Study Area

The study was conducted within the Chakana Reserve and the Antisana Water Conservation Area, located in the provinces of Pichincha and Napo in the northeastern Andes of Ecuador. Both reserves belong to the buffer zone of the Antisana National Park (Figure 1). They are limited to private ranches where cattle raising is the main activity. The nearest town is Pintag located approximately 35 km away (MAE, 2015). However, there are small human settlements along the main access road to the reserves.

Figure 1.

Location of camera-trapping stations for 37 carcasses monitored in the Chakana Reserve and Antisana Water Conservation Area, Ecuador, between 2014 and 2016.

The Chakana Reserve covers 5010 ha in an elevation range of 3000 to 4800 m asl. The reserve was a cattle ranch until 2011, when Jocotoco Foundation purchased it. The cattle population was reduced and nowadays, tourism is its main activity (Jocotoco, 2021). The Antisana Water Conservation Area belongs to the Water Protection Fund (FONAG) and the Quito Metropolitan Public Water and Sanitation Company (EPMAPS), it has an area of 8487 ha in an elevation range of 3720 to 4760 m asl. This area was a ranch with over 22,000 animals for 100 years. Since 2011, farm animals were eradicated, and the area was devoted to ecosystem restoration and conservation activities (EPMAPS & FONAG, 2018).

The area where both reserves are located has a mean annual temperature that varies between 3 and 17 °C (Chuncho & Chuncho, 2019; MAE, 2013) and is characterized by the presence of evergreen shrubland and paramo grasslands, an ecosystem dominated by grasses (Calamagrostis spp.), cushion plants (Azorella sp., Asteraceae) and shrubs (Chuquiraga jussieui, Gynoxys sp., Puya sp.) (Chuncho & Chuncho, 2019; MAE, 2013).

Data Collection

Between November 2014 and September 2016, we systematically monitored 37 carcasses using camera traps. We randomly placed carcasses at seven stations distributed between 3500 and 4150 m asl (Figure 1). Those sites were selected after locating natural carcasses where we observed condors landing and feeding relatively close to roosting or nesting sites. We placed carcasses in the same locations due to the accessibility and security of camera traps. The carcasses were placed monthly at two camera stations. This rotation meant that some camera stations had no carrion between months. Since the camera stations were separated by less than 5 km and located at about the same elevation, the environmental conditions were assumed to be the same as those described for the study area.

Thirty-five carcasses were experimental, and two further carcasses were found killed by predators. The researchers provided experimental carcasses after local cowboys (Chagras) killed them. The carcasses comprised cattle (Bos taurus) (89.1% of the total number of carcasses), horses (Equus caballus), and alpacas (Vicugna pacos) (each 5.4% of the total number). The average weight of the monitored carcasses was 385 kg, 464 kg, and 60 kg respectively (García Neder et al., 2009; Tirira, 2017).

We monitored carcasses by photo-trapping with a camera trap. We used Bushnell 8MP Trophy Cam HD Hybrid and Reconyx HC600 Hyperfire camera traps. These cameras operated 24 hours per day and were programmed to capture a burst of three photographs with a one-minute interval between detections. Each camera was mounted on a wooden pole at a distance between 5 and 10 m from the carcass. Photographic data were retrieved every 15 days and entered into a database. We registered species that fed on carrion, pecked/bit, chewed, or torn carcasses. We considered independent events as consecutive photographic records of the same species separated by at least 60 minutes. For each independent event, we recorded the time, the time spent at the carcass, the species photographed, the abundance, and the number of days since carcass placement.

Data Analysis

We calculated the sampling effort (trap-nights) using R version 4.0.0 (R Core Team, 2021) through the R package “camtrapR” (Niedballa et al., 2016). We obtained the value from the total number of camera traps distributed in the field, multiplied by the total number of days in which they were active for 24 hours. We calculated a species accumulation curve and three estimators (Chao, Jackknife 1, and Bootstrap) to determine the representativeness of the scavenger community. The species accumulation curve and the estimators were calculated by the R package “biodiversityR” (Kindt & Coe, 2005).

