The crab-eating fox, Cerdocyon thous (Linnaeus, 1766), is the wild canid with the broadest distribution in South America (Maffei and Taber 2003). It ranges from northern Venezuela to Argentina (Lucherini 2015; Castelló 2018). Notably, recent records suggest that this species is expanding its range from the south-eastern Atlantic region of South America to the central Pacific coast of Central America by crossing the Darién forests (Tejera et al. 1999; Hody et al. 2019). Furthermore, its distribution extends from north-eastern to southern Argentina (Fariñas-Torres et al. 2024; Manzione et al. 2025). Cerdocyon thous is a highly adaptable canid, occurring in diverse ecosystems, including tropical and subtropical forests, montane regions, dry areas, plains, and savannas, from sea level to altitudes of 3690 m (Escobar-Lasso et al. 2016; Castelló 2018).

Cerdocyon thous exhibits an omnivorous and opportunistic feeding strategy, consuming a wide variety of food items including fruits, plants, molluscs, insects, fish, amphibians, reptiles, birds, and both small and medium-sized mammals (Bueno and Motta-Junior 2004; Gatti et al. 2006; Pedó et al. 2006; Canesini et al. 2008; Dutra-Vieira et al. 2024). This dietary flexibility allows it to exploit diverse habitats and seasonal food availability (Gatti et al. 2006; Pedó et al. 2006). However, this species is also known to prey on small domestic animals, such as poultry, which frequently leads to conflict with local communities. These interactions often result in negative perceptions and, in some cases, retaliatory killing or persecution as a means of conflict resolution (Kihn et al. 2022).

This species is known for its ecological plasticity and tolerance of anthropogenic disturbance (da Silva et al. 2020). It occupies a range of modified habitats, including grasslands, agricultural areas, and peri-urban and urban settings (Dias et al. 2014; Jiménez et al. 2017; da Silva et al. 2020). However, such tolerance also exposes it to significant threats, particularly road mortality, making C. thous one of the most frequently reported mammal species in wildlife–vehicle collisions across its geographical range (Vázquez Rodríguez et al. 2016; Bauni et al. 2017; Rojano Bolaño and Ávila Avilán 2021). In peri-urban areas, C. thous often interacts with domestic animals, in particular with free-ranging dogs, Canis lupus familiaris (Linnaeus, 1758), which compete for similar resources such as prey (Lacerda et al. 2009; Sánchez-Londoño 2014; Lessa et al. 2016) and territory (Jiménez et al. 2017). These interactions can alter the activity patterns of C. thous (Jiménez et al. 2016), increase disease exposure (Brandão et al. 2020; Acevedo and de Lima 2021; Dutra-Vieira et al. 2024), and facilitate the bidirectional transmission of pathogens (Nava et al. 2017; Ramos et al. 2020; Jiménez-Ruiz et al. 2024), particularly in peri-urban areas where anthropogenic ecosystem changes have intensified the spatial overlap between wildlife and domestic animals.

Cerdocyon thous exhibits considerable ecological variation across its range, potentially representing distinct evolutionary and ecotypic forms adapted to local environments (Martinez et al. 2013; Santos et al. 2024). Populations are known to be omnivorous and nocturnal in open areas, such as the Cerrado and Pampas biomes (de Matos Dias and Bocchiglieri 2016). Forest-dwelling individuals, especially those in the Amazon, may exhibit different foraging strategies, trophic roles, and habitat-use patterns (Castelló 2018; Dutra-Vieira et al. 2024). Populations that occupy mosaics composed of native and exotic vegetation tend to have a generalist diet, consuming a mix of animal prey, and both native and exotic fruit and other plant materials (Rocha et al. 2008), and animals may modify their activity patterns based on factors such as human density and available habitat (Santos et al. 2024). Such differences suggest that C. thous may exhibit localised adaptations despite taxonomic continuity (Dutra-Vieira et al. 2024), with important implications for understanding its ecological niche, ecological plasticity, and for informing conservation strategies across its broad geographic distribution.

