Introduction

Vertebrate scavengers play an essential role in ecosystems by consuming dead animals, which accelerates the process of organic matter decomposition and nutrient recycling1,2,3. Two functional groups use carrion as a food resource. Obligate scavengers (i.e., vultures) are species that feed almost exclusively on carrion, while facultative scavengers (e.g., mammalian carnivores, avian species like raptors and corvids) often, but not always, feed on carrion3,4,5. Vultures help to eliminate potential foci of disease transmission due to their gregarious habits and ability to consume large amounts of biomass in relatively short periods3,6. Despite this, anthropogenic threats such as poisoning with pesticides or rodenticides, lead contamination, human-wildlife conflict, and illegal hunting make them one of the most threatened groups of birds worldwide4,7,8,9,10.

Carcass size is one of the main factors influencing their detection and consumption patterns11,12. Species richness and abundance in vertebrate scavenger communities often vary depending on carcass size and biomass availability11,12,13. Likewise, carcass consumption patterns (e.g. detection time, carcass consumption time and rate) varies due to climate-related factors, such as temperature or precipitation14,15. Also, it has been demonstrated that the type of landscape where the carcasses appear is a factor influencing the time taken to detect and consume them16. However, there is a notable geographical bias in the amount of information about the influence of climatic factors and landscape on scavenger communities, as available studies have mainly been conducted in septentrional countries12,17,18. Although there are no distinct seasons per se in the Neotropics, some regions exhibit cyclical rainfall patterns that mark two seasons19; therefore, it is foreseeable that these dynamics may influence the scavenger communities. In globally threatened ecosystems such as tropical dry forests, which are highly dependent on their rainfall patterns, it is essential to assess the influence of carcass size, rainy and dry seasons on the composition and dynamics of vertebrate scavenger communities11,20,21.

Interactions among scavenging species can facilitate food detection and reduce carcass consumption time22,23. For example, some vulture species play an essential role in maximizing the foraging success of other vulture species (i.e., heterospecifics) by providing visual cues to food location and opening up carcasses, enabling access to food22,23,24. To prevent agonistic encounters, some scavengers avoid using the resource in the presence of species that may exhibit aggressive behaviors, such as domestic dogs (Canis lupus familiaris)25,26. The lack of information on carrion use, foraging behavior, and the richness and abundance of vertebrate scavengers in tropical dry forests makes it difficult to design and implement effective conservation strategies (e.g. multi-species action plans) for species with declining populations, such as some New World vultures8,21,27,28,29,30. Given the vulnerability of the tropical dry forest, which has historically lost 75% of its original extent21,31,32scientific research on vertebrate scavenger communities in this critically threatened ecosystem is a priority.

A comprehensive understanding of climate-related and landscape factors that influence vertebrate scavenger communities and their dynamics, is essential to developing conservation measures and maintaining the ecosystem services2,6,16. By consuming carcasses and maintaining healthy ecosystems, vulture species are essential for biodiversity conservation and achieving sustainable development goals33,34. We aimed to describe the scavenging guild exploiting carrion and to evaluate the influence of carcass size and season on the vertebrate scavenger community (species richness and abundances) and carcass consumption patterns (detection time, total consumption time, consumption rate) in the tropical dry forest of western Ecuador. We predicted that (1) larger carcasses (medium-sized carcasses in the experimental system), as a resource with greater available biomass, would have a higher number of scavenger species and a higher abundance, that (2) scavenger abundance would differ between the rainy and dry seasons, that (3) carcass consumption time and detection time would be lower on medium-sized carcasses and during the dry season due to lower canopy cover, and that (4) carcass consumption rate would be higher on medium-sized carcasses11,35.

Methods

Study area

The study was conducted in three private nature reserves located in the dry forest of western Ecuador. The Lalo Loor Reserve and the Jama-Coaque Reserve are located in Manabí Province, while the Bosque Protector Cerro Blanco is situated in Guayas Province (Fig. 1). The Lalo Loor Reserve has 180 ha and is managed by the Ceiba Foundation36. The Jama-Coaque Reserve has 840 ha belonging to the Third Millennium Alliance37and the Bosque Protector Cerro Blanco has an area of 6,078 ha and is managed by the Pro-Bosque Foundation38.

The Manabí and Guayas reserves are distanced from each other by approximately 200 km but share the same climatic and environmental characteristics. The altitude of the reserves ranges from 80 to 500 m above sea level. The study area has a rainy season, between January and May, characterized by rainfall ranging from 500 to 1000 mm, and a dry season from June to December, with precipitation below 500 mm. The average annual temperature is 26° C, with a 1 to 2° C decrease during the dry season39,40.

