Exploring Vertebrate Behavior at Mesocarnivore Carcass Sites: Implications for Non-Trophic Parasite Transmission

The spatiotemporal distribution of parasites in ecosystems is heterogeneous. High infection risk is often associated with aggregations of animals around attractive resources. Here, we explore the behavior of potential hosts of non-trophically transmitted parasites at mesocarnivore carcass sites. We used videos recorded by camera traps at 56 red fox (Vulpes vulpes) carcasses and 10 carcasses of other wild carnivore species in three areas of southeastern Spain. In general, scavenging species, especially wild canids, mustelids and viverrids that display rubbing behavior, were more exposed to infection risk at carnivore carcass sites than non-scavenging and domestic species. The red fox was the species that most frequently contacted carcasses and marked and rubbed carcass sites. Foxes contacted heterospecic carcasses more frequently and earlier than conspecic ones, and close contact was more frequently observed at heterospecic carcasses. Thus, foxes seemed to avoid contacting carcasses in those periods and at carcass types of maximum risk of acquiring parasites. Overall, our results suggest that infection risk at carnivore carcass sites may take place mainly for visitor-carcass contact or contact with feces and urine rather than direct contact between visitors. Moreover, contact events between scavengers and carnivore carcasses are far more frequent than consumption events, which indicates that scavenger behavior is more constrained by the risk of acquiring meat-borne parasites than non-trophically transmitted parasites. This study contributes to lling key gaps in understanding the role of carrion in the landscape of disgust, which may be especially relevant in the current global context of emerging and re-emerging pathogens. Cazorla (22.2% of total carcasses in this area), three in Murcia (15.8%) and two carcasses of other mesocarnivores in Espuña (20.0% of total non-fox carcasses). Contact between individuals of different visiting species at carcass sites was recorded only once, between a golden eagle (Aquila chrysaetos) and a griffon vulture (Gyps fulvus) in Cazorla. suggesting scavenger behavior is more constrained by the transmission risk of meat-borne than the risk of non-trophically transmitted parasites. Overall, this study contributes to lling major gaps in the empirical knowledge of the role of carrion in the landscape of disgust (Moleón and Sánchez-Zapata 2021), and shows the promising and varied opportunities of studying animal behavior associated with carrion resources. The impact that emerging and re-emerging diseases associated with wildlife are having on modern societies makes it necessary to conduct these types of studies, providing key scientic evidence to improve our understanding of the epidemiological factors that occur in the wild.


