SciELO - Scientific Electronic Library Online

 
vol.25 número2Environmental predictors of demography in the tuco-tuco of the dunes (Ctenomys flamarioni)Variabilidad genética de visones americanos Neovison vison asilvestrados en la provincia de Santa Cruz: ¿se cumple la paradoja genética de las especies invasoras? índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

  • No hay articulos citadosCitado por SciELO

Links relacionados

Compartir


Mastozoología neotropical

versión impresa ISSN 0327-9383versión On-line ISSN 1666-0536

Mastozool. neotrop. vol.25 no.2 Mendoza dic. 2018

 

ARTÍCULO

Ant diversity in the diet of giant anteaters, Myrmecophaga tridactyla (Pilosa: Myrmecophagidae), in the Iberá Nature Reserve, Argentina

 

Nadia L. Jiménez1, 2, Yamil E. Di Blanco2, 3 and Luis A. Calcaterra1, 2

1 Fundación para el Estudio de Especies Invasivas (FuEDEI), Hurlingham, Buenos Aires, Argentina. [Correspondence: Nadia L. Jiménez <nadinelis@hotmail.com>]

2 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), CABA, Argentina.

3 Asociación Civil Centro de Investigaciones del Bosque Atlántico (CeIBA); Instituto de Biología Subtropical (IBS), nodo Puerto Iguazú, CONICET-Universidad Nacional de Misiones (UNaM); Puerto Iguazú, Misiones, Argentina.


ABSTRACT

The giant anteater, Myrmecophaga tridactyla, globally categorized as a vulnerable species, has disappeared in several regions of its original distribution in Argentina. A program to reintroduce the species has been conducted in the Iberá Nature Reserve in Corrientes province since 2006. The diet of released giant anteaters was studied to determine the identity of their prey, and establish whether they have preference for ants or termites or, rather, prefer certain feeding habitats (e. g., open or closed). Twenty two fecal samples were randomly collected during 2008-2013, and heads and mesosomes were recovered. We identified 12 taxa of ants and only one taxon of termites. Observed taxa represent around 80% of the taxa expected to be eaten by anteaters. Camponotus was the most common ant genus, and Acromyrmex and Solenopsis werethe numerically most abundant genera. The ant taxa ingested by M. tridactyla were reflective of their natural availablility in the area, suggesting that giant anteaters had no preference for any particular prey. They mainly consumed ant species of the genera Solenopsis, Camponotus and Acromyrmex with conspicuous nests that occur mostly in open habitats of the reserve and not in the most preferred habitat (forest). One possible explanation is that anteaters reduce their foraging search time, and consequently the time they are out in open habitat, so avoiding predation risk and thermal injuries. Thus, conservation of both open and closed habitats would be essential for maintaining the reintroduced populations of giant anteaters.

RESUMEN

Diversidad de hormigas en la dieta del oso hormiguero gigante, Myrmecophaga tridactyla (Pilosa: Myrmecophagidae), en la Reserva Natural Iberá, Argentina.

El oso hormiguero gigante, Myrmecophaga tridactyla, es una especie catalogada como vulnerable tanto a nivel mundial como en Argentina, donde desapareció en varias regiones de su distribución nativa. Un programa de reintroducción de especies se lleva a cabo desde 2006 en la Reserva Natural Iberá en Corrientes. Se estudió la dieta de los osos hormigueros gigantes liberados para determinar la identidad de sus presas y conocer si poseen preferencia por hormigas o termitas o por sitios de alimentación (abiertos o cerrados). Se colectaron 22 muestras de heces entre 2008-2013, recuperándose cabezas y mesosomas. Estos tagmas fueron identificados en 12 taxones de hormigas y uno de termitas. Los taxones observados representaron el 80% de lo esperado a ser ingerido por los osos. Camponotus fue el género más común, y Acromyrmex y Solenopsis los más abundantes numéricamente. Los taxones ingeridos por M. tridactyla se relacionaron positivamente con la disponibilidad del área, sugiriendo que los osos no tienen preferencia por presas en particular. Principalmente consumieron hormigas con nidos conspicuos de Solenopsis, Camponotus y Acromyrmex que se encuentran mayormente en áreas abiertas de la reserva y no en el bosque, hábitat preferido. Creemos que debido a la alta disponibilidad de esas presas, los osos reducen su tiempo de búsqueda de forrajeo, y por ende, su tiempo de permanencia en áreas abiertas, evitando así riesgos de prelación y daños térmicos. La conservación de ambos hábitats, abiertos y cerrados, sería esencial para el mantenimiento de la población reintroducida de osos hormigueros gigantes.