The landscape at the three sites consists of remnants of dry semi-deciduous forest, urbanized areas, and agricultural land. The vegetation is xerophytic and deciduous, dominated by shrub species. The tree species reach a canopy height of 10 to 25 m. During the dry season, between 25 and 75% of the tree and shrub species lose their leaves, and the herbaceous vegetation disappears. The representative plant species of the area are Ceiba trischistandra, Cavanillesia platanifolia, and Cecropia angustifolia39,40.

The three reserves are part of the Tumbes-Chocó-Magdalena hotspot. A total of 543 vertebrate species have been recorded in the area surrounding the reserves, including three species of obligate scavengers: the Black Vulture (Coragyps atratus), the Turkey Vulture (Catharthes aura), and the King Vulture (Sarcoramphus papa). There is no information or estimate of the vulture population size in the study area. However, there are informal records of breeding pairs and chicks of both the Turkey and the King Vultures. The most common large herbivores are the Collared Peccary (Dicotyles tajacu) and deer of the Mazama genus. Native carnivores include the Sechura Fox (Lycalopex sechurae), the Tayra (Eira barbara), and the Ocelot (Leopardus pardalis)39,41. In the study area, natural carcasses consist mainly of medium-sized wild herbivores. However, it is possible to find larger carcasses with less frequency of domesticated introduced species such as cattle (Bos taurus), horses (Equus caballus), or donkeys (Equus asinus) that occasionally die inside or near the reserves.

Fig. 1
figure 1

Location of the three study sites in the dry forest of western Ecuador and the camera-trapping stations within each study area. The map was created using QGIS 3.28 software (URL: https://www.qgis.org/).

Full size image

Data collection

Between November 2022 and May 2023, we systematically monitored 64 experimental carcasses at eight camera-trapping stations, which were placed in the Cerro Blanco Protected Forest (4 stations), the Jama-Coaque Reserve (3 stations), and the Lalo Loor Reserve (1 station) (Fig. 1). The camera-trapping stations were separated from each other by at least 1 km. At each camera-trapping station, we simultaneously placed two carcasses of different sizes at least 100 m apart: a small-sized carcass weighing between 1.5 and 4 kg of domestic chickens (Gallus gallus), and a medium-sized carcass weighing between 15 and 40 kg of domestic pigs (Sus scrofa var. dom) or domestic goats (Capra hircus).

We placed a total of four sets of carcasses. At each camera-trapping station, we placed carcasses on two occasions in the dry season (November and December) and on another two occasions in the rainy season (March and April). We did not place carcasses in January and February to avoid the transition period between the rainy and dry seasons. Carcasses placed at the same camera-trap station were approximately 30 days apart from the previous one. These were monitored from the moment they were placed until their biomass was consumed entirely, leaving only skin, bones, fur, or feathers42.

We used one camera trap for each carcass placed in the field. The camera traps were placed 50 cm above the ground and 6–7 m from the carcasses. We set the camera traps to hybrid mode (capturing two photos and a 20-second video) with a 30-second delay and a “Normal” sensor level. The time elapsed between each of the two photos and the video was of one second by camera default. The use of videos made it possible to clearly verify the number of individuals and their stages in case the photos were not clear enough. We checked the memory cards and batteries every 15 days after placing the carcasses.

At each camera-trapping station, we recorded the time and date of carcass placement, the weight of the carcass (in kg), and the time and date of retrieval from the camera trap. We estimated canopy cover at the time of carcass placement by taking a photo pointing skyward and determining the percentage of the photo that was covered by the tree canopy. Then, we obtained the following data on the scavenger community from the analyzed photos and videos: total scavenger species richness, vulture species richness, facultative scavenger species richness, abundance of each species, vulture abundance, and mammalian scavenger abundance. Abundance was calculated as the highest number of individuals of the same species recorded in the same photo or photo-sequence. It was estimated by recognizing individuals by their plumage or fur patterns, age, and sex11especially for S. papa, C. aura, L. pardalis, and E. barbara. The abundance of vultures and carnivorous mammals was estimated as the sum of the number of individuals of each species for each group in the same carcass.

We estimated three variables related to carcass consumption. (1) Detection time was the time in hours it took the scavenger species to locate the carcass. (2) Total consumption time, which was the time in hours when the carcass was consumed entirely since it was placed in the field. Meanwhile, (3) the consumption rate was estimated based on the relationship between carcass weight (kg) and total consumption time (h)11,12,16.