Introduction
Pathogen distribution is spatially and temporally heterogeneous, so epidemiological landscapes frequently consist of hotspots for transmission risk within a matrix of area with reduced or even no exposure to parasites (Bousema et al. 2012;Buck et al. 2018; Weinstein et al. 2018a). Infection risk hotspots may be driven by the presence of attractive resources that favor aggregation of animals, such as water points and food-rich patches, or by speci c sites where pathogens are more like to accumulate, such as latrines (Buck et al. 2018; Weinstein et al. 2018aWeinstein et al. , 2018b. Epidemiological risk may also be increased by species-speci c behaviors, such as social interactions between individuals of gregarious species or family groups, or during certain times of year, such as the rutting season (Altizer et al. 2003; Patterson and Ruckstuhl 2013; Ezenwa et al. 2016). Infection risk also depends on the diversity of susceptible and alternative hosts in the environment (Johnson and Thieltges 2010). In this context, when food resources and other points of attraction are apparently infected, hosts must weigh the perceived infection risk against foraging gains and other bene ts (Weinstein et al. 2018b). Understanding host behavioral responses to potential risk of infection associated with food resources is relevant from an ecological and evolutionary perspective, but also provides a solid basis for better interpreting the epidemiological risk factors that favor the transmission of pathogens in the wild (Hart 1990;Kuris 2003;Penczykowski et al. infective stages, within-host replication rate, pathogen infectivity, the route of infection, the number of host species that are susceptible, and the life cycle they present (Poulin 2007; Alizon and Michalakis 2015; Acevedo et al. 2019;Brouwer et al. 2019;). The persistence outside the host of infective stages can vary from a few hours to many years, depending on pathogen characteristics and environmental factors (Traversa et al. 2014; Chenais et al. 2018). With regard to carcasses, it is assumed that, in general terms, the number of infective forms and their survival decreases as the distance to the carcass site increases and over time, although few studies have investigated this topic (Turnbull et al. 1998;Fialho et al. 2018; Rossi et al. 2019).
Among the non-trophically transmitted pathogens that cause the greatest impact on wildlife is the mite Sarcoptes scabiei, an obligate permanent parasite that causes sarcoptic mange (Niedringhaus et al. 2019). This multi-host ectoparasite is widely distributed and affects a broad range of mammals, including ungulates and carnivores (Carricondo-Sánchez et al. 2017; Pisano et al. 2019;Turchetto et al. 2020). These mites live in the epidermis of their hosts, and can be transmitted through direct contact between animals or indirectly when a susceptible host acquires free mites that have shed the skin of an infected animal, especially in dens and other sheltered sites where S. scabiei may survive for several days (Pence and Ueckermann 2002). Another infectious agent of major concern, due to its health impact on wildlife populations, is the bacterium Bacillus anthracis, which causes anthrax in ungulates and, to a lesser extent, in carnivores (Hugh-Jones and de Vos 2002). After the death of the infected animal, this virulent pathogen produces spores around the carcass that can persist in the environment for years, infecting new hosts via ingestion or inhalation (Bellan et al. 2013;Turner et al. 2014). Other widely distributed, nontrophically transmitted infectious agents that can seriously affect wild carnivore populations are rabies, distemper virus and canine parvovirus, which can be acquired through the saliva, respiratory secretions and feces of infected animals, respectively (Truyen et al. 1998;Nouvellet et al. 2013;McElhinney et al. 2014).
One of the paradigmatic hosts of these pathogens is the red fox (Vulpes vulpes), the most broadly distributed mammalian carnivore worldwide. This generalist species feeds upon a wide array of trophic resources, including vertebrate and invertebrate prey, plants Our main goal is to explore the behavior of potential hosts of non-trophically transmitted pathogens at carnivore carcass sites, with a special emphasis on the red fox. For this purpose, we monitored the decomposition process of fox and other mesocarnivore carcasses in several areas that differ in their communities of vertebrate carnivores and levels of anthropization. Analyzed behaviors include direct contact, marking and rubbing, either on the carcass or in its vicinities. Our main hypothesis is that the risk of acquiring pathogens through direct contact is dependent on both time since the carcass became available and carcass type (conspeci c vs. heterospeci c regarding the consumer), and that hosts rely on indirect cues to shape their behavior at carcass sites. Overall, we predict that risky behaviors will be more frequent at late stages of carcass decomposition and in heterospeci c carcasses. This study may provide important insights to further understand the landscape of disgust associated with carrion, as well as the possible epidemiological consequences of this host behavior (Buck et al. 2018;Weinstein et al. 2018a; Doherty and Ruehle 2020; Moleón and Sánchez-Zapata 2021). This kind of study may be especially relevant in the current SARS-CoV-2 pandemic context, which has highlighted the need to investigate the forms of transmission of this emerging pathogen (Wong et al. 2020) in wild species, especially in mesocarnivores (Leroy et al. 2020;Tiwari et al. 2020).