Key words: Conservation; Diet preference; Feces; Prey availability; Wild anteaters.

Palabras claves: Conservación; Disponibilidad de presas; Heces; Oso hormiguero salvaje; Preferencia.

Recibido 16 febrero 2018.
Aceptado 12 junio 2018.

Editor asociado: F. Bozinovic


INTRODUCTION

The giant anteater (Myrmecophaga tridactyla Linnaeus, 1758) is a large insectivorous mammal (Pilosa: Myrmecophagidae) native to Central and South America. This species is considered highly vulnerable in its homeland as a consequence of road kills, deforestation (mostly for agriculture), grassland burning, and hunting (International Union for Conservation of Nature 2013). Anteater is a common name for the four extant mammal species of the suborder Vermilingua (meaning "worm tongue”) that almost exclusively eat ants (Hymenoptera: Formicidae) and termites (Isoptera). The other three species are the silky anteater Cyclopes didactylus (Cyclopedidae) and two Myrmecophagidae: the southern collared Tamandua tetradactyla and the northern collared Tamandua mexicana.

The giant anteater is the largest extant Xenarthra, reaching up to two meters of total body length, including the tail, and weighting 40 kg in adulthood (Drumond 1992). Myrmecophaga tridactyla is a solitary species that does not show strong sexual dimorphism (Camilo-Alves 2003). Because of its low metabolic rate and body temperature (27-33°C), anteaters are very susceptible to extreme temperatures (McNab 1984). In consequence, their daily activity pattern changes from diurnal in the colder seasons to nocturnal in the warmer periods of the year (Camilo-Alves & Mourão 2006; Di Blanco et al. 2012, 2016). The giant anteater has home ranges between 2 and 25 km2 (Medri & Mourâo 2005). Their home ranges contain open areas where they spend most of the time feeding, and forested areas, closed scrubs or high grasslands that they use as anti-predatory and thermal shelter (Camilo-Alves & Mourão 2006; Mourão & Medri 2007; Di Blanco et al. 2015, 2017).

Myrmecophaga tridactyla is one of the most specialized mammalian predators that eats almost exclusively ants and termites (Redford 1985; Rodrigues et al. 2008), for which they have extremely long tongues and powerful claws adapted to destroy anthills and termite mounds. Early studies established that giant anteaters feed mainly on ants, and occasionally on termites (Montgomery 1985; Redford 1985, and references therein). Several studies have been additionally conducted on the diet of wild and captive giant anteaters in central (Cunha et al. 2015), southwestern (Medri et al. 2003) and central eastern (Shaw et al. 1987; Drummond 1992) Brazil and Colombia (Sandoval-Gómez et al. 2012). However, hardly anything is known from Argentina, the southern limit of their distribution, with exception of a recent study conducted by Gallo et al. (2017) in the Chaco ecoregion, northwestern Argentina. This work provides novel information on the presence of ants found in feces of wild giant anteaters, although relative abundances are not reported.

The wide geographical distribution of M.tridactyla ranges from southern Mexico to northern Argentina (Chebez & Cirignoli 2008). However, it is considered vulnerable in Argentina (Superina et al. 2012), where it disappeared in the provinces of Tucumán and Córdoba, and presumably also in Corrientes since around the middle of the twentieth century (Chebez & Cirignoli 2008), due to the fragmentation of its habitat by urbanization, spread of agriculture and other factors like direct mortality by bulldozer clearing, fire, fights with domestic dogs and their owners, road kills and deliberate hunting (Jiménez-Peréz 2013), a process also observed in Brazil (Diniz & Brito 2013).

A multispecies reintroduction program to restore a large ecosystem in Corrientes was initiated in the Iberá Nature Reserve (INR) in 2007 (Zamboni et al. 2017). This ambitious program started with the reintroduction of giant anteaters in the INR, where deforestation and traditional cattle management may have caused the local extinction of the species (Di Blanco 2015; Jiménez-Peréz et al. 2016). From 2007 to 2015, 47 individuals were released successfully in two locations of the INR, and it is believed that at least 28 anteaters were born there (Jiménez-Peréz et al. 2016; The Conservation Land Trust 2017). All anteaters released were monitored to assess, among other things, their habitat use, home ranges (Di Blanco 2015; Di Blanco et al. 2012, 2015, 2017) and activity patterns (Di Blanco 2015; Di Blanco et al. 2016), although nothing is known about their local diet.