Despite its broad range, C. thous remains a rare and poorly understood species in Ecuador. Historical records suggest its occurrence in the Andes and the coastal regions of the country (Tirira 2009; Lucherini 2015; Ramírez-Chaves and Pérez 2015). However, evidence of its presence in Ecuador is limited to a single specimen (Tirira 2009), leading to uncertainty and controversy over its historical and current occurrence in Ecuador (Ramírez-Chaves and Pérez 2015). Since the 1990s, however, there have been increasing reports of this species in the northern Amazon of Ecuador (Tirira 2017). Two main hypotheses have been proposed to explain these observations: that they represent sporadic occurrences, or that they are the result of a recent range expansion into the Amazonian lowlands (Tirira et al. 2018). The objective of this study was to confirm the presence of C. thous in the northern Ecuadorian Amazon, using verified roadkill specimens and camera trap detections, and to provide new information on this species’ diet, ectoparasites, and mitochondrial genetic variation. These data offer novel insights into the ecology, biogeography, and potential differentiation of C. thous populations in the tropical rainforests of eastern Ecuador.

The specimens of Cerdocyon thous documented in this study represent some of the very few verified museum records of this species in Ecuador, reconfirming its presence in the country’s northern Amazon region. Although Tirira et al. (2018) hypothesized that C. thous may be expanding into Ecuador via recent colonization events, the evidence presented here (including low but confirmed photographic detections, roadkill records, and genetic distinctiveness) suggests that C. thous is present in intervening forest landscapes and areas of anthropogenic disturbance in the northern Amazon of Ecuador. Additional records are needed to determine whether these records represent a recent expansion or a long-established cryptic population. Three additional roadkill records from the northern Amazon on the citizen-science platform iNaturalist (https://ecuador.inaturalist.org/) suggest that C. thous is likely using road corridors to move through or into disturbed areas (Auz-Cerón et al. 2023; Santos et al. 2024). However, C. thous remains absent from national-scale roadkill datasets (Medrano-Vizcaíno et al. 2023), possibly due to low natural abundance or misidentification; C. thous is sympatric with two other native canid species in the Ecuadorian Amazon (Tirira et al. 2018; this study). Its absence from roadkill datasets is in contrast with high rates of roadkill reported for other areas of the Amazon region (Vázquez Rodríguez et al. 2016; Bauni et al. 2017; Rojano Bolaño and Ávila Avilán 2021).

Despite intensive camera-trap surveys across disturbance gradient in the northern Ecuadorian Amazon (Mena et al. 2020; Auz-Cerón et al. 2023; Jácome-Negrete et al. 2023), C. thous has rarely been recorded, implying either low abundance in intact forests or a strong preference for anthropogenically modified habitats. The low photographic capture rate in the western section of Yasuni Biosphere Reserve (this study) further supports this pattern, especially when compared to other canid species such as A. microtis and S. venaticus, which are recorded more frequently; this low capture rate is in contrast with assessments made in other areas of the species’ distribution (Jiménez-Alvarado et al. 2017; Caballero-Gini et al. 2020; Fonseca-Prada et al. 2022. Similarly, Fariñas-Torres et al. (2024) also reported low photographic rates in the southern portion of the species’ range, likely related to its occurrence near the edge of its known distribution. This scarcity may reflect either true rarity or avoidance behaviour (Smith et al. 2023; Crisfield et al. 2024). As a generalist species often associated with open or disturbed habitats, C. thous may be less suited to dense, undisturbed rainforest conditions (Maffei and Taber 2003; Jiménez et al. 2017; Castelló 2018).

