The culpeo has been described as a generalist species, able to harness a diversity of food resources (Jiménez et al., 2008). In the high areas of the PNP the species has a predominantly carnivorous diet (98% of items are of animal origin). In this sense, the results match those reported in the pioneering studies of Crespo and Carlo (1963) in Argentina (prey animals with 96.9% FO), and Fugger (1979) in Chile (with 95% and 97% FO in Ayacucho and Cuzco, respectively). However, the trophic pattern found in southern Ecuador, regarding the type of prey and its frequency in the diet, is completely different from those reported to date.
Indeed, in contrast to studies carried out in other areas of Latin America, the culpeo in southern Ecuador seems to prey mainly on wild ungulates (Mazama and Pudu genus), which were the bulk of the diet in terms of both frequency of occurrence and biomass contribution. Therefore, the high representation of their remains in the scats (i.e. 70% FO) supports the idea that the culpeo actively hunts deer, being able to be an important predator of these species in the region. This pattern differs from all published studies, where cervids have little or even no representation. For example, in a study of the culpeo’s diet in the paramos of Peru (Romo, 1995), where deer were present, no deer remains were found in the culpeo scats. In the PNP cervids belonging to the Mazama and Pudu genus were the most consumed prey throughout the year, with a peak frequency in June, and high values for FO during April, May and August. Perhaps this pattern relates to a higher temporal availability of pregnant females, cubs and juveniles during these months. In Ecuador it is thought that deer can breed throughout the year, but unfortunately data on deer phenology in the different Ecuadorian regions is lacking. Months showing high consumption of deer could also coincide with a period of low availability of other potential prey. Small mammals were the second most consumed prey group in terms of FO (although its contribution to consumed biomass was apparently irrelevant) and consumption levels were particularly high during August, perhaps indicating a significant temporal variation in their abundance, population age structure or mean body weight throughout the year (Malo et al., 2013). Moreover, it has been observed that the higher the consumption of deer the lower the consumption of small mammals (precisely during January, February, March and September). Taking all the above into account, this variation could indicate temporary changes in the availability of these prey groups (Malo et al., 2013).
It is likely that the high consumption of deer (and of other animals such as carnivorous) in the PNP is in part due to the culpeo’s scavenging activity. Although scavenging on deer was not detected in Peru (Romo, 1995), general scavenging activity by culpeos has been reported on many occasions (Walker et al., 2007). However, PNP rangers have observed direct predation of culpeos on deer (PNP staff, pers. comm.), which is not strange when considering that consumed deer were medium-small species, mainly Mazama rufina (8-14 kg) and Pudu mephistophiles (5-10 kg) (see Tirira, 2007). Indeed, predation and attacks on animals of equal or even greater size, such as offspring and subadults of alpacas, vicuñas, llamas and guanacos, as well as sheep and goats, have been reported in Argentina and Chile (Novaro et al., 2000; Pia, 2003, 2004; Zacari and Pacheco, 2005). For example, Novaro et al. (2009) photographed and verified an episode of direct attack on young guanacos (33 kg), also observing continuous reactions of defense by guanaco family groups as soon as culpeos were sighted. Apparently, any animal weighing less than 35 kg can be potential prey for culpeos, given that this predator is able to kill individuals 6.6 times heavier than itself (see Jiménez and Novaro, 2004). Therefore, although culpeos could take deer meat from carcasses (left behind by the scarce numbers of pumas living in the area, for instance), it seems certain that in the PNP they often actively hunt deer.
The forest rabbit was the fourth most consumed species (showing a similar significance in biomass contribution) by the Andean fox. It is known that lagomorphs are important prey for this carnivore, with numerous studies highlighting the consumption of the European hare (Johnson and Franklin, 1994; Corley et al., 2005; Walker et al., 2007) and the European or wild rabbit (Iriarte et al., 1989; Rubio et al., 2013). It is interesting to note that only in Ecuador and northern Peru do the culpeo and the forest rabbit overlap in their respective ranges of distribution, the latter being the only lagomorph present in Ecuador. The consumption of forest rabbits tends to present temporal variation, showing FO peaks in the months of April and June (as with deer, although FO values for rabbits are much smaller). In fact, a negative correlation appears between the consumption of cervids and that of rabbits during April and February, which seems to be a trend for the remaining months. As previously commented in relation to deer, it is possible that the consumption pattern mimics the variation in rabbit availability in the environment (for example, due to potential seasonal breeding patterns of rabbits, although unfortunately no data is available concerning rabbit reproduction in the area).