Food and habitat are the most common foci in resource selection studies (Manly et al. 2002), and the distribution of food is often one of the most important factors defining spatial characteristics of a given species. Open savannas at INR showed more ant species, individuals, biomass, and functional groups of ants than other habitats (Calcaterra et al. 2010a). Giant anteaters spend most of their time feeding in open areas (Camilo-Alves & Mourão 2006; Mourão & Medri 2007). However, this habitat type was avoided by anteaters in the INR, probably because the vegetation cover had a buffering effect from extreme temperatures and the trees acted as a refuge, decreasing risk of predation (Di Blanco et al. 2015). Still, there is no clear evidence of the presence of potential predators (e. g., jaguar and cougar) in the INR during the last decades, with the only exception of a cougar that was recorded several times by camera traps in 2008 (Di Bitetti et al. 2010). However, there were no records of anteaters being either damaged or killed by the cougar after those records. In spite of this, it has been hypothesized that food availability may not be the main factor affecting habitat use of reintroduced giant anteaters (Di Blanco et al. 2015). If avoiding risk of predation and seeking protection from extreme temperatures are more important than prey availability, it would be expected that giant anteaters in the INR would consume a higher abundance of forest prey species than open savanna prey species.

The main objective of this study was to determine the relative abundance and richness of ants and termites in the diet of M. tridactyla released into the INR, and secondarily, to determine whether giant anteaters show preferences in their diet and habitat use. We aim at answering questions such as: 1) do anteaters consume taxa in proportion to their natural abundance?; 2) do they prefer ants above termites?; 3) do they select prey on the basis of their size?; and 4) do they preferentially feed in closed, forested areas or in open habitats, such as open savannas?

MATERIALS AND METHODS

Study area

The Iberá Natural Reserve (INR) is located in the center of the province of Corrientes, Argentina, and comprises 13000 km2 that protect varied landscapes, including wetlands, temporary freshwater lakes, grasslands, open and closed savannas, and hygrophilous forests (Canziani et al. 2003). The climate is subtropical with average temperatures of around 15-16 °C, and absolute minimum temperature of -2 °C in winter; summer temperatures average 27-28 °C,  with an absolute maximum temperature of 44°C. Mean annual rainfall is around 1500-1800mm (Neiff & Poi de Neiff 2006).

Collection and preservation of samples

From October 2007 to December 2013, 31 giant anteaters (17 males and 14 females) were released in El Rincón del Socorro (RS, the first reintroduction site), a private ranch that is part of the INR, next to the town of Carlos Pellegrini (Di Blanco et al. 2012, 2015; Jiménez-Pérez 2013). All anteaters released were fitted with harnesses equipped with very high frequency transmitters (Telonics, Mesa, Arizona). Released animals were of different ages (Table 1) and originated from the Argentine Chaco region (Jiménez-Pérez 2013). Animals were located by "homing in”, following the radio signal until the animal was actually seen or heard, and then followed for varying periods of time (Di Blanco et al. 2012, 2015).

Table 1
Name, sex, age and release date of anteaters released in the Iberá Natural Reserve and collection date and season of their feces. F=Female, M=Male, U=Unidentified anteater, A=Autumn, S=Summer, Sp=Spring, W=Winter.

To assess the diet of anteaters, their feces were randomly collected in 21 opportunities between January 2008 and August 2013 during animal tracking or monitoring, in habitats with high visibility or open understory grass layers in savannas and forests in RS. A total of 22 samples belonging to 10 (32%) identified (5 of each sex) and four unidentified anteaters of the 31 released giant anteaters were collected (Table 1) and preserved in 70 and 96% alcohol vials.

Sample examination

A subsample of 5 grams (dry weight) of feces was taken from each container. Samples were washed and shredded under a dissecting microscope in order to extract all recognizable cuticular structures of ants and termites (Medri et al. 2003; Sandoval-Gomez et al. 2012; Gallo et al. 2017). Once separated, structures were preserved in 70% ethyl alcohol for subsequent identification. Different identified tagmata (heads, mesosomes, and gasters) were stored in separate vials for identification to the lowest possible taxonomic level (species, genus or subfamily) using available keys (Kusnezov 1978; Palacio & Fernández 2003; Bolton 2013; and other keys and photographic material available in Antweb: https://www.antweb.org/). The number of structures (heads and mesosomes) of each taxon in each sample was counted. Samples were deposited in the Fundación para el Estudio de Especies Invasivas (FuEDEI) entomological collection.

