The canine rabies elimination program in the city of Arequipa, Peru faces a unique threat that has hitherto been unreported. The geographic positioning of Arequipa amidst the climatically unforgiving Andean high desert was believed to shield the urban fauna (1) from contact with local wildlife disease reservoirs (2). Since the reintroduction of the rabies virus in Arequipa, 394 rabid dogs have been detected, averaging nearly one case per week in most years. As a result, present rabies control efforts, like those in most Latin American countries (3, 4), have focused mostly on populations of owned dogs (those with human caretakers, whether confined or allowed to roam) (1). Control efforts have included a combination of annual, city-wide mass vaccination campaigns, localized vaccination rings around positive cases, passive surveillance of dog populations, as well as methods that have been proven around the world to be ineffective, such as sporadic culling of free-roaming dogs (5, 6). Despite these efforts, cases have continued to spread over the past seven years (7). The scope of current control paradigms would be insufficient to address the complexity of the local disease ecology if feral dogs (1) are present in Arequipa.

Feral dog populations have broad negative impacts, from diminishing environmental quality (8–10), to displacing native wildlife (11, 12), to affecting the locals' livelihoods by targeting livestock (8, 13). Most importantly, they pose a significant health risk as potential vectors to a variety of pathogens that affect wildlife, domestic animal species, and human populations (14), such as canine rabies and distemper viruses. Aside from the human impact of rabies and the potential role of feral dogs in the virus transmission, there is also the risk of feral dogs introducing the rabies virus to vulnerable wildlife populations (10, 14, 15), such as the common Andean fox. The establishment of a new native vector and the potential for a novel rabies virus transmission cycle—such as the historical spillover from domestic dogs to red foxes in Europe (16)—demonstrates how such shifts can complicate efforts to eliminate the disease from a given region.

Recent accounts from peri-urban dwelling inhabitants have detailed numerous sightings and occurrences in which “cave dogs from the hills” have come into the periphery of the city to hunt local domestic animals, including adult pigs and sheep. Herders of medium-sized livestock have reported packs of dogs jumping fences, destroying corrals, and killing farm animals. The presence of feral, cave-dwelling dogs in peri-urban areas on the city's margins may represent an overlooked challenge for rabies control. Because unowned dogs are excluded from vaccination and sterilization campaigns, and dogs outside city boundaries are not systematically monitored by rabies surveillance systems (3), important knowledge gaps remain regarding their potential role in rabies transmission dynamics. Surveillance and understanding of the behavioral ecology and extent of interactions between populations of owned and unowned dogs in the region are essential to explain the potential contribution of unowned dogs to the local propagation of the canine rabies virus (17–19). In this context, characterizing the feral cave-dogs' diet provides insight into their ecological role, their interactions with humans, domestic animals, and wildlife, and the resources that sustain them. Diet directly reflects whether these dogs rely primarily on anthropogenic sources (garbage, backyard livestock) or on hunting/scavenging of domestic and wild animals (20). This distinction has important epidemiological implications: scavenging and predation increase the frequency of direct contact with unvaccinated dogs, cats, and livestock, thereby expanding potential rabies transmission networks across populations that are otherwise considered distinct (21).

In response to these growing concerns, our objectives were to establish the presence of feral dogs in peri-urban areas of Arequipa and examine the potential challenges these animals pose for rabies control programs, as well as the broader One Health implications for local wildlife conservation and the sustainability of small-scale livestock farming.

Our field investigations confirm the presence of feral, cave-dwelling dog populations in peri-urban areas bordering the city of Arequipa, Peru—an observation consistent with previous reports in Arequipa, and a problem that has been reported in other countries in the Americas (9, 12, 14, 25). Here, we provide new quantitative, demographic, and dietary data that substantiate and expand upon those findings. We identified feral dog populations inhabiting caves in the southeastern peri-urban outskirts of Arequipa, particularly in Alto Selva Alegre District. Following the survey presented in this study, we conducted exploratory visits to three additional districts surrounding the city and observed similar conditions—hundreds of caves inhabited by feral dogs. The presence of these dogs challenges previous assumptions that urban dog populations are largely isolated from wildlife disease reservoirs and unmanaged canine populations (7, 26).