Relative frequency of detection was calculated as the percentage ratio of independent events of a given species relative to the total number of all scavenging species detections. Minimum abundance corresponded to the highest number of individuals of the same species recorded in an independent event. It was estimated as from individuals recorded together in a single photo or individuals recognizable by sexual and age. The 37 carcasses were monitored until complete consumption (bones, fur, and skin remaining) (Orihuela-Torres et al., 2021). The persistence time was the number of days which scavengers fed on the carcasses; from the date of placement until the last scavenger species was recorded. The average persistence time was calculated from the persistence time of all the carcasses and each carcass species. We estimated the residence time of each independent event as the difference between the time of the first and the last recorded photograph. The residence time of a scavenging species was estimated as the sum of the time between their independent events.

Activity patterns were described for species with at least 30 independent events between all stations (Fonseca-Prada et al., 2023; Zapata-Ríos & Branch, 2016). Species were classified into four categories, (a) diurnal, species detected in the period 06:00 to 18:00 h; (b) nocturnal, species detected between 20:00 and 04:00 h; (c) crepuscular, species recorded from 04:00 to 06:00 h or from 18:00 to 20:00 h; and (d) cathemeral, species with 24-hour activity (Monroy-Vilchis et al., 2011; Mosquera-Guerra et al., 2018). Due to equatorial latitude, sunrise and sunset are regular and transformation was unnecessary to account for differences across seasons. The graphical representation was produced using the R package “Circular” (Lund et al., 2017).

We divided carcasses into two categories to describe changes in the activity patterns of native scavenging species. We divided the data into carcasses with high activity of free-ranging dogs (dogs` records represented more than 25% of the total independent events) and carcasses with low/null activity of free-ranging dogs (dogs´ records represented less than 25% of the total independent events). We chose 25% as an approximation of dogs´ relative frequency of detection. We represented the activity patterns using the R package “Overlap” (Meredith & Ridout, 2021).

We fitted Generalized Linear Mixed Models (GLMM) by the R package “lme4” (Bates et al., 2015) to test the influence of free-ranging dogs´ presence in native scavenging species. We included the species-independent events as the response variable and the free-ranging dogs´ presence as the predictor variable. We also added the number of days since carcass placement as covariable and a random intercept of the carcass ID. We used the negative binomial distribution to adjust data overdispersion (Lindén & Mäntyniemi, 2011).

Independent events

We obtained a total of 1010 independent events and a total sampling effort of 1065 trap-nights. We recorded seven vertebrate species feeding on carcasses: three mammal and four bird species. The species accumulation curve reached the asymptote (Figure 2). We got a 100% of representativeness according to all the species richness estimators calculated (Chao1=7±0; Jackknife1=7±0; Bootstrap= 7.17±0.4 (stimation±standard deviation)).

Figure 2.

Species accumulation curve of vertebrate scavenger species studied between 2014 and 2016 in the buffer zone of Antisana National Park. The grey area represents the confidence interval at 95%.

The species with the highest number of independent events were the Andean fox (Lycalopex culpaeus, n= 279), the Carunculated Caracara (Phalcoboenus carunculatus, n=257), the free-ranging dogs (n=233), and the Andean Condor (n=191). These species accounted for 95% of the total number of independent records. For the Black-chested Buzzard-Eagle (Geranoaetus melanoleucus, n=33), the Black Vulture (Coragyps atratus, n=10), and the Andean bear (Tremarctos ornatus, n=7), we obtained a combined value of 50 records, which represented 4.95 % of the total number of records (Table 1).

Table 1.

Vertebrate scavenger species recorded at 37 carcasses in the buffer zone of Antisana National Park, between 2014 and 2016.