Data analysis

Richness was estimated using EstimateS 9.1.0 software (Colwell 2013) with a presence-absence matrix. Sample-based taxon accumulation curves (genera and subfamilies for ants, and orders for termites) were used to compare density of taxa (number of taxa per sampling unit using both heads and mesosomes) as an indicator of sampling efficacy. Curves were obtained after 100 randomizations. Three nonparametric indexes (Chao 1, ACE and Bootstrap) were used to estimate the total number of taxa expected to occur in the diet of the giant anteaters.

Nonmetric multidimensional scaling (NMDS) based on a Bray-Curtis dissimilarity matrix with presence-absence of grouped head and mesosome data were used to compare similarity patterns in ant and termite taxa consumed by M. tridactyla per season (22 samples) and sex (18 samples). The ordinations were performed with taxa that occurred in four or more samples. Thus, only eight of the 13 taxa (Acromyrmex, Atta, Camponotus, Crematogaster, Dolichoderinae, Pheidole, Solenopsis, Isoptera) were used in both analyses. These analyses were tested statistically using an analysis of similarity (ANOSIM and post hoc Bonferroni pairwise comparisons; Clarke & Green 1988) based on 1000 permutations in the Past 3.16 software (Hammer et al. 2001). The Sörensen similarity index was also used to compare the similarity in taxon composition between seasons and sexes with the same matrix, using in NMDS. As there was a large unbalance in the number of samples per year, analysis and comparisons between years were not carried out.

Abundance of each taxon in the diet of anteaters was calculated on the basis of the two most informative types of structures (heads and mesosomes). The mean number of structures and taxa found in each season and sex was compared using one-way analysis of variances (ANOVA) with InfoStat software (Di Rienzo et al. 2015). Spearman’s correlation between the number of heads and mesosomes was also calculated, in order to determine if both structures are similarly recovered using Past 3.16 software (Hammer et al. 2001).

The presence and relative abundance of ant taxa found in the feces of M. tridactyla were related to the presence and relative abundance of the ant species previously found in the INR by Calcaterra et al. (2010b, 2014) at a large (~14000 ha) and small (~500ha) spatial scales, respectively. On a large scale, ants were sampled in (grazed and non-grazed) savannas and grasslands, while on a small scale, sampling was carried out in (burned and unburned) grasslands and open and closed savannas, plus (unburned) forests. However, only ants found in undisturbed (non-grazed and unburned) habitats were used in the analysis. The small scale corresponds to the area where the first anteater individuals were released and established (Di Blanco et al. 2012, 2015, 2016). In both cases, samplings were conducted using five unbaited pitfall traps every 10 meters along one transect in each site; traps were exposed for 48 hours (Calcaterra et al. 2010b, 2014). A lineal regression was calculated using InfoStat software (Di Rienzo et al. 2015) between the relative abundance of ants obtained in the feces and the overall availability of ants found in each one of these two spatial scales (grouping ants from all habitats), that is, without discriminating by habitat type usually used by anteaters (Di Blanco et al. 2012, 2015, 2017) because it is impossible to know in which habitat/s anteaters ate before defecating.

RESULTS

Taxon richness

A total of 13 taxa were identified from the feces of giant anteaters; 12 taxa of ants (species, genera or subfamily) and one taxon of termites (subfamily). Accumulation curves of estimated taxa did not reach the asymptote, indicating that more taxa are expected to occur in the diet of the anteaters (Fig. 1). According to two nonparametric richness estimators (Chao 1 and ACE) and an extrapolation of 200 samples, an average of 18.3 taxa was expected to occur, whereas, using a Bootstrap estimator, only 14.5 taxa were expected. Thus, the number of taxa observed in this study (13) represents between 71.0 and 89.7% of the total number of taxa expected to occur in the diet of anteaters in the study area.


Fig. 1
. Taxon accumulation curves from collected heads and mesosomes from feces (all samples pooled) of released giant anteaters in the Iberá Nature Reserve, and the three nonparametric indexes with the most stable asymptote.