Scat analysis, a widely used and cost-effective method for studying carnivore diets in free-ranging populations, provided robust qualitative and semi-quantitative evidence of the main taxa consumed by these dogs, even when direct observation was impractical (27). To reduce bias from differential digestibility and from the challenge of distinguishing scavenging from predation, we combined macroscopic identification with expert consultation and cross-referenced remains against local fauna lists. While scat analysis cannot capture every dietary detail, it was sufficient for our ecological and epidemiological objectives and offered unique insights into the subsistence strategies of these populations. Our dietary analysis of 194 fecal samples revealed evidence of a varied and opportunistic diet, highlighting that these cave-dogs are closely linked to human communities and other dog populations, creating additional pathways for rabies transmission beyond the vaccinated owned-dog population. Almost half of all recovered skeletal elements were unidentifiable—a proportion consistent with expected bone degradation from digestion and gnawing (28). The avian remains likely include domestic chickens, as suggested by the size and morphology of claw sheaths, skin, and bones. Informal conversations with local residents further support the notion that feral dogs hunt backyard livestock in addition to scavenging from household garbage (29–31), which is frequently discarded into dry riverbeds due to lack of formal waste management. Notably, canine skeletal remains comprised over one-third of all identifiable bones. Several fecal samples contained phalanges and metapodial elements from medium- to large-sized dogs, suggesting possible scavenging or intra-species predation—behaviors that may indicate nutritional stress or limited food availability. Although local accounts describe feral dogs hunting cats and other dogs in packs, further behavioral observation is needed to clarify whether these animals primarily scavenge (8, 32) or actively prey on conspecifics (33) and other species. Reliance on both anthropogenic resources and hunting of domestic or wild animals demonstrates that these dogs can persist independently in marginal habitats, supporting their potential role as reservoirs of infection.

Canine rabies remains endemic in the city of Arequipa despite ongoing control and elimination efforts (34, 35). Although suboptimal vaccination coverage among owned dogs is the primary factor sustaining transmission (36), unrecognized or unsurveyed subpopulations—such as feral, cave-dwelling dogs—may also contribute to the persistence of the virus (14). As the region approaches elimination thresholds, the relative importance of these overlooked groups is likely to increase. Rabies elimination strategies hinge on achieving and maintaining 70% vaccination coverage across the entire at-risk canine population (37). However, most vaccination campaigns are designed to reach owned dogs within established administrative boundaries, potentially excluding adjacent feral populations that remain unvaccinated and unsurveyed (3). Furthermore, these mass dog vaccination campaigns are only parenteral delivery, facing inherent limitations when targeting inaccessible or elusive populations such as feral cave-dwelling dogs. Our findings highlight the need to complement existing approaches with alternative tools that can extend coverage to inaccessible groups. Oral rabies vaccination (ORV) represents one promising strategy that has been successfully implemented in wildlife reservoirs and is gaining traction as a feasible option for dog populations with limited human contact (38). In the context of Arequipa, deploying ORV baits in peri-urban and cave-dwelling habitats could increase vaccination coverage among feral dogs that are unlikely to be captured or handled during conventional campaigns. Integrating ORV with community-based surveillance, humane dog population management, and ecological monitoring would strengthen a One Health approach, addressing both the epidemiological and environmental dimensions of rabies control.

Our findings underscore the spatial and ecological connectivity between urban, peri-urban, and rural dog populations, highlighting the need for expanded surveillance and control efforts that include feral dog groups. The presence of these unsurveyed populations may undermine progress toward regional and global rabies elimination goals (39). While the expansion of rabies in southern Peru has largely been attributed to human migration and the translocation of owned dogs (40), the widespread presence of feral dogs across the Arequipa region suggests that their potential role in disease persistence and spread warrants consideration. Notably, among the observed litters, the vast majority of puppies were female—an unusual deviation from the expected 1:1 sex ratio. Informal conversations with local residents indicated that people, including dog vendors, frequently visit the area to take male puppies, either for sale in local markets—where male dogs are preferred—or for use as guard animals. This ongoing removal and movement of unvaccinated puppies from unmonitored populations into the city presents an overlooked risk that should be considered in the design and implementation of rabies control programs.

In addition to their potential role in rabies epidemiology, the presence of these dogs has other important implications for public health. Settlements within a few hundred meters of these caves report some of the highest rates of dog bites globally, yet display low rates of post-exposure prophylaxis and healthcare-seeking behavior (41). While residents do not distinguish between city dogs and feral cave dogs in reports of bites, the spatial proximity and possible movement between these populations complicate surveillance and control efforts.

