OUP user menu

Seasonal Diet of the Pampas Fox (Lycalopex gymnocercus) in the Chaco Dry Woodland, Northwestern Argentina

Omar Varela, Ainhoa Cormenzana-Méndez, Lucía Krapovickas, Enrique H. Bucher
DOI: http://dx.doi.org/10.1644/07-MAMM-A-125.1 1012-1019 First published online: 15 August 2008


The seasonal diet of Pampas foxes (Lycalopex gymnocercus) was examined from 431 scat samples collected during 1 year in the dry Chaco woodland of Salta Province, northwestern Argentina. Fleshy fruits dominated the diet during the dry and wet seasons, representing 69% of the total frequency and comprising 91% of the total volume. Most of the fruit diet was contributed by 5 woody plant species. Mistol (Ziziphus mistol, Rhamnaceae), was the most important fruit resource throughout the year. Arthropods (particularly beetles and scorpions) and small mammals (mainly rodents and marmosets) were the most frequently consumed animal prey. Seasonal shifts in the frequency of the main food items in the diet were not significant, but there was considerable seasonal variation in the frequency of fruit species. Diversity of prey in the diet was highest in the wet season. The Pampas fox is a generalist feeder that can alter its diet to changes in food availability. However, its diet in the Chaco, which is dominated by fruit, differs from its diet in Pampas grassland habitat, where it consumes mostly meat.

Key words
  • Argentina
  • diet
  • dry Chaco
  • frugivory
  • Lycalopex gymnocercus
  • Pampas fox

Based on the breadth and variability of their diets, carnivorous mammals are often characterized within a spectrum ranging from specialists to generalists (Jędrzejewska and Jędrzejewski 1998; Jędrzejewski et al. 1989). Generalist species distributed across different habitats and ecoregions may reveal large variation in their diets, whereas specialists tend to show a rather similar trophic niche despite significant variation in habitat characteristics (Elmhagen et al. 2000; Kaneko et al. 2006). A comparative analysis of resource use along geographic gradients by a given species is therefore a useful tool for understanding its degree of trophic specialization, information that in turn may be useful in designing conservation strategies or improving the efficacy of management efforts (Benson and Chamberlain 2006; Cole et al. 1995; Connell 1975; Tokeshi 1999).

Canids (wolves, coyotes, and foxes) are a conspicuous component of mammalian communities in South America, and are usually the dominant group in the carnivore guild (Medel and Jaksic 1988). They inhabit several kinds of habitat and show high flexibility in their ecological requirements (Johnson et al. 1996). Usually, foxes are considered generalist omnivores that feed opportunistically on a wide variety of foods (Sillero-Zubiri et al. 2004). The Pampas fox (Lycalopex gymnocercus; formerly Pseudalopex gymnocercus) is one of the commonest carnivores in South America (Medel and Jaksic 1988). A small (4- to 6-kg), solitary, and mainly nocturnal species, it is relatively abundant in grasslands and open woodlands throughout Bolivia, Paraguay, Brazil, Uruguay, and Argentina, particularly in the Chaco, Monte, and Pampas ecoregions (Redford and Eisenberg 1992). The distribution of the Pampas fox in Argentina ranges from the province of Formosa in the north to the province of Rio Negro in the south (Redford and Eisenberg 1992).

Most of available information about food habits of the Pampas fox refers to populations from the grassland regions of northern Patagonia (Castillo 2002; Crespo 1971; Farías and Kittlein 2008; García and Kittlein 2005). Food habits of this canid in scrublands and sand dune habitats (García and Kittlein 2005) and forests (Vieira and Port 2007) have been less studied. These studies concur in defining the Pampas fox as an omnivore generalist, with a diet dominated by animal prey, particularly wild mammals and insects.