Data analysis

We analyzed 62 carcasses because one camera trap was stolen, and another had a damaged SD card. We constructed Generalized Linear Mixed Models (GLMMs) to assess differences in scavenger species richness and abundance, using the total scavenger species richness, vulture species richness, total carnivore mammal species richness, species abundance, vulture abundance, and mammal abundance as response variables. Carcass size and season were used as predictor variables. Camera-trapping stations were treated as a random factor to control for the lack of independence of carcasses placed simultaneously at the same camera-trapping station. Negative binomial family models were fitted to address overdispersion43. The R package “lme4” fit the mixed models44.

The variables of carcass detection time, total consumption time, and consumption rate were normalized by the natural logarithm12 and selected as response variables to fit Linear Mixed Models (LMMs). The predictor variables were carcass size, vulture abundance, and percentage of canopy cover. Variance inflation factor (VIF) analysis was used to measure the multicollinearity of model variables, and variables with values greater than five were excluded45. No interactions between any of the model terms were included.

We performed the model selection based on the Akaike information criterion (AICc). Only models with a ΔAICc < 2 were considered valid46. If we obtained more than one candidate model, we performed model averaging using the package “MuMIn”47. We used the R package “MuMIn”47 to calculated the R2 marginal (R2m) to evaluate the proportion of variance explained by the fixed effects alone, and R2 conditional (R2c) to see the proportion of variance explained by the fixed and the random effects. All tests were performed using R software version 4.4.148. We considered the results to be statistically significant when p < 0.05.

Results

Scavenger community

We detected seven species of vertebrate scavengers, including three obligate scavengers: the Black Vulture (Coragyps atratus), the Turkey Vulture (Cathartes aura) and the King Vulture (Sarcoramphus papa), and four species of mammalian facultative scavengers: the Ocelot (Leopardus pardalis), the Tayra (Eira barbara), the Crab-eating Raccoon (Procyon cancrivorus) and the domestic dog (C. l. familiaris) (Fig. 2). No facultative avian scavenger species were detected.

The species that most frequently visited the carcasses were C. aura (87.1%) and C. atratus (85.5%), followed by S. papa (48.4%), L. pardalis (25.8%), and E. barbara (11.3%). While the species that visited the carcasses least frequently were C. l. familiaris (3.2%) and P. cancrivorus (1.6%). We detected that C. aura and C. atratus visited 100% of medium-sized carcasses and between 69% and 80% of small carcasses in both the dry and rainy seasons, and S. papa visited mainly medium-sized carcasses in the dry season (86.7%) and the rainy season (62.5%). In contrast, C. l. familiaris and P. cancrivorus were only recorded in small carcasses during the dry season, and E. barbara was not recorded in medium-sized carcasses during the rainy season (Fig. 3).

Vultures were the most abundant species, with groups of 15.66 ± 11.15 (mean ± SD) individuals. Coragyps atratus was the most abundant species, recorded with a mean abundance of 10.48 ± 9.75 and a maximum of 49 individuals. It was followed by C. aura and S. papa with mean abundances of 3.35 ± 2.14 and 1.82 ± 2.85 individuals, respectively. Cathartes aura was recorded in groups of up to nine individuals, while S. papa was recorded in groups with a maximum of 11 individuals. Facultative mammalian scavengers were less abundant, with a mean of less than one individual per carcass. L. pardalis had the highest mean in this group, with 0.34 ± 0.62 individuals (Supplementary Table 1).

Fig. 2
figure 2

Native vertebrate scavenger species recorded in the dry forest of western Ecuador. (A) Black Vulture (Coragyps atratus), (B) Turkey Vulture (Cathartes aura), (C) King Vulture (Sarcoramphus papa), (D) Ocelot (Leopardus pardalis), (E) Tayra (Eira barbara), (F) Crab-eating Raccoon (Procyon cancrivorus).

Full size image
Fig. 3
figure 3

Percentage of small-sized and medium-sized carcasses visited by seven species of vertebrate scavengers during the rainy and dry season in the dry forest of western Ecuador.

Full size image

Carcass consumption patterns

The most frequent detector species was C. aura, which detected 54.84% and 58.06% of small and medium-sized carcasses, respectively. Leopardus pardalis detected 22.58% of small-sized carcasses, while C. atratus detected 22.58% of medium-sized carcasses. Mean (± SD) carcass detection time was 20.16 ± 18.99 h. Small-sized carcasses were detected in a mean time of 25.07 ± 23.82 h, and medium-sized carcasses were detected in a mean time of 15.25 ± 10.78 h (Supplementary Table 1).

Mean (± SD) carcass total consumption time was 47.64 ± 31.73 h. In medium-sized carcasses, the mean time of total consumption was 40.87 ± 21.37 h; in small-sized carcasses, the mean time of total consumption increased to 54.40 ± 38.67 h. The consumption period of small-sized carcasses varied between 3.73 and 149.13 h (Supplementary Table 1).