Worldwide, feral dogs also pose threats to local economies and biodiversity. Their presence is often first reported through farmers' and herders' complaints about livestock attacks (29–31). In ASA and surrounding districts, small-scale animal husbandry remains a key economic activity. Repeated predation on domestic animals undermines household income and food security. Moreover, wildlife authorities have documented feral dog predation on protected species, including South American camelids and Andean condors (42, 43). The potential for interspecific competition and disease transmission with native carnivores such as the Andean fox is a further concern. Feral dogs may act as reservoirs for pathogens such as canine distemper virus and sarcoptic mange, both of which can cause substantial mortality in wild canid populations (44, 45).

This study provides the first systematic documentation of feral dogs inhabiting caves in peri-urban Arequipa, highlighting an overlooked ecological niche with major implications for rabies control. A notable strength is the combination of direct and indirect evidence (photographs, sightings, feces, tracks, and food remains) to validate cave occupation, which increases confidence in the observations. The georeferencing of cave sites adds spatial epidemiological value, enabling future surveillance and interventions. Finally, the study addresses an urgent operational gap for public health programs, bridging ecological fieldwork with zoonotic disease control. The study also has limitations. First, it was observational and we did not measure rabies virus circulation limiting the ability to infer the direct role of cave dogs in transmission. Second, the sample size and geographic scope were restricted to a subset of peri-urban Arequipa, which may not capture the full variability of dog ecology across the city or other regions. Third, while photographic verification reduced double counting, individual identification methods such as tagging or genetic sampling were not used, leaving potential for misclassification. Lastly, the reliance on indirect signs of cave occupation (e.g., feces, paw prints) introduces some uncertainty, as these may persist after dogs abandon a cave.

Despite these limitations, the findings underscores the need for coordinated, cross-sectoral strategies to address the ecological and public health impacts of feral dog populations—an approach that has advanced slowly in Arequipa (46). Surveillance programs targeting cave-dwelling dogs are essential to understanding their movements, population dynamics, and role in zoonotic disease transmission. Sterilization and vaccination efforts could help reduce their ecological impact and interrupt rabies transmission chains. However, high population turnover remains a significant barrier to the long-term effectiveness of these interventions. Importantly, this challenge is not limited to feral populations: high turnover is also observed among owned dogs in Arequipa, particularly in periurban communities (47). Although we did not record any dead puppies during our surveys, this absence should not be interpreted as evidence of low pup mortality. Puppies are more difficult to detect due to their smaller size, greater use of dens, and rapid removal by scavengers or conspecifics. High juvenile mortality has been reported consistently in stray and feral dog populations, and our cross-sectional approach was not designed to track changes in litter size or survival over time. Future research using longitudinal monitoring of identified litters is needed to accurately estimate juvenile mortality, a critical parameter for modeling population turnover and rabies transmission dynamics. However, the high mortality observed among feral dogs may not represent a stable demographic trend. Reports from residents indicate that periodic poisoning or culling events occur in response to perceived threats posed by feral dog packs. Combined with the likely high rate of dog abandonment in peri-urban areas, these human-mediated factors may sustain the population despite insufficient reproduction. In areas like Arequipa city, vaccinated and sterilized feral dogs may be rapidly replaced by unvaccinated, intact dogs—often due to abandonment from nearby periurban neighborhoods. Establishing the prevalence of infectious diseases such as rabies, distemper, and mange in these feral and periurban owned dog populations will inform both wildlife conservation efforts and the design of effective public health interventions (48–50). Ultimately, controlling rabies and mitigating other health and environmental risks posed by feral dogs in peri-urban Arequipa will require adaptive, integrated One Health strategies (51). The One Health approach has been used successfully to implement cross-sectoral vaccination and surveillance programs in Africa, Asia, and the Americas to reduce dog-mediated rabies (52–54, 56), and it is a pillar of the global Zero by 30 initiative led by WHO, FAO, OIE, and partners (39, 55). By documenting the presence, ecology, and epidemiological significance of feral cave-dwelling dogs in Arequipa, our study contributes to this integrative perspective, highlighting an overlooked interface between domestic animals, wildlife, and human populations that must be addressed to achieve sustainable rabies elimination.