In the Chaco woodland, the 2nd largest biome of South America, the Pampas fox is probably the most common native carnivore. The species is very abundant in areas close to human dwellings, to the point of being considered a threat to domestic animals. Food availability in the Chaco is likely to differ from that in the Pampas grasslands because the preferred mammal prey species in the Pampas, such as the European hare, do not occur in the Chaco, and small rodents are usually less abundant, particularly in overgrazed areas (Bucher 1982). The combination of reduced availability of mammals and abundance of fruit and seeds in scats provided the stimuli for the present study. Here we report the seasonal diet of the Pampas fox throughout a 1-year study in the Chaco of Salta Province, northwestern Argentina, based on analysis of 431 scats. Our objectives were to describe and quantify the seasonal composition of the diet of Pampas foxes, and to compare our results with those reported for other biomes in Argentina and Brazil.

Materials and Methods

Study area.—Fieldwork was conducted at Los Colorados Biological Station and in the neighboring Campo Grande Ranch, both sites located in Salta Province, Argentina (24°43′S, 63°17′W). The area is a vast plain at about 200 m above sea level. The climate is subtropical semiarid, markedly seasonal, with a distinct warm and wet season in midspring–summer (October–March) and a dry, cool season in autumn–winter (April–September). Annual precipitation averages 550 mm, with 80% concentrated between December and February (Bianchi and Yañez 1992). Mean annual temperature is 21.7°C and mean maximum and minimum temperatures are 35.8°C (December) and 7.9°C (July), respectively (data from the locality of J. V. González 1944–1950—Galmarini and Raffo del Campo 1964). Vegetation is the typical semiarid woodland of the Western Chaco (Cabrera 1994). Dominant trees are Schinopsis lorentzii (Anacardiaceae) and Aspidosperma quebracho-blanco (Apocynaceae). The most abundant middle-sized trees are Ziziphus mistol (Rhamnaceae), Prosopis elata (Fabaceae), and Prosopis nigra (Fabaceae). The dense 3- to 4-m-high shrub layer is dominated by species of Acacia (Fabaceae), Capparis (Capparaceae), and Celtis (Ulmaceae), intermingled with grasses (Várela 2001). The area around the biological station (10,000 ha) is fenced and has been protected from grazing by domestic herbivores for the last 30 years. In contrast, the neighboring Campo Grande Ranch has been subjected to continued and severe overgrazing for a longer period, mostly by cattle and goats.

Food habits.—We used scat analysis to examine the diet of the Pampas fox. This is a noninvasive and relatively reliable method, and commonly used in studies of carnivore diets (Corbett 1989; Putman 1984; Reynolds and Aebischer 1991). Fresh scats were collected from different sites frequently visited by Pampa foxes, including paths, dirt roads, and bald spots (“peladares”), during 5 days in each month (from February 1995 to January 1996). All scat fragments within 0.5 m2 were considered as a single defecation. We took special care not to collect more than 1 scat from the same place, which could have been from the same individual. Given the spatial and temporal distance between sampling sites, we considered each scat as an individual foraging event. Scats were identified by size or diameter, shape, texture, characteristic odor, deposition place, presence of hairs ingested while grooming, and tracks close to the fecal samples. Those scats that could not be accurately attributed to foxes were discarded.

Fecal samples were individually stored in numbered paper bags and labeled with date of collection and location. Scats were air dried until constant weight and then analyzed in the laboratory following a standardized procedure (Korschgen 1987). Individual scats were weighed, manually crumbled, and then strained through a 0.5-mm-mesh sieve to discard the inconspicuous and unidentifiable fraction, which was excluded from the analysis. Sieve contents were placed in a petri dish and examined under a 4–40 × magnification binocular microscope. All distinguishable macroscopic components (head capsules, mandibles, legs, hairs, bones, teeth, jaws, claws, scales, vertebrae, feathers, bones, feet, seeds, and any other identifiable remains) were removed and identified to the finest taxonomic resolution possible. Mammals, reptiles, and arthropods were identified by comparison with collections of reference material deposited at the Fundación Miguel Lillo (Tucumán, Argentina). Fruits and seeds were identified by comparison with a reference collection from the study area. Birds were identified only to class.