By using heads for identification, a higher number of taxa was found (13) than when using mesosomes (7). The mean (ñ SD) richness of heads per sample was 5.1ñ1.6 taxa (range: 2-8 genera per sample); there were no significant differences between seasons (F3, 18=1.71, P=0.201) or sexes (F1, 16=0.08, P=0.775). The mean (ñ SD) richness of mesosomes was 4.0ñ1.3 genera per sample (ranging from 1 to 6). There was no difference in taxa per sample between sexes (F1, 16=2.87, P=0.110), however, there was a marginally significant difference between seasons (F3, 18=2.98, P=0.059), but it was only attributed to differences between autumn and summer.

All ant diversity indicators (grouping head and mesosome data) were relatively similar in all seasons, except in autumn (Table 2), and between sexes (Table 3). Consequently, the mean number of taxa per sample did not differ significantly between seasons (F3, 18=1.39, P=0.277) or sexes (F1, 16=1.36, P=0.261). It is interesting to note that a single sample can contain more than a half of the total taxa found (up to 61.5%).

Taxon composition

Table 2
Diversity indicators based on taxa found per season grouping heads and mesosomes.

Table 3
Diversity indicators based on taxa found per sex, grouping heads and mesosomes.

The visual ordination of the samples did not show separation in the ingested taxa composition between seasons (NMDS: R2=0.766, stress=0.24; Fig. 2a; ANOSIM, R=0.047, P=0.280) or between sexes (NMDS: R2=0.826, stress=0.23; Fig. 2b; ANOSIM, R=-0.007, P=0.470). The overlapping between seasons was mostly due to the high similarity between winter and the other three seasons (summer, spring and autumn with Sörensen indexes of 0.73, 0.91 and 0.75, respectively). The lowest similarities were observed between summer, spring and autumn (0.67 for all combinations). The highest similarity was observed in the taxa consumed by both sexes (Sörensen index equal to 1).


Fig. 2
. Nonmetric multidimensional scaling ordination plot from presence of heads and mesosomes of taxa found in feces samples of giant anteaters in (a) different seasons (R2=0.766, stress=0.24) and (b) sexes (R2=0.826, stress=0.23). The ellipses represent 95% confidence. Note that autumn does not have a confidence ellipse because with two points the program is not capable of calculating it.

Most of the heads and mesosomes found belong to the genera Camponotus (such as C.rufipes and C.punctulatus), Crematogaster, Pheidole (e. g., P. aberrans), and Solenopsis (mostly S.invicta). Few heads of Dolichoderinae (presumably Dorymyrmex or Linepithema), Gnamptogenys, Odontomachus, Pseudomyrmex, Trachymyrmex, Wasmannia (only W.auropunctata) were found. Leafcutter ants, mainly from the Acromyrmex genus (e.g., A. heyeri, A. hispidus), and fewer still of the genus Atta (e. g., A. sexdens, A. vollenweideri) were recorded. Many wings and several heads with ocelli (e. g., Camponotus), typically from sexual caste and some sclerites from gasters and legs of ants were also found. Unexpectedly, only a few nasutti morph (Termitidae: Nasutitermitinae) termite heads were recovered. All the Nasutitermes species can be recognized by their soldiers that have a pointed snout at the front of their heads (called nasus). Some mites (Acari) and a Coleoptera head were collected, but they are not considered an item eaten by anteaters.

Abundance

In a few samples, 22 whole ants were found in perfect condition and were easily determined to species: 16 workers belonged to S. invicta and 6 workers to A. hispidus.

Heads. A total of 11965 heads was obtained from giant anteater feces: 11911 of ant heads (99.55%) and only 54 (0.45%) termite heads; 11627 (97.6%) ant heads were assigned to 11 genera, while 61 (0.5%) could only be assigned to the Dolichoderinae subfamily (probably Dorymyrmex or Linepithema genera). The remaining 223 heads (1.9%) could not be identified. The most common genus was Camponotus, present in 100% of the samples, followed by Acromyrmex and Solenopsis found in 95.5% and 90.9% of the samples, respectively. The most abundant genus was Solenopsis which comprised 41% of the total amount of heads, followed by Acromyrmex with 31.9% (Table 4, Fig. 3). The number of heads per sample did not differ between seasons (F3, 18=0.65, P=0.590) or sexes (F1, 16=0.25, P=0.624).

Table 4
Number (percent of samples) of heads and mesosomes of ants and termites recovered from 22 samples of 5 grams of feces of released anteaters in Iberá Nature Reserve. Ant availability was obtained from the literature at two spatial scales: large scale (~14000 ha in Calcaterra et al. 2010b) and small scale (~500 ha in Calcaterra et. al. 2014).