Mean (± SD) carcass consumption rate was 0.47 ± 0.75 kg/h. The mean consumption rate was higher in medium-sized carcasses (0.86 ± 0.91 kg/h) than in small-sized carcasses (0.08 ± 0.11 kg/h).

Influence of carcass size and season on scavenging dynamics

Species richness and abundance of facultative scavengers did not show significant differences (p-value > 0.05) influenced by the carcass size or season (Supplementary Table 2). The model-averaged indicated that carcass size influenced all the variables analyzed, except the scavenger species richness. While the season influenced the vulture abundance and the abundance of S. papa. The camera-trapping station (random effect) highly increased the variance explained for the variables: vulture abundance, the abundance of C. atratus and S. papa (Table 1). Medium-sized carcasses had higher scavenger species richness, higher vulture species richness and abundance than small-sized carcasses. During the dry season, both small and medium-sized carcasses recorded higher scavenger species richness and higher vulture abundance compared to the rainy season (Fig. 4).

Table 1 Generalized linear mixed models (GLMMs) of the variables of the scavenger community in the dry forest of Western ecuador. Model-averaged coefficients, standard errors (SE), p-values, R2 marginal (R2m) evaluate the proportion of variance explained by the fixed effects alone, and R2 conditional (R2c) evaluates the proportion of variance explained by the fixed and the random effects. Significant p-values (< 0.05) are highlighted in bold.
Full size table
Fig. 4
figure 4

Boxplot of the differences in the variables of the scavenger community studied in the tropical dry forest of Ecuador according to carcass size and season (rainy vs. dry).

Full size image

Carcass total consumption time was not explained by any of the variables included in the LMMs. In contrast, Carcass detection time was influenced by carcass size, with a low R2 marginal value, explaining low variability in the model. Also, carcass size had a significant influence on the consumption rate. The camera-trapping station (random effect) increased the variance explained for the carcass detection time and consumption rate (Table 2). Larger carcasses showed a higher consumption rate compared to smaller ones.

Table 2 Linear mixed models (LMMs) of the carcass consumption variables. Model-averaged coefficients, standard errors (SE), p-values, R2 marginal (R2m), which evaluates the proportion of variance explained by the fixed effects alone, and R2 conditional (R2c), which evaluates the proportion of variance explained by the fixed and the random effects, are shown. Significant p-values (< 0.05) are highlighted in bold.
Full size table

Discussion

We found that the vertebrate scavenger assemblage in the dry forest of western Ecuador was dominated by the obligate scavengers (i.e., vultures), with three vulture species recorded scavenging on experimental carcasses: C. atratus, C. aura, and S. papa. Facultative scavengers were recorded to a lesser extent, only involving mammalian carnivore species: L. pardalis, E. barbara, P. cancrivorus, and C. l. familiaris (Fig. 3). Our results indicate that the carcass size influenced all the variables analyzed, except the scavenger species richness. The season influenced the abundance of vultures recorded on carcasses, especially that of S. papa (Fig. 4). Furthermore, the total consumption time of carcasses was not explained by any of the variables in the model. Still, the carcass size influenced the detection time and consumption rate.

Our findings suggest that in the Neotropical dry forest, vultures may be the dominant species on carcasses (Fig. 3). Similarly, in the Brazilian Cerrado dry forest, the scavenger community was dominated by vultures, such as S. papa, C. aura, and C. atratus11. This probably occurs because the mammalian facultative scavengers in the Neotropics tend to be medium-sized species (e.g., L. pardalis, E. barbara). In contrast, in African savannas and European temperate forests, the facultative scavengers tend to be larger mammalian carnivores such as lions (Panthera leo), spotted hyenas (Crocuta crocuta), and bears (Ursus arctos), which may be more dominant and monopolize carcasses35,49,50. The greater dominance of vultures on carrion may also be due to their efficiency in locating carcasses (via visual and olfactory abilities) compared to mammals (only via olfactory abilities)22,23,51.

Contrary to our predictions, the total scavenger species richness in the tropical dry forest was not significantly influenced by carcass size or season. This is opposite to studies conducted in other tropical ecosystems where medium and large carcasses recorded more scavenging species than the smaller ones11,35. This pattern is repeated in various ecosystems worldwide, where larger carcasses recorded scavenger communities composed of more species2,14,16. However, vulture abundance was influenced by carcass size. The larger carcasses have more biomass for consumption, allowing more animals to feed on them2,35.