Interspecific competition may also influence the occurrence of C. thous. Although direct evidence from Amazonian forests is lacking, studies in other regions (Bubadué et al. 2016) have shown that C. thous can compete with other canid species when sympatric, which could partly explain its tendency to use more degraded zones or ecotones rather than intact habitats (Maffei and Taber 2003; Jiménez et al. 2017; Castelló 2018). In these environments, it interacts with domestic mammals, such as domestic dogs, which are potential competitors for territory and food resources (Lacerda et al. 2009; Sánchez-Londoño 2014; Lessa et al. 2016; Jiménez et al. 2017) and are also vectors of pathogens that affect wild carnivores (Guzmán et al. 2024). In the western section of the Yasuni Biosphere Reserve, the photographic capture rate of domestic dogs increased in CFT2018 (0.42 ± 0.16 SE) compared to CFT2015 (0.05 ± 0.012 SE), indicating growing interactions between free-ranging dogs and native wildlife. This trend underscores the urgent need to implement measures that minimize dog–wildlife interactions, particularly in biodiversity-rich ecosystems such as Yasuni National Park.

The diet of C. thous has been the subject of several studies, especially in the Atlantic Forests and the Brazilian Pampa (Bueno and Motta-Junior 2004; Canesini et al. 2008; Dutra-Vieira et al. 2024). However, there are few reports on its diet in the Amazon rainforest (Dutra-Vieira et al. 2024). For the Ecuadorian Amazon, no data related to the diet of this canid have been documented (Vallejo 2022). The stomach-content analysis of the roadkill specimen from Dureno revealed a predominantly insectivorous diet, with insects and arachnids comprising nearly 40% of identifiable items, similar to that observed in other areas of the Brazilian Amazon (Dutra-Vieira et al. 2024). Vertebrate prey included one amphibian (Rhinella marina) and one fossorial snake (Amerotyphlops reticulatus). Although C. thous is known to consume snakes (Castellari Gonzalez et al. 2016; Dutra-Vieira et al. 2024) the presence of A. reticulatus in its diet represents a novel prey record for the species. These patterns are consistent with findings from forested areas in the Brazilian Amazon, where C. thous diet tends to include more invertebrates and small vertebrates compared to populations in more open environments (Dutra-Vieira et al. 2024). In contrast, studies from the Brazilian Cerrado, Atlantic Forest, and Pampas show diets richer in fruits, plant matter, and small to medium-sized mammals (Bueno and Motta-Junior 2004; Gatti et al. 2006; Pedó et al. 2006; Canesini et al. 2008). This opportunistic behaviour may reflect ecological adaptation to resource availability and lower visibility in dense forest environments, where large prey is harder to detect and capture. Such dietary plasticity reinforces the ecological generalist nature of C. thous but also highlights potential regional adaptations among forest-dwelling populations that warrant further investigation.

Across its geographical range, C. thous has been reported as a host for approximately 21 species of ticks. Most reports indicate a high prevalence of infestation by ticks of the genus Amblyomma (Ixodida, Ixodidae; Appendix 1), which commonly parasitise both wild and domestic animals during their adult stage (Nava et al. 2017). The presence of R. microplus, a species of tick introduced to the Neotropics from Europe via domestic cattle, represents the first such case of this parasite on C. thous in Ecuador. This ectoparasite has been previously recorded parasitising C. thous in Brazil and Venezuela (Jones et al. 1972; Rodrigues and Daemon 2004; Martins et al. 2014; Ramos et al. 2020) and is known to be a competent vector of various haemoparasites (Cruz-González et al. 2024). Its detection in a wild canid from a region undergoing increasing deforestation and cattle expansion raises concerns about spillover risks from domestic to wild species. Forest degradation may facilitate tick–host transmission cycles by increasing contact among wildlife, dogs, and livestock (Brandão et al. 2020; Cruz-González et al. 2024; Guzmán et al. 2024). The role of C. thous as a potential reservoir or host for pathogens in such altered landscapes merits focused study, especially in proximity to human settlements.