Study areas
Fieldwork was carried out in three mountainous areas of southeastern Spain: Sierras de Cazorla, Segura y Las Villas Natural Park (hereafter Cazorla; 2,099 km 2 , 38º09'N 2º44'W), Sierra Espuña Regional Park (hereafter Espuña; 178 km 2 , 37º51'N 1º32'W) and periurban areas of Murcia city (hereafter Murcia; 415 km 2 , 37º57'N 1º02'W). Natural vegetation in these three areas is dominated by pine forests (mostly Pinus halepensis at low altitudes and P. nigra and P. pinaster at higher altitudes), aromatic shrubs, and patches of oak forests (Quercus ilex and Q. faginea) (Rivas-Martínez et al. 1987). There is an altitudinal and meteorological gradient from Cazorla (500-2,107 m a.s.l.; mean annual temperature: 12-16ºC; mean annual precipitation: 300-950 mm) to Espuña In general, vertebrate communities are much richer in Cazorla, which holds a large resident population of obligate scavengers (i.e., vultures) and a wide variety of facultative scavengers. The scavenging community is similar in Espuña, though vultures are less abundant. In Murcia, vultures are mostly absent, and domestic carnivores, such as the dog (Canis lupus familiaris) and the cat (Felis silvestris catus), are more frequent. The fox is the commonest wild mammalian carnivore in the three study areas, though it is more abundant in Espuña than in Cazorla;

Data collection
A total of 66 mesocarnivore carcasses were monitored in Cazorla (n = 27 foxes), Murcia (n = 19 foxes) and Espuña (n = 20 carcasses, including ten foxes, four stone martens Martes foina, three Eurasian badgers Meles meles, two common genets Genetta genetta and one wildcat Felis silvestris silvestris) from November 2016 to March 2018. The main research model was the fox because it is the most abundant carnivore in the studied areas. Hereafter, carcasses of carnivores other than foxes are designated as "other carcasses". Carcasses came from authorized hunting (only in the case of foxes) and recent road kills (foxes and other carnivores). Immediately after collection, carcasses were eviscerated, and a serum sample was taken from each animal to perform enzyme-linked immunosorbent assays for antibody detection (ELISA kits, Ingenasa®, Madrid, Spain) against some infectious diseases (canine distemper virus CDV, feline coronavirus FCoV, canine and feline parvovirus CPV/FPV, feline leukemia virus FeLV and feline immunode ciency virus FIV). In addition, muscle samples from the base of the tongue, the forearms and diaphragm were processed by arti cial digestion to detect the presence of Trichinella spp. larvae (Kapel et al. 1994;Gamble et al. 2000). Carcasses used in the study were free from these pathogens, and no lesions compatible with sarcoptic mange, mycosis or other pathologies were detected. In the case of hunted foxes, the tissues adjacent to the shot were removed to eliminate any trace of lead. Carcasses were monitored using automatic cameras (Bushnell Trophy Cam and Bushnell Aggressor) until complete consumption (i.e., no remains, or only fur left) or for a maximum of 10 weeks if the carcass was not completely consumed (i.e., bones and skin remained). Cameras were placed in discreet locations close to the carcasses (3-4 m) and were programmed to record a 15-second video every minute when detecting movement. Each carcass site was visited weekly to check batteries and memory cards. Cameras provided information on the presence of vertebrate species and their behavior at carcass sites. Recorded vertebrate species were classi ed into three groups: "red fox", "other mammals" and "birds" (the reptile Timon lepidus was also included in this last group, due to the scarce number of events in which this species was recorded). Based on O'Brien et al. (2003) and Ridout and Linkie (2009), we de ned independent events for each carcass as: a) consecutive videos of unequivocally different individuals of the same species or individuals of different species; b) when individual identi cation was not possible, consecutive videos of individuals of the same species taken more than 30 minutes apart; and c) nonconsecutive videos of individuals of the same species. For each event, we recorded a) the species group, b) the number of different individuals, c) the existence of direct contact between the visitor and the carcass, d) the existence of marking behavior (urine and feces deposition), e) the existence of rubbing behavior, and f) the minimum distance between the visitor and the carcass ("contact": distance = 0 cm; "close": distance > 0-50 cm; "moderate": distance > 50-200 cm; "far": distance > 200 cm). These distance intervals were also used to classify marking and rubbing sites.