The absence of facultative avian scavenger species is striking despite the presence of raptor species that could potentially feed on them, such as the Red-throated Caracara (Ibycter americanus) and the Crested Caracara (Caracara plancus)52. This could be because vultures dominate carcasses and exploit them faster than other birds35,53. Likely, the high abundance of vultures in the carcasses, which are larger than other raptors, scares away facultative avian scavengers that prefer to use other resources available in the forest.

The influence of the camera-trapping station on the abundance of S. papa, may be related to the type of landscape in which the carcass appears19,53. S. papa seems to have a more specialized foraging behavior, restricted to the carcasses available within or near natural landscapes5. This reliance on the dry forest could make them more dependent on medium-sized carcasses and more vulnerable to reductions in natural food sources. On the other hand, C. atratus and C. aura have a flexible and opportunistic diet, frequently feeding on fish scraps along beaches, roadkill animals, and garbage dumps in cities54,55,56. These urban and suburban areas provide a stable food source for their diet5, and their visits to the forest in our study area may be more related to searching for additional resources. This may explain the higher percentage of visits to small-sized carcasses compared to S. papa (Fig. 3).

The rainy and dry seasons affected the vulture abundance, as well as that for C. atratus and S. papa (Fig. 4). This could be due to variations in canopy cover since the dry forest is an ecosystem characterized by semi-deciduous vegetation21,40. The abundance of the two vulture species was higher in the dry season. The lower canopy covers likely facilitated the detection of carcasses by visual cues in species that rely primarily on their sense of sight, such as C. atratus and S. papa51.

The presence of the detector species C. aura and L. pardalis, which rely mainly on their sense of smell to locate food5,22would explain why the detection time was not influenced by the season or carcass size. Likewise, changes in canopy cover would not be significant as they do not directly affect the ability to detect carcasses. Species with a developed sense of smell play an important role in ecological communities, as their presence alerts other scavenging species to the existence of available food, favoring the use of this resource22,51.

In the three protected areas studied, natural carcasses consist mainly of the remains of ungulate herbivores’ corpses, such as D. tajacu and Mazama sp39,41. As well as introduced domestic species such as C. hircus, S. scrofa var. dom., and cattle (Bos taurus) that occasionally die inside or near forested areas. The availability of carcasses in natural conditions, especially the larger ones, would depend on the presence of predators and their successful hunting events35,57. It is of great concern that no large felids (i.e., puma Puma concolor, jaguar Panthera onca) have been recorded in the last decade in the three protected areas58,which would probably reduce the possibility of finding carcasses from predation events. Specific seasonal factors could also influence the abundance and availability of food resources. However, data related to climatic factors, such as temperature and rainfall, were not directly measured, which could provide a better explanation of their influence on the assemblage and consumption patterns of vertebrate scavengers.

Implications for conservation

The high abundance of vultures could explain the low species richness of mammalian carnivores found and the absence of avian facultative scavengers in the dry forest of western Ecuador, as obligate scavengers are more efficient at detecting and consuming carrion16,22,23. We detected mixed groups of vultures up to 51 individuals, which could have restricted the use of carcasses by mammals57,59. This suggests that in the critically endangered tropical dry forest20obligate scavengers are probably fulfilling an essential role in removing organic matter and potential sources of infection for humans, which could hardly be covered by facultative scavengers alone. The loss of vultures could have important implications for ecosystem services and human well-being2,6.

Our findings provide evidence of the significance of carcass size for vulture abundance and consumption rates, as well as the importance of rainy and dry seasons for the total abundance of vultures and each vulture species. Meanwhile, the lack of significant explanatory variables for the total consumption time indicates the need for further studies on scavenger communities and their consumption patterns in tropical dry forests. It is necessary to prioritize research efforts on understudied species with globally declining populations, such as S. papa, which is listed as Near Threatened in Ecuador8,11,60,61.

We also highlight the record of domestic dogs on carcasses in the Ecuadorian western dry forest. These alien species compete for food with native scavengers26,62 and are considered a major threat to vultures in Ecuador, either through direct attacks on individuals62,63  by causing fake news and conflicts with rural inhabitants, which leads in turn to unintentional subsequent poisoning of obligate scavengers7,33,64,65. Expanding the study area to obtain a more representative sample size and achieve a comprehensive understanding of interspecific interactions in scavenger communities, particularly those involving vulture species, is essential22,23,24. It is also necessary to estimate the populations of vulture species in order to evaluate their changes over time2,8,66as well as to understand human-vulture interactions67,68. This will facilitate the design and implementation of multi-species conservation plans suited to the particular social-ecological contexts of critically endangered neotropical ecosystems.