This study also reports a new pattern found in relation to the consumption of carnivorous species (Palomares and Caro, 1999), which show a FO of around 10% (and a biomass contribution of 9%). In a few previous studies examining culpeo’s diet, carnivorous species are described as prey in less than 3% FO in the culpeo diet: for example, in Argentina Johnson and Franklin (1994) recorded the consumption of patagonian hog-nosed skunk (Conepatus humboldti) with an 0.8% FO; Walker et al. (2007) used the carnivorous taxon as a consumed group with an appearance in faeces of 2.5%; while in Zapata et al. (2005) this prey group represents 1.8% of the FO. Thus, the present work reports the highest value of consumption of carnivorous species, including those of medium size such as the Andean coati, the skunk, and also species of the genus Didelphis. In addition, hair remains of puma were found in scats collected during different samplings and times, thus belonging to three different individuals. In fact, puma remains were also found in the culpeo’s diet elsewhere (see Pia, 2013). There are two possible explanations for this finding: one, pumas were consumed as carrion (the most probable if these individuals were adults); and two, there was active predation on young pumas. In support of the latter it is known that coyotes kill adult bobcats (see Palomares and Caro, 1999), whose sizes could be similar in some cases to those of adult culpeos (especially males) and young pumas (especially females). Because these felids can kill culpeos (Pacheco et al., 2004; De Oliveira and Pereira, 2014), young pumas could be killed in turn by adult culpeos if given the opportunity, as a culpeo’s strategy for reducing the risk of intra-guild predation (e.g. Palomares and Caro, 1999). However, having only the evidence of hair remains in scats we were unable to differentiate between the two possibilities.
Interestingly, we have not found remains of livestock in scats, despite the presence of livestock in the PNP (cattle, horses, poultry, etc). It is possible that due to the high density and availability of wild prey in the PNP, attacks from Andean foxes to livestock do not occur, as often happens with large predators and other canid species (e.g. Imbert et al., 2016). Considering other prey groups, it is also notable that in the PNP bird consumption seems minimal (2.6% FO). Again, this pattern significantly differs from those found during other research, where birds attain a FO of almost 12% (Marquet et al., 1993; Romo, 1995; Walker et al., 2007). In the case of the nine-banded armadillo the FO found (1.3%) was similar to that described for other species of the same family (see Novaro et al., 2000; Zapata et al., 2005). The consumption of invertebrates was also low (0.7% FO). The importance of this group for the culpeo differs greatly among studies, as it is present in some (Ebensperger et al., 1991; Correa and Roa, 2005; Achilles, 2007; Guzmán-Sandoval et al., 2007) but not in others (Leon-Cobos 1994; Pia, 2013). Finally, the estimated biomass provided by birds, armadillos and invertebrates was almost irrelevant.
Fruit consumption showed low values (3% FO) compared to those observed in other studies, where the FO reaches values of almost 70% (Romo, 1995). Although this is not the only study where a low consumption is described (see Marquez et al., 1993), the culpeo in the PNP did not use this resource significantly at any particular time of the year, showing instead a markedly carnivorous diet. Some studies have observed a negative correlation between consumption of fruits and abundance of rodent (Castro et al., 1994; Silva et al., 2005) and invertebrate (Cornejo and Jiménez, 2001) prey availability. In the desert of Peru, Romo (1995) found a 70% FO of Vaccinium berries in the culpeo’s diet. This source of food is also abundant in the PNP, but consumption levels remain low. Thus, this could indicate a continuous supply of their most consumed prey throughout the year. Indeed, deer have long life cycles and their populations fluctuate less than those of smaller species such as rodents or rabbits. However, fruits might be more relevant in the diet if longer time series are considered, given that it has been observed that its importance can vary between years (Arim and Jaksic, 2005; Silva et al., 2005). Therefore, more studies are needed to assess the role of the Andean fox as a seed disperser in the high-Andean ecosystem.
In the study area the trophic diversity of Andean foxes fluctuated across time. During most of the year Shannon-Wiener’s index values remain below the average (i.e. H’ = 1.7), showing a relatively low dietary diversity. In fact, diversity values are higher than, or close to, the mean value during only four months. March was the month showing the highest value of trophic diversity, while January presented the lowest one, when the diet was clearly dominated by cervids. Lower trophic diversity values (H’ = 1.2 and H’ = 1.3) were described in Central Chile (see Rubio et al., 2013), whereas a high diversity value (H’ = 3) was found in the south of Peru (Cornejo and Jiménez 2001).
The results of this work, which is the first systematic approach to the feeding ecology of the Andean fox in Ecuador, show that the species could behave as a facultative specialist rather than a generalist (see Glasser 1982, 1984). This hypothesis seems more likely when studies from other countries are also taken into account. Thus, as described for other carnivores such as the European wildcat (Felis silvestris) (Malo et al., 2004; Lozano et al., 2006), the red fox (Vulpes vulpes) (Delibes-Mateos et al., 2008) and the pine marten (Martes martes) (Rosellini et al., 2008), the culpeo could specialise on different prey groups or mammal species according to their availability. In contrast to other regions, in the PNP the culpeo has specialised in consuming cervids (probably including carrion), although small mammals or alternative food types could become the bulk of the diet if deer availability decreases. Consequently more studies on feeding ecology, in particular accounting for prey availability, are needed to test this hypothesis. Furthermore, the important consumption of carnivorous species in the PNP seems to demonstrate that the culpeo is a top predator in the high Andes. If so, we can speculate that the culpeo, along with the puma, could regulate the populations of meso-predators, having a strong influence on both carnivore community structure and the functioning of the ecosystem functioning (see Ripple et al., 2014).This hypothesis has important consequences for conservation purposes. Nevertheless, more studies on population dynamics and the effects upon food-webs will be needed to gain a better understanding of the ecological role played by the culpeo, or Andean fox, in the high mountains of Ecuador.