Data analyses: weekly behavioral patterns of species visiting the carcasses
We explored the general spatiotemporal patterns of mesocarnivore carcass use by the studied vertebrate communities. First, for each study area and carcass type (foxes and others), we calculated, on a weekly basis, the proportion of carcasses that were contacted (i.e., with at least one direct contact event), marked (i.e., with at least one marking behavior event), rubbed (i.e., with at least one rubbing behavior event on the carcass or on the ground next to it), and visited but not contacted (i.e., no contact events recorded), for all vertebrates together and separately for each vertebrate group. For each study area and carcass type, we also estimated the number of contact, marking, rubbing and no contact events per carcass. Second, we calculated the accumulated number of carcasses that were a) detected, b) contacted (i.e., at least one contact event), c) marked (i.e., at least one marking event), and d) rubbed (i.e., at least one rubbing event) each week by foxes.

Data analyses: determinants of fox behavior
We used generalized linear models (GLMs) to analyze the factors in uencing "time of rst contact" (only carcasses with at least one contact event by foxes were used; n = 54). We conducted two separate analyses, using these two different datasets: 1) all fox carcasses in the three study areas; and 2) both fox and other carcasses in Espuña only. The rst analysis is mainly aimed at exploring the general behavior of foxes at conspeci c carcasses, while the second is aimed at determining if fox behavior is in uenced by carcass type. Time of rst contact was calculated as the time elapsed since carcass detection by foxes until the rst contact event by foxes. The sample unit for these analyses was the carcass. The explanatory variables for the rst analysis were study "area" (Cazorla, Espuña, Murcia), "habitat" (close, open), "year", "season" (winter: November-February; spring: March and April), "hour" of carcass placement (morning: from dawn to 12:00h; afternoon: from 12:00h to dusk), and carcass "detection time" by foxes (i.e., time elapsed since carcass placement and its detection by fox, expressed in days).
The explanatory variables for the second analysis were "carcass type" (fox, other) and carcass "detection time" by foxes.
We then ran univariate models (Gaussian error distributions and identity functions) with all the possible explanatory variables for each case. Model selection was based on Akaike's Information Criterion, which allows the identi cation of the most parsimonious model (lowest AIC) and ranks the remaining models. For each model, the AIC value was corrected for small sample sizes (AICc). Then, delta AICc (ΔAICc) was calculated as the difference in AICc between each model and the best model in the evaluated set, and models with ΔAICc < 2 were considered to have similar support (Burnham and Anderson 2002). We calculated the deviance (D 2 ) explained by each candidate model according to this formula: D 2 = (null deviance -residual deviance) / null deviance *100 (Burnham and Anderson 2002).
Finally, we used Chi-square analyses to compare the minimum distance between visiting foxes and the carcass a) among study areas (only fox carcasses) and b) carcass types (only in Espuña). All analyses were done with R Studio software v1.0.143 (R Core Team 2018).

Visiting species
We recorded a total of 2,383 events (58.9% in Cazorla, 23 (Tables S1 and S2). Domestic species (mainly dogs, but also cats, goats and sheep) were rarely recorded (1.4% of total events; Table S2). The fox was the most frequently recorded species in the three study areas (40.3% of total events). Mean number of different individuals per event was 1.1 ± 0.9 (range: 1-29), and groups of visitors (i.e., more than one individual) were recorded at 8.0% of total events. Groups were more frequently recorded for carrion crow (Corvus corone), wild boar (Sus scrofa), mou on (Ovis aries musimon) and Eurasian jay (Garrulus glandarius) in Cazorla. The fox was very rarely observed in groups (Table  S2).