Fig. 3
. Relative abundance of identified heads (black bars) and mesosomes (grey bars) of ants and termites from feces of giant anteaters.

Mesosomes. A total of 9216 ant mesosomes were recovered from giant anteater feces, which could be assigned to 7 genera of ants. Again, Camponotus was the most common genus, present in 100% of the samples, followed by Acromyrmex (90.9%) (Table 4). Solenopsis and Acromyrmex were the most abundant genera: 42.2% and 34.8% of the total number of mesosomes, respectively (Fig. 3). No termite mesosome was recovered from feces of giant anteaters. The number of mesosomes per sample did not differ between seasons (F3, 18=0.90, P=0.462) or sexes (F1, 16=0.87, P=0.365).

Diet preference

As the numbers of heads and mesosomes of each taxon (ants and termites) were strongly correlated (R2=0.915, P<0.001), taxon preference was based only on heads because they provide greater resolution. The number of heads of each taxon found in the diet of M. tridactyla was related to the availability analyzed on different spatial scales (large and small) in the INR (Calcaterra et al. 2010b, 2014, respectively) (Table 4). On a large scale, the relationship was low and non-significant (R2=0.25, F1, 16=2.39, P=0.166). Nevertheless, the relationship was higher and significant on a small scale (R2=0.42, F1, 20=6.52, P=0.031).

DISCUSSION

Diet of giant anteaters reintroduced in the INR was mainly composed of ants, and very secondarily, of termites. This agrees with the diet of wild giant anteaters observed by Redford (1985) and Shaw et al. (1987) in the Brazilian states of Goias and Minas Gerais, respectively, and recently by Gallo et al. (2017) in the Chaco ecoregion in Argentina. The latter found 22 morphospecies in 14 ant genera and indicated that termites were recovered in fewer numbers than ants. In spite of the low number of taxa and individuals of termites recovered from the feces of anteaters, they appeared in the half of the samplescollected in the INR.

Medri et al. (2003) observed that ants were consumed on a higher proportion than termites (81 against 19%) only in June (winter). However, winter was the season with more termites in the diet of M. tridactyla in the INR. It could be due to the lower availability of ants recorded in winter in the INR (Calcaterra et al. 2014) that would force the anteaters to consume more termites, as they do not decrease their abundance as ants do. This difference could also be mostly because of latitudinal variations in prey availability. In other cases, differences could be local or longitudinal. For example, Shaw et al. (1987) found in Serra da Canastra National Park in the Minas Gerais state (Brazil), a 9:1 ant:termite ratio, whereas Drummond (1992) found, in the same place, a 1:1 ratio (local variations). Meanwhile, Redford (1985) got the same relationship as Shaw et al. (1987) in the Emas National Park, in Goias state, in Brazil (longitudinal differences); whereas, Cunha et al. (2015) found slightly more termites (17%) than ants (11%) in the stomach content of a road-killed giant anteater in the Goias state (Brazil).

Regarding the natural diversity of ants and termites in Argentina, a total of 661 ant species in 71 genera, and 7 subfamilies, have been reported from all the biomes (Cuezzo 1998), whereas only around 80 termite species in 4 subfamilies have been found in 7 phytogeographical provinces: Yungas (12 spp.), Chaqueña (78 spp.), Paranaense (41 spp.), Espinal (17 spp.), Pampeana (7 spp.), Monte (7spp.), and Subantárctica (1 spp.) (Torales et al. 2005, 2009). These data indicate much higher ant than termite diversity, as observed in the diet of the anteaters in the Chaco ecoregion (Gallo et al. 2017). The same pattern has been reported for the INR, where more than 100 ant species and only a dozen of termite species occur (Calcaterra et al. 2010a; b, 2014; Jiménez-Peréz 2013). An alternative explanation for the discrepancy in the quantity of ants and termites recovered from feces of M. tridactyla in the INR and other sites could be that the cuticle of the termites is more labile (fragile) than the cuticle of the ants, and thus, it could be digested more easily by the gastric acid in the stomach of the anteaters. This supposition arose from observing that the nasuti heads found in the feces were mostly in very bad condition, sometimes only differentiated by the frontal tube. Nasutitermes have strictly chemical defense mechanisms and their soldiers have vestigial mandibles and a frontal tube which is their only weapon against predators (Scholtz et al. 2008).