Data analysis.—We determined the importance of the each food item in the diet in terms of 4 parameters: frequency of occurrence, percent frequency, relative frequency, and relative volume. We defined frequency of occurrence as the total number of scats in which a given food item was found. Percent frequency is the percentage of the total scats in which each food item was found. Relative frequency is the percentage of the total frequency in which each food item occurred. Relative volume is the percentage of the total volume in which each food item occurred (Korschgen 1987). The volume of each scat sample was measured using a graduated cylinder to the nearest 1 ml. We were unable to quantify the volume for most of the animal remains present in the scats because not all animal species could be identified to species. To minimize over-estimation of food items that contained a high proportion of nondigestible material (e.g., arthropods with chitinous exo-skeleton, fruit cuticles and fibers, nuts, and hard seeds) when reported by frequency of occurrence (Dickman and Huang 1988; Korschgen 1987) items <1 ml of volume were excluded from the analysis.

We used chi-square tests (Zar 1996) to compare the total frequency of occurrence of the main food items (fleshy fruits, arthropods, and vertebrates) as well as to test for seasonal differences in the frequency of occurrence of the main food items and among the frequency of fruit species and seasons. Seasons, defined according to rainfall pattern (dry and wet seasons), determined pulses of fruit scarcity and abundance (Varela 2004). Differences were considered significant at the probability level of P < 0.01.

Trophic niche breadth and trophic diversity between seasons were calculated using the standardized Levin index (BstdKrebs 1989) and the Shannon–Wiener index (Krebs 1989), respectively, based on frequency of occurrence of a single food item. By food item, we considered any taxon that could be identified in the scats. For some categories, this taxon was the species, but higher taxonomic levels (genus or order) were used for others. The Levins index formula is: Embedded Image where n is the number of food categories and p is the proportion of records in each food item (i). The standardized form of the formula is: Bstd = (B – 1)/(Bmax – 1), where Bmax is total number of food categories recognized. The scores of this index range between 0 (minimum niche breadth) and 1 (maximum niche breadth). The Shannon–Wiener diversity index (H) was calculated as H’ = —∑pi × log2 pi, where pi is the proportion of the frequency recorded for the item “i,” and log2 is the logarithm in base 2. The diversity values obtained were subsequently compared by Hutcheson's test (Krebs 1989). Rarefied species accumulation curves (sensu Gotelli and Colwell 2001) were produced by using Estimate S 8.0 (Colwell 2006) to determine whether an asymptote had been reached, and therefore whether the diet of Pampas foxes in the study area was adequately sampled.


Sampling effort.—We collected and analyzed a total of 431 scats of Pampas foxes over a 1-year period, 209 during the dry season (April–September) and 222 during the wet season (October–March). Based on the cumulative curves of food items (rarefied curves) for the wet and dry seasons, which reached an asymptotic at a sample size of about 200 scats, we inferred that the sample size was sufficient to characterize the seasonal diet of Pampas foxes in this study.

Diet composition.—We identified 36 different food items (12 plants and 24 animals) >1 ml in volume (Table 1). This number is an underestimate because we were unable to identify all samples to species. The diet was predominantly composed of fleshy fruits (including seeds), followed by arthropods and vertebrates (Fig. 1). Fleshy fruits occurred in 96% of the scats, making up 91% of the total scat volume (Fig. 1), and were significantly more frequent than animal food categories (χ2 =339.3, d.f. = 2, P < 0.0001).

Fig. 1

Seasonal composition of major food items in the diet of Pampas foxes (Lycalopex gymnocercus) in the Chaco dry woodland of Salta, Argentina. Bars indicate the percentage of the total frequency or total volume in which each food item occurs. Number in parentheses indicates sample size.

View this table:
Table 1

Diet composition of the Pampas fox (Lycalopex gymnocercus; scat analysis) in the Chaco dry woodland of Salta, northwestern Argentina, sampled during the wet and dry seasons of 1995. Values indicate number of each item that occurred in scats (FO), percentage of the total scats in which each food item was found (FO (%)), and percentage of the total volume in which each food item occurred (RV (%)). Unidentified items (NI) are specified in parentheses, n = number of scats. The life forms are defined according to Zuloaga and Morrone (1996). A plus (+) indicates species included in the Pampas fox diet that are listed for the 1st time.