General patterns of contact, marking and rubbing behaviors
Contact events represented 40.6% of the total recorded events (Fig. 1, Table 1). Considering all study areas together, the fox was the species that most frequently contacted carcasses (45.0% of total contact events; Fig. 1, Tables 1 and S2). Intraspeci c contact was recorded at 100% of carcasses in Cazorla, 63.2% in Murcia, and 60.0% (fox carcasses) and 30.0% (other carcasses) in Espuña. In foxes, intraspeci c contact was detected in 43.4% of the total events recorded. In Espuña, events (especially contact events) of foxes and other mammals, but not of birds, were more frequently recorded at carcasses of other mesocarnivores (Fig. 1). Contact of both domestic and wild species with the same carcass took place at six carcasses in Cazorla (22.2% of total carcasses in this area), three in Murcia (15.8%) and two carcasses of other mesocarnivores in Espuña (20.0% of total non-fox carcasses). Contact between individuals of different visiting species at carcass sites was recorded only once, between a golden eagle (Aquila chrysaetos) and a griffon vulture (Gyps fulvus) in Cazorla. Table 1 Carcass use patterns per study area and carcass type, according to different vertebrate species groups. Number of monitored carcasses is indicated for each study area and carcass type. The number of carcasses visited, contacted, marked and rubbed by each vertebrate group is shown together with the percentage relative to the total carcasses monitored per area and carcass type (in parentheses). Mean number of events per carcass ± SD is shown for total, contact, marking and rubbing events. We considered carcasses contacted, marked and rubbed as those carcasses with at least one event with contact, marking or rubbing by a given vertebrate group. Similarly, we considered contact, marking and rubbing events as those events with at least one contact, marking or rubbing behavior recorded.  Fig. 1). Regarding total marking events, urination was more frequently recorded than defecation in foxes (85.2% of total marking events) and other mammals (73.9%).

Weekly patterns in fox behavior
Carcasses in all the study areas were detected by foxes from the rst week. The number of red fox contact events peaked in the second to sixth week in the case of fox carcasses in all areas. In Espuña, the peak for other carcasses took place in the second week, i.e., two weeks earlier than the peak for fox carcasses in this area. While the rst contacts with fox carcasses in Cazorla and Murcia, and with other carcasses in Espuña, were recorded in the rst week after their deployment, the rst events of contact with fox carcasses in Espuña were detected in the second week. In Espuña, foxes contacted more heterospeci c carcasses than conspeci c ones (Table 1, Fig. 1).
The accumulated number of fox carcasses contacted by fox ranged between 100% in Cazorla to 60% in Espuña; in the latter area, foxes contacted 90% of carcasses of other carnivores (Fig. 2). While marking by foxes was anecdotal for other carcasses (10%), foxes marked 40-74% of fox carcasses (Fig. 2). At conspeci c carcass sites, rubbing by foxes was less frequent than marking in all study areas, while the opposite was true for heterospeci c carcass sites (Table 1, Fig. 2).
According to the GLMs, with regards to fox carcasses in the three study areas, the time elapsed between carcass detection and contact by foxes was mostly related to habitat (Table 2), with foxes contacting carcasses sooner in open habitats (Table 3). Regarding carcasses of fox and other carnivores in Espuña, the time of rst contact by foxes was mainly dependent on carcass type (Table 2), with foxes contacting heterospeci c earlier than conspeci c carcasses ( Table 3); hour of carcass placement and carcass detection time by foxes also had an in uence (Tables 2 and 3). However, selected models explained little of the variability in the response variable, as revealed by their low D 2 values (< 11%; Table 2), which indicates that fox behavior was mostly conditioned by other variables not taken into account in this study. Table 2 AICc-based model selection to assess the factors in uencing "time of rst contact" by foxes of mesocarnivore carcasses in three study areas of southeastern Spain ("among areas" comparisons) and on conspeci c and heterospeci c carcasses in one of these study areas ("fox vs. other carcasses" comparisons). Explanatory variables include study "area", "habitat", "year", "season", "hour", and "carcass type" (see main text for details on the variables  Table 3 Generalized linear models (GLMs) showing the relationship between "time of rst contact" by foxes with the explanatory variables included in the selected models ("habitat": open, close; "carcass" type: fox, other; "detection time": carcass detection time by foxes; "hour": morning, afternoon). Only selected models are shown, ordered from highest to lowest D 2 . The estimate of the parameters (including the sign), the standard error of the parameters (SE) and the degree of freedom of the models (df) are shown. Fox behavior in relation to distance to carcass Most of the recorded events involving foxes occurred close to the carcasses (Fig. 3). The average distance between foxes and conspeci c carcasses was similar in the three study areas (Cazorla and Murcia:  Fig. 3).