Minimum Abundance

The Andean Condor and the Carunculated Caracara were recorded in groups over 10 individuals. They were followed by free-ranging dogs, with groups up to five individuals, Andean foxes (four individuals), and Black Vultures (three individuals). The average minimum abundance of Carunculated Caracara was 2.28. Andean Condor, free-ranging dogs, and Black Vulture' average minimum abundance was between 1.82 and 1.90 individuals (Table 1). Andean foxes average minimum abundance was 1.10. All the records of Andean bear and Black-chested Buzzard-Eagle were of only one individual (Table 1).

Time of Food Consumption

The average carcass persistence time was 29±14 (mean±standard deviation) days. Cattle carcasses were fully consumed in an average of 31±14 days. For horse and alpaca carcasses, the average persistence time was 14±14 and 18±5 days, respectively.

We recorded 2407 hours of activity in front of the camera traps. Free-ranging dogs were present at carcasses for 40.39% (972 hours) of the total activity time, followed by the Andean fox (784 hours), the Carunculated Caracara (436 hours), and the Andean Condor (174 hours). These three native species along occupied 57.93% of the total time spent scavenging the carrion. The Black-chested Buzzard-Eagle, Black Vulture, and Andean bear together remained at the carrion for 1.68% of the time (Figure 3).

Figure 3.

Percentage of total persistence time of seven vertebrate scavenger species at 37 carrion carcasses monitored between 2014 and 2016 in the buffer zone of Antisana National Park.

Activity Patterns

The scavenging birds were diurnal, the Andean Condor and the Carunculated Caracara fed on carrion between 06:00 and 18:00 h. More scavenging bird records were in the early morning hours with another peak in activity between 14:00 and 15:00 h. Mammals had a cathemeral pattern with activity recorded across the 24-hour cycle. The Andean fox concentrated its activity at night, between 19:00 and 06:00 h. In contrast, free-ranging dogs mainly accessed the carrion in the morning hours, with a peak observed between 06:00 and 07:00 h (Figure 4).

Figure 4.

Activity patterns of the most abundant vertebrate scavenger species recorded at 37 carcasses monitored between 2014 and 2016 in the buffer zone of Antisana National Park. A. Free-ranging dogs Canis lupus familiaris, B. Andean Fox Lycalopex culpaeus, C. Carunculated Caracara Phalcoboenus carunculatus, D. Andean Condor Vultur gryphus.

Twenty-two carcasses were considered as low-activity and fifteen as high-activity. We found an increase in the records of the Andean fox in high-activity carcasses, between 18:00 and 24:00 h (Figure 5). The peak activity of the Andean Condor changed from 06:00 to 11:00 h in low-activity carcasses to 10:00 to 11:00 h in high-activity carcasses (Figure 5).

Figure 5.

Activity patterns of free-ranging dogs and four species of native vertebrate scavengers recorded at 37 carcasses studied between 2014 and 2016 in the buffer zone of Antisana National Park. The Y-axis is the Kernel density of species-independent events. A. Andean Condor Vultur gryphus, B. Andean Fox Lycalopex culpaeus, C. Carunculated Caracara Phalcoboenus carunculatus. Black-continued line represents the activity pattern of native scavenger species. The blue-dashed line represents the activity pattern of free-ranging dogs in carcasses with low activity. The red-dashed line shows the activity pattern of free-ranging dogs in carcasses with high activity.

Impacts of Free-Ranging Dogs

Free-ranging dogs visited 28 carcasses and were absent from nine of them. The GLMMs showed that free-ranging dogs´ presence negatively influenced the number of independent events of the Andean Condor, the Andean fox, the Carunculated Caracara, and the Black-Chested Buzzard Eagle (p-value<0.05; Table 2). The number of days since carcass placement was significant for all the scavenging species except the Carunculated Caracara (p-value=0.69; Table 2).

Table 2.

Results of the GLMMs testing the influence of free-ranging dogs´ presence in the number of independent events of the Andean Condor, the Andean fox, the Carunculated Caracara, and the Black-chested Buzzard Eagle.