Food itemsWet season (n = 222)Dry season (n = 209)
Species (family) (life form)FOFO (%)RV (%)FOFO (%)RV (%)
Fleshy fruits21496.489.219593.392.5
+Acacia aroma (Fabaceae) (shrub)156.83.85827.822.8
+Acanthosyris falcata (Santalaceae) (tree)115.04.700.00.0
+Bromelia hieronymi (Bromeliaceae) (herb)62.70.9125.70.8
+Capparis speciosa (Capparaceae) (shrub)
+Celtis pallida (Ulmaceae) (shrub)
Geoffroea decorticans (Fabaceae) (tree)10.50.300.00.0
+Grabowskia duplicata (Solanaceae) (shrub)10.5<
+Prosopanche americana (Hydnoraceae) (parasite)83.60.4104.80.9
+Prosopis (P. nigra + P. elata) (Fabaceae) (tree)5926.623.3167.74.4
+Prosopis torquata (Fabaceae) (shrub)4821.65.04220.15.7
+Ximenia americana (Olacaceae) (shrub)52.30.400.00.0
+Ziziphus mistol (Rhamnaceae) (tree)14063.148.215172.257.6
Anoploderma (Cerambycidae)10.500.0
Aplagiognathus (Cerambycidae)62.700.0
+Calocomus desmaresti (Cerambycidae)10.500.0
Dichotomius (Scarabaeidae)2913.173.3
+Entomoderes satanicus (Tenebrionidae)104.5104.8
Glyphoderus (Scarabaeidae)10.510.5
Megelenophorus (Tenebrionidae)00.010.5
Pelidnota (Scarabaeidae)10.500.0
Pinotus (Scarabaeidae)10.500.0
+Psectrascelis ursina (Tenebrionidae)3114.02813.4
Strategus (Scarabaeidae)2913.12210.5
+Scotobius hystricosus (Tenebrionidae)167.2115.3
Orthoptera (NI)
+Scolopendra viridicornis (Scolopendridae)10.510.5
Timo genes (Bothriuridae)104.5115.3
+Timogenes dorbignyi (Bothriuridae)10.573.3
Wild mammals146.31.32612.42.9
Thylamys pusillus (Didelphidae)00.052.4
+Sylvilagus brasiliensis (Leporidae)00.010.5
+Dolichotis salinicola (Caviidae)10.500.0
Galea musteloides (Caviidae)10.541.9
Microcavia australis (Caviidae)31.441.9
+Graomys griseoflavus (Muridae)41.894.3
Domestic mammals20.90.510.50.1
Bos taurus (Bovidae)10.50.310.50.1
Capra hircus (Bovidae)
Birds (NI)135.91.373.30.9
Sauria (lizards) (NI)52.321.0
Serpentes (snakes) (NI)52.352.4

The fruit diet comprised 13 species of fleshy-fruited plants. Shrubs and trees were the best-represented life forms among the 4 life forms present (Table 1). Z. mistol was the predominant fruit in the diet, making up 48% of the total frequency of occurrence and 58% of the total fruit volume. Most of the fruit diet was contributed by Z. mistol, Prosopis torquata, P. nigra, P. elata, and Acacia aroma, which together accounted for 87% of the total frequency and for 94% of the fruit volume (Fig. 2). Other fruits (corn and dry fruits of 3 wild grass species) that were infrequent in the scats (volume < 1 ml) were not considered in the analysis.

Fig. 2

Seasonal composition of fruit species in the diet of Pampas foxes (Lycalopex gymnocercus) in the Chaco dry woodland of Salta, Argentina. Bars indicate the percentage of the total frequency or total volume in which each food item occurs. Number in parentheses indicates sample size.