Discussion
We carried out a detailed behavioral study of carnivore vertebrates, including scavenging and non-scavenging species, at mammalian carnivore carcass sites, which may represent hotspots for non-trophically transmitted pathogens. Visitor behavior at carnivore carcass sites Carnivore carcass sites were visited by a rich community of vertebrates, though their behavior differed widely among species groups, study areas and carcass types. The long persistence of mesocarnivore carcasses in the environment due to their relatively low consumption rate which lasted up to ten weeks. Contact between the visitor and the carcass was frequently recorded. However, direct contact between two different visitor species was hardly ever recorded, and it was never observed between mammals. This contrasts with herbivore carcasses, in which mammalian scavengers may have more opportunities for contact (Borchering et al. 2017), especially in the absence of competition with vultures (Ogada et al. 2012). In carnivore carcasses, visits of mammals are more spaced than in herbivore carcasses, where many scavengers can gather in the short interval during which meat is available. Thus, at carnivore carcass sites, infection risk may take place mainly for visitorcarcass contact rather than direct contact between visitors.
Visitor behavior at carcass sites was highly dependent on the scavenging habits of the species. In our study, scavenging species were responsible for most contact events (53.1-96.5%, depending on the study area; see Table S2). Contacts by non-scavengers were mainly by small passerine birds that were observed taking hair from carcasses for nest construction ( Rubbing, or scent-rubbing, is also very frequent in mammals such as carnivores, though the eco-evolutionary signi cance of this behavior is far from clear (Rieger 1979; Gosling and McKay 1990). In our study, direct contact with the carcass was much more frequent in rubbing events than in marking ones, which suggests that the risk of acquiring multi-host pathogens transmitted directly through non-trophic mechanisms, such as S. scabiei ( and viverrids that display rubbing behavior. In addition to ectoparasites, as the carcass decays, diverse endoparasite infective stages can spread around the carcass, such as Toxocara canis eggs, an intestinal nematode transmitted by fecal-oral route that affects domestic and wild canids (Roddie et al. 2008). Thus, touching, rubbing against the carcass or sni ng it can also be a route of contagion for this and other directly transmitted endoparasites.
Domestic species, represented by livestock (goats and sheep) and pets (dogs and cats), were recorded in a very low proportion of total and contact events, even for the most anthropized area (Murcia). This suggests that carnivore carcasses are not important hotspots of pathogen transmission for these species, at least in our study areas. There is general concern for rabies circulation among dogs, other domestic animals, wildlife and humans in several parts of the world (Hughes and Macdonald 2013; Nadin-Davis et al. 2021), though there are no cases of rabies in our study areas. However, these interactions must be considered to study other pathogens with high epidemiological relevance at the wildlife-domestic-human interface, such as SARS-CoV-2, which is characterized by rapid spread and interspecies-jumping capacity (Leroy et al. 2020). Further studies should be promoted in regions where potential contact between wildlife and domestic animals is higher.
Fox behavior in relation to carcass type We found important behavioral differences of red foxes at conspeci c and heterospeci c carcasses in Espuña. Foxes contacted heterospeci c carcasses more frequently and earlier than conspeci c ones, as con rmed by the GLMs, and close contact was more frequently observed at heterospeci c carcasses than at fox carcasses. Similarly, rubbing by foxes was more frequent at heterospeci c than conspeci c carcass sites in Espuña. All of this is in accordance with the hypotheses that, in general, infection risk is higher for phylogenetically related species (Huang et al. 2014), and that carnivores avoid feeding upon conspeci c carcasses because the risk of acquiring species-speci c meat-borne pathogens is at a maximum (Hart 2011;Moleón et al. 2017  Fox marking behavior was also conditioned by carcass type, as urination and defecation were more frequent for conspeci c carcasses. This behavior does not entail, a priori, a direct contact with the carcass, so the risk of acquiring some pathogens that are usually transmitted by direct contact and have reduced mobility outside the host, such as lice and especially S. scabiei, is greatly reduced (Millán et al. 2016). This also suggests that marking behavior of the red fox is weakly inhibited by the infection risk associated with the presence of carcasses. In mammalian carnivores, marking is mainly associated with intraspeci c communication (e.g. Sillero-Zubiri and Macdonald 1998). However, why foxes marked more conspeci c than heterospeci c carcasses is unclear. A possible explanation could be that fox carcasses are more attractive as long-term marking points than carcasses of other mesocarnivores. This is because the persistence of fox carcasses in the environment is higher than that of other mesocarnivore carcasses, as foxes are more prone to feed upon heterospeci c carrion (

