian carnivore occupancy and richness by integrating local camera trap surveys. Glob Ecol Biogeogr 26(8):918–929. https://doi.org/10.1111/geb.12600
Richard-Hansen C, Vié J-C, Vidal N, Kéravec J (1999) Body measurements on 40 species of mammals from French Guiana. J Zool 247:419–428. https://doi.org/10.1111/j.1469-7998.1999.tb01005.x
Ripple WJ, Estes JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, Berger J, Elmhagen B, Letnic M, Nelson MP, Schmitz OJ, Smith DW, Wallach AD, Wirsing AJ (2014) Status and Ecological Effects of the World’s Largest Carnivores. Science 343(6167):1241484. https://doi.org/10.1126/science.1241484
RStudio: Integrated Development for R. RStudio Team, RStudio PBC (2020) http://www.rstudio.com/
Santos F, Carbone C, Wearn OR, Rowcliffe JM, Espinosa S, Lima MGM, Ahumada JA, Gonçalves ALS, Trevelin LC, Alvarez-Loayza P, Spironello WR, Jansen PA, Juen L, Peres CA (2019) Prey availability and temporal partitioning modulate felid coexistence in Neotropical forests. PLoS ONE 14(3):e0213671. https://doi.org/10.1371/journal.pone.0213671
SERFOR (2018) Libro Rojo de la Fauna Silvestre Amenazada del Perú (Primera edición). Serfor (Servicio Nacional Forestal y de Fauna Silvestre). https://www.serfor.gob.pe/portal/wp-content/uploads/2018/10/Libro-Rojo.pdf
Silveira L, Jácomo A, Furtado M, Torres N, Sollmann R, Vynne C (2009) Ecology of the Giant Armadillo (Priodontes maximus) in the Grasslands of Central Brazil. Edentata, 8–10, 25–34. https://doi.org/10.1896/020.010.0112
Suzán G, Armién A, Mills JN, Marcé E, Ceballos G, Ávila M, Salazar-Bravo J, Ruedas L, Armién B, Yates TL (2008) Epidemiological Considerations of Rodent Community Composition in Fragmented Landscapes in Panama. J Mammal 89(3):684–690. https://doi.org/10.1644/07-MAMM-A-015R1.1
Terborgh J, Estes J (2010) Trophic Cascades: Predators, Prey, and The Changing Dynamics of Nature. Bibliovault OAI Repository, the University of Chicago Press
Valsecchi J, Monteiro MCM, Alvarenga GC, Lemos LP, Ramalho EE (2023) Community-based monitoring of wild felid hunting in Central Amazonia. Anim Conserv 26(2):189–198. https://doi.org/10.1111/acv.12811
Vaughan CS, Hawkins LF (1999) Late dry season habitat use of common opossum, Didelphis marsupialis (Marsupialia: Didelphidae) in neotropical lower montane agricultural areas. Revista de Biología Tropical 47(1–2):263–269
Whitworth A, Beirne C, Huarcaya P, Whittaker R, Serrano Rojas L, Tobler SJ, M. W., MacLeod R (2019) Human disturbance impacts on rainforest mammals are most notable in the canopy, especially for larger-bodied species. Divers Distrib 25(7):1166–1178. https://doi.org/10.1111/ddi.12930
Williams-Linera G (1990) Vegetation Structure and Environmental Conditions of Forest Edges in Panama. J Ecol 78(2):356–373. https://doi.org/10.2307/2261117
Zeilhofer P, Cezar A, Tôrres NM, de Almeida Jácomo AT, Silveira L (2014) Jaguar Panthera onca Habitat Modeling in Landscapes Facing High Land-use Transformation Pressure—Findings from Mato Grosso, Brazil. Biotropica 46(1):98–105. https://doi.org/10.1111/btp.12074

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Moderate Levels of Human Disturbance Impact Persecuted Mammals in the Peruvian Amazon Rainforest

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study Area

Focal Species

Camera Trapping

Covariates

Data Analysis

Results

Discussion

Declarations

References

Additional Declarations

Status:

Version 1