The most important animal remains were arthropods (26%), followed by mammals (9.3%), birds (4.6%), and reptiles (3.2%). Most of the arthropod remains were coleopterans (21.3%; particularly Tenebrionidae and Scarabidae) and scorpions (7.4%). The most frequently found mammals were small rodents (6.3%) and mouse opossums (1.2%). Medium-sized mammals such as Dolichotis salinicola (Rodentia) and Sylvilagus brasiliensis (Lagomorpha) appeared in only 1 scat sample. Remains of domestic mammals (goats and cows) were occasionally present in the scats (0.7%) and contributed a negligible volume (0.3%). Birds and reptiles (lizards and nonvenomous snakes) accounted for small part of the overall volume (1.1% and 0.9%, respectively). Neither amphibians nor fishes were present.

Seasonal use of food.—There were no significant differences in the number of occurrences of the main food categories (fleshy fruits, arthropods, and vertebrates) between seasons (chi-square test of association: χ2 = 3.13, d.f. = 2, P =0.21; Fig. 1). However, the frequency of occurrence of fruit species (Fig. 2) was significantly different between seasons (chi-square test of association: χ2 = 90.74, d.f. =11, P < 0.0001). The partial components of chi-square showed that the differences occurred mainly due to the unequal frequency of fruits of A. aroma and Prosopis between seasons. Fruits of A. aroma were more frequent than expected in the dry season during its peak of availability, whereas the opposite happened with the frequency of occurrence of fruits of Prosopis (P. nigra and P. elata). Fruits of Z. mistol were highly consumed in each season (Fig. 2).

Trophic niche breadth.—Dietary breadth of Pampas foxes was slightly higher in the wet season (Bstd = 0.205) than in the dry season (Bstd = 0.195). Likewise, the trophic diversity of the diet was significantly higher in the wet season (Hwet = 3.82) as compared to the dry season (Hdry = 3.44; t = 3.22, d.f. = 895, P < 0.01).


Our study confirms the generalist–omnivorous nature of the diet of the Pampas fox, as compared with previous studies (Crespo 1971; Farías and Kittlein 2008; García and Kittlein 2005; Vieira and Port 2007). Prey types ingested by Pampas foxes in our study area are similar to what was found in other regions of the fox's distribution, although several species identified here have not been previously reported as part of the diet of Pampas foxes (Table 1). We cannot determine whether the observed differences resulted from differences in seed availability between sites or, alternatively, the new diet items did not occur in grassland sites where the diet of Pampas foxes has been reported.

The most striking difference between our results and those of previous studies is the considerably higher proportion of fleshy fruits taken by foxes from the Chaco, which contrasts with the prevalence of animal matter in the diet of Pampas foxes from other regions. Fleshy fruits comprised more than two-thirds of the diet of Pampas foxes in the Chaco dry woodland of Salta, as compared with only one-fourth of the diet in the grasslands of La Pampa Province (Crespo 1971). Moreover, recent studies (Castillo 2002; Farías and Kittlein 2008; García and Kittlein 2005; Vieira and Port 2007) conducted in different habitats (grasslands, scrublands, sand dunes, and forests) also agree in that Pampas foxes feed mainly on animal prey (particularly mammals and insects). In such studies, fruits were important in certain seasons only, whereas in the Chaco dry woodland fruits dominated the diet of Pampas foxes throughout the year. The ability of Pampas foxes to survive on a low-meat diet in the Chaco is comparable to the differences in diet of Darwin's fox (Lycalopex fulvipes) between the Island of Chiloe and continental areas of Chile (Jimenez et al. 1991).

The regional differences observed in the diet of Pampas foxes possibly are a response to food availability. Although our study does not provide data on this aspect, it is well known that availability of fleshy fruits is relatively higher in the Chaco woodland than in the grassland habitats, where most of the fruit-producing trees and shrubs are rare or absent. Moreover, rodents are relatively scarce in the Chaco (Chani et al. 1998). European hares (Lepus europaeus), the most commonly consumed prey in the Pampas (Crespo 1971), also are absent in our study area.

The use of fleshy fruits by Pampas foxes closely agrees with fruiting patterns in the study area. Fleshy fruits were slightly more diverse in the diet during the wet season, when more plant species have ripe fruits. The high rate of fruit consumption by Pampas foxes in the Chaco dry woodland is comparable to few examples recorded in the canid literature (Geffen et al. 1992; González del Solar et al. 1997; Hockman and Chapman 1983; Jimenez et al. 1991) and confirms recent studies showing the species’ importance as a fruit-dispersal agent (Varela 2004; Varela and Bucher 2006).