Among the ants, workers of the Solenopsis genus were the main component (41%) of the diet of giant anteaters in the INR. The following most important ant genera were Acromyrmex (32%) and Camponotus (17%). The latter was the only genus present in all samples in the INR and it was the richest genus with five morphospecies as shown in the work of Gallo et al. (2017). These three genera comprised around 90% of the total heads counted and were also recovered from feces of M. tridactyla in Brazil (Medri et al. 2003). Leaf-cutting ants from the Atta genus (e. g., A. sexdens, A. vollenweideri) are represented in low numbers in the samples despite being present in the INR (Calcaterra et al. 2010a; b, 2014). These ants seem to be highly preferred by M. tridactyla in Colombia (Sandoval-Gómez et al. 2012) and denote a very important food source in the tropics, where they are more common, abundant and have larger colonies than the Acromyrmex species.

The diet composition of M. tridactyla in the INR indicates that it consumes mostly terrestrial ant species with conspicuous nests containing from hundreds of big workers(e. g., C. punctulatus, C. rufipes, A. heyeri, A.hispidus) to several thousand small workers (e. g., S.invicta). A colony of S. invicta can contain up to 250000 workers (Tschinkel 2006). These ant species nest mostly in open habitats, such as grasslands or open savannas with grass predominance (Calcaterra et al. 2010a; b, 2014), where M. tridactyla searches its food (Silvestre et al. 2003; Sandoval Gómez et al. 2012; Di Blanco et al. 2015). Because of their conspicuous nests containing a large quantity of ant biomass easy to be consumed, giant anteaters can ingest a large quantity of biomass in a short time (Fernández 2003; Miranda et al. 2009). A similar selection was observed in the yellow armadillo, Euphractus sexcinctus, in abandoned rice fields with a high density of nests of C. punctulatus in the INR (Calcaterra et al. 2010a). This foraging behavior contrasts with that observed in the most forested Chaco ecoregion, where giant anteaters presumably spent more time searching than eating the ants present mostly in the leaf litter (Gallo et al. 2017). There, the most consumed ants were the army ants of the genera Eciton or Labidus, followed by above-ground ants and some few arboreal ants. This also suggests the high plasticity of giant anteaters in terms of their diet.

Redford (1985) postulated that M. tridactyla did not eat according to prey size or nutritional quality in captivity, but it did in the wild. They spent more time on nests with winged individuals (larger ones and present only in the warm season), which is probably related to their high nutritional value. This difference was attributed to the foraging behavior of giant anteaters that is different in captivity, where they simply have to take the prey, while in the wild they have to search and capture them. The choice of preys depends on their nutritional value, availability and response to attack. In this work, many winged individuals (sexual), a few wings, and several heads with ocelli, typically from this caste, were found. Similarly, winged termites represent a bigger source of food than worker caste to other vertebrates, like birds (Eisenmann 1961).

The size of ant workers consumed by giant anteaters in the INR was variable; they consumed both big Camponotus and Acromyrmex workers, and smaller Solenopsis workers. However, although colonies of Solenopsis species (e.g., S. invicta) have many more workers than colonies of bigger carpenter and leaf-cutting ants (e. g., C. punctulatus, C. rufipes, A. heyeri, A. hispidus), the average weight of S. invicta workers (0.15 mg) is around sixteen times lower than the average weight of grouped workers of the biggest Camponotus and Acromyrmex species (~2.5 mg) (Calcaterra et al. 2010a). Thus, the quantity of biomass of Acromyrmex consumed by giant anteaters was much higher than that of S. invicta.

The periods of activity of M. tridactyla are also important as different ant species have different annual and daily foraging patterns, and this may influence their diet. Medri et al. (2003) recorded only a few ant and termite nests attacked by giant anteaters in Brazil between April and October, the coldest half of the year, when ant foraging activity is overall lower. They ate nine ant species, of which individuals of Solenopsis (probably conspicuous fire ant nests) were the most frequently attacked (82% of total attacked nests) and only two termite species were attacked in June (winter). In the warmer season, nests are bigger and contain a higher number of workers (sexual and brood). However, since giant anteaters mainly fed directly from the nests when breaking them (Medri et al. 2003), maybe their diet was not so much based on whether they found foraging ants than the size or age of the colony. On the other hand, there was a strong overlapping of most seasons, mainly because of the higher similarity between winter and the other seasons, including autumn although only two samples were taken in this season, whereas between 5 and 8 samples were taken in each of the other seasons. Nevertheless, reintroduced giant anteaters presented seasonal variations in habitat selection, using grasslands more frequently in winter (Di Blanco 2015). This would probably be more related to seasonal variability in ant biomass availability (Calcaterra et al. 2014) than to differences in ant composition. A more balanced design among seasons (and years) could confirm if M. tridactyla effectively showed seasonal variations in their diet in the INR. No difference was found between sexes in the composition of taxa (or abundance of tagma) ingested by anteaters.