The mitochondrial Cyt-b analysis revealed that the Ecuadorian C. thous specimens form a distinct haplotype, with low but consistent genetic divergence from populations elsewhere. Phylogenetic analysis and genetic distance data suggest some geographic differentiation among South American populations of C. thous, although the limited sample size and reliance on mitochondrial markers constrain the strength of this inference. The Ecuadorian individuals form a well-supported haplotype; however, the possibility that they are related to populations from poorly sampled regions, such as trans-Andean Colombia, cannot be excluded. Brazilian specimens form a monophyletic group, while Argentine populations cluster closely in the phylogenetic tree; notably, the smallest observed genetic distance (0.58%) is between individuals from Argentina and Ecuador. This pattern contradicts the relative placement of these populations in the phylogenetic tree and suggests a more complex evolutionary history, potentially involving recent introgression or secondary contact, although additional data would be required to evaluate these possibilities. Currently, five subspecies (Cabrera 1931, 1958; Berta 1982; Bisbal 1988) of C. thous are recognized: C. t. thous (the Guianas, eastern Amazonia, and northern Brazil), C. t. aquilus (the savannas and forests of Colombia and Venezuela), C. t. azarae (north-eastern and central Brazil), C. t. entrerianus (southern Brazil, northern Argentina, Paraguay, western Bolivia, and Uruguay), and C. t. germanus (central Colombia); however, these require taxonomic evaluation to test their validity.

Representative genetic sequences are lacking for several subspecies and for many parts of the species’ geographical range, limiting our ability to confidently assign the observed lineages to any specific subspecies. However, our findings suggest some level of geographic structuring and, potentially, the emergence of a unique forest-associated evolutionary lineage. Although current data are insufficient to assign these individuals to a specific subspecies, their genetic differentiation, combined with ecological and geographical isolation, points to the possibility that the Ecuadorian population represents an evolutionarily significant unit (ESU). Additional sampling and integrative analyses, including morphological, genetic, and ecological data, are essential to test whether the Ecuadorian population corresponds to a previously recognised subspecies, represents a distinct geographic variant, or constitutes an undescribed evolutionary unit.

This study presents the first verifiable ecological and genetic documentation of Cerdocyon thous in the Ecuadorian Amazon, confirming its presence through museum specimens, camera trap detections, and molecular data. Our findings challenge the prevailing notion of the species as a recent colonizer in the Ecuadorian Amazon, instead suggesting the existence of a cryptic, potentially established population that persists in disturbed forest mosaics shaped by anthropogenic activity.

The low capture rates in intact rainforest environments, contrasted with roadkill and photographic records from altered landscapes, indicate that C. thous may be functionally excluded from undisturbed habitats by ecological filters such as dense vegetation structure with limited prey visibility, and interspecific competition with forest-adapted canids. Dietary analysis revealed a predominantly insectivorous and opportunistic feeding strategy that differs from the frugivorous diets reported for populations in open biomes, underscoring the species’ remarkable ecological plasticity and potential local adaptations to Amazonian rainforest environments. The detection of the invasive cattle tick (R. microplus) on a wild C. thous individual signals not only increased contact with livestock and domestic animals, but also the emerging risk of pathogen spillover in landscapes undergoing rapid deforestation and agricultural expansion. Mitochondrial sequence data place the Ecuadorian individuals in a distinct, well-supported haplotype, contributing new evidence of geographical genetic structuring across the range of C. thous and hinting at the presence of a potentially unique evolutionary lineage in the north-western Amazon. While our data are insufficient to delimit a new subspecies or evolutionary significant unit formally, the combination of ecological, morphological, and genetic distinctiveness calls for further integrative taxonomic assessments.

Collectively, these results highlight the urgent need to reassess the conservation status of C. thous populations in Amazonian Ecuador. Future research should prioritize broader sampling across environmental gradients, the ecological role of C. thous in disturbed forests and its interactions with domestic species. Given its apparent association with anthropogenic habitats, C. thous may serve as a sentinel species for ecosystem change, zoonotic risk, and the resilience of medium-sized carnivores in fragmented Neotropical landscapes. Safeguarding its future requires not only taxonomic clarity but also the development of landscape-level conservation strategies that address habitat degradation, road expansion, and domestic dog-mediated disease transmission in one of the most biodiverse regions of the planet.