It has been well documented that the diet of canids responds to shifts in seasonal availability of food (Castillo 2002; Elmhagen et al. 2000; Hockman and Chapman 1983; Kaneko et al. 2006; Motta-Junior et al. 1996; Pigozzi 1992). According to the energy optimization model (MacArthur and Pianka 1966), generalist species increase their feeding efficiency by shifting to other food resources when fleshy fruits become scarce. Our results show that the proportion of fruit in the diet of Pampas foxes remained relatively constant throughout the seasons, despite changes in the proportions of fruit species.

Of particular interest is the high proportion of fruit consumed by Pampas foxes during the period of highest fruit scarcity (dry season: July–October), when only a few woody plants species (e.g., Z. mistol, A. aroma, and Caesalpinia paraguariensis) are fruiting in the study area. Preference for fruits of Z. mistol may be explained by several reasons: the long fruiting season of this species (from mid-October to late July); regular productivity (we observed plants fruiting every year during 10 consecutive years); prolonged availability of undamaged fruits on the ground, which is favored by the low water content of the fruits as well as low humidity and relatively low temperatures during the dry season; high palatability observed in captive individuals; and high availability (mistol is the 2nd most abundant woody species with fleshy fruits in the study area—Varela 2004). The fruit of Z. mistol also may be preferred because of its nutritional content, although data on this aspect are not available. Chemical studies of fruits of congeneric species, such as Z. jujuba (Guil-Guerrero et al. 2004), Z. mauritiana (Muchuweti et al. 2005), and Z. obtusifolia (Everett 1986), showed the presence of fatty acids, proteins, and sugars.

Although a wide variety of fruit species were available, the Pampas fox appears to be highly selective. Fruits of some very abundant species (e.g., Castela coccinea and Capparis tweediana) were not included in the diet of Pampas foxes. Chemical compounds (triterpenoids, alkaloids, and phenols) present in these fruits may act as feeding deterrents for the three-banded armadillo (Tolypeutes matacusBolkovíc et al. 1995) and probably also for the Pampas fox. The fruits of C. tweediana might have similar compounds, because it has been reported that Chaco aborigines washed the seeds to eliminate the chemical compounds present in the fruit pulp (Arenas 1981).

Meat from domestic animals (young goats and cows) was occasionally present in the scats. It is likely that Pampas foxes occasionally scavenge on carrion (Crespo 1971), because it is unlikely that Pampas foxes would be able to kill a goat or sheep. We conclude therefore that in our study area predation by foxes on domestic livestock is insignificant.

Our findings support previous studies indicating that the Pampas fox is an omnivorous and generalist feeder that can adapt its diet to changes in food availability (Castillo 2002; Crespo 1971; Vieira and Port 2007). However, examination of our data also shows the high dominance of fruit (especially of Z. mistol) as a staple in the diet of Pampas foxes in the Chaco, suggesting that in the Chaco, Pampas foxes have a more significant role as fruit eaters and seed dispersers than as vertebrate predators.


We thank the Fundación para Desarrollo del Chaco for giving permission and logistical support to conduct this research at Los Colorados Biological Station. Fieldwork was financially supported by the World Wildlife Fund (grant FA91) and Consejo Nacional de Investigaciones Científicas y Técnicas of Argentina, granted to EHB, and a research fellowship provided by Consejo Nacional de Investigaciones Científicas y Técnicas of Argentina to OV. Several researchers (G. Claps, P. Ortiz, P. Jayat, J. C. Moretta, and M. Tome) provided substantial assistance in the identification of animal specimens. M. Beleizán provided invaluable assistance in fieldwork. J. Brasca checked the English language. We are especially grateful to F. Beleizán and L. Gil from Campo Grande for their friendship and hospitality, which made our stay in the study area comfortable and pleasant.


  • Associate Editor was Rodrigo A. Medellín.

Literature Cited

View Abstract