Conclusions
Here we disentangled the behavior of animals visiting mesocarnivore carcass sites, which may have important implications not only for understanding the epidemiology of non-trophically transmitted parasites, but also in eco-evolutionary terms. Contact events between scavengers and carcasses were far more frequent than consumption events ; Muñoz-Lozano et al. 2019; Gonzálvez 2020), suggesting that scavenger behavior is more constrained by the transmission risk of meat-borne parasites than the risk of acquiring nontrophically transmitted parasites. Overall, this study contributes to lling major gaps in the empirical knowledge of the role of carrion in the landscape of disgust (Moleón and Sánchez-Zapata 2021), and shows the promising and varied opportunities of studying animal behavior associated with carrion resources. The impact that emerging and re-emerging diseases associated with wildlife are having on modern societies makes it necessary to conduct these types of studies, providing key scienti c evidence to improve our understanding of the epidemiological factors that occur in the wild.

Declarations
Funding MM was supported by a research contract Ramón y Cajal from the MINECO (RYC-2015-19231). This study was funded by the Spanish Ministry of Economy, Industry and Competitiveness and EU ERDF funds through the project CGL2017-89905-R.

Con icts of interest
None.

Availability of data and material
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Code availability
Not applicable.

Authors' contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by MG, CMC and MM. The rst draft of the manuscript was written by MG. CMC and MM revised the manuscript, and all authors commented on previous versions of the manuscript. All authors read and approved the nal manuscript.
Ethics approval Not applicable.

Consent to participate
Authors have permission to participate.

Consent for publication
Authors have permission for publication. Figure 1 Weekly variation in patterns of use of mesocarnivore carcasses by red fox, other mammals and birds in three areas of southeastern Spain. A) Weekly percentage of contacted (i.e., with at least one contact event), non-contacted (i.e., visited, but no contact events recorded), marked (i.e., with at least one marking event), and rubbed (i.e., with at least one rubbing event) carcasses by red fox, other mammals and birds per study area and carcass type. B) Weekly number of contact, non-contact, marking, and rubbing events by red fox, other mammals and birds per study area and carcass type. For a given week, the number of events is divided by the total number of carcasses studied in each study area, and the number of carcasses available is given in parentheses. Panels for carcasses of carnivores other than foxes are in boxes.

Figure 2
Accumulated weekly number of detected (i.e., with at least one event recorded), contacted (i.e., with at least one contact event), marked (i.e., with at least one marking event), and rubbed (i.e., with at least one rubbing event) carcasses by the red fox per study area and carcass type.
Dotted horizontal gray lines represent the accumulated number of available carcasses. For a given week, the number of carcasses available is given in parentheses. Panel for carcasses of carnivores other than foxes is in the box.