The spatial scale of studies is important to understand biological process, such as food resource preferences. In our case, the relationship (available/ingested ants) was higher and significant on a smaller scale that corresponds to the area where the first anteaters were released in INR (Calcaterra et al. 2014). Thus, it suggests that anteaters could be consuming their prey according to their relative abundance on this scale (~500 ha), which is similar to their usual home range (Di Blanco 2015), or even closer to the individual home range in the INR (an average of 2100 ha between males and females, Di Blanco et al. 2017) than on a large scale (~14000 ha) that is closer to the range used for all grouped individuals of the first reintroduced population (~12000 ha) (Di Blanco et al. 2015). We believe that the loss of relationship on a large scale could be due mostly to a poorer spatial match between the feeding sites (from where ants were incorporated to the feces) and ant collection sites (Calcaterra et al. 2010b).

This study revealed that it was possible to recover most taxa expected to be present (80%) in the diet of M. tridactyla in the INR. The most recognizable structures were heads and mesosomes and followed by sclerites from gasters, legs of ants and several termite heads. However, legs and sclerites of gasters could not be associated with any kind of taxa. Although it was difficult to identify ants recuperated from the feces of giant anteaters at taxonomic levels lower than genus (e. g., Gallo et al. 2017), this methodology seems at least to recover most of the ants ingested by the dwarf armadillo, Zaedyus pichiy (Cingulata: Dasypodidae) (Superina & Elizalde 2011). In this study, the number of workers of leaf-cutting ants (Acromyrmex lobicornis) recovered from feces was around 87% of the total number of ants effectively consumed in the steppe habitat in Rio Negro province, in the Argentine Patagonia. These studies illustrate that feces examination can be an appropriate method for diet estimations based mainly on heads of each taxon. Curiously, other methodologies to study anteater diets that presumably preserved heads and mesosomes better, such as stomach content of anteaters, have been less successful to recover taxa; such is the case of a road-killed giant anteater in central Brazil, in which 71% of the stomach content could not be identified, or even differentiated between ants and termites (Cunha et al. 2015).

In summary, reintroduced giant anteaters mainly consumed terrestrial ants, as in most of the previous studies. This species, despite their specialized diet and the geographic origin of reintroduced individuals, showed high plasticity in terms of prey consumption, eaten in proportion to their abundance in nature. Thus, the higher biomass was mostly composed of the most commonly distributed and numerically abundant species (e. g., C.punctulatus, C.rufipes, A. heyeri, S. invicta) that occur mostly in open habitats, such as grasslands and open savannas of the INR. Our results also supported previous assumptions that preys consumed do not reflect habitat selection patterns found for this reintroduced population. Therefore, food resource distribution would be a poor predictor of habitat selection for giant anteaters in this study site, where open habitats are more available than forested areas. A possible explanation could be that high prey availability allows anteaters to limit foraging search time and the time spent in open habitats decreasing the risk of predation and thermal damage. Giant anteaters generally benefit from habitat heterogeneity, where they can use vegetation cover for protection and open habitats for foraging (Cardoso Da Silva & Bates 2002; Prada & Marinho-Filho 2004; Desbiez & Medri 2010; Vynne et al. 2011; Quiroga et al. 2017).

Maintenance of this habitat mosaic of open and closed habitats seems to be essential for the maintaining of self-suitable populations of giant anteaters restored in the INR. The conserving of a greater surface of these habitats, some of which are currently being threatened by the advance of the agriculture frontier, as the savannas, is also crucial for increasing the size of their populations, which will guarantee the future survival of this species within and outside of this protected area.

ACKNOWLEDGEMENTS

We are grateful for the assistance and support provided by the staff of Conservation Land Trust Argentina in the Rincón del Socorro ranch; especially Karina Spørring and Alicia Delgado for facilitating feces samples of anteaters for this study. We also thank Mario Di Bitetti, Talia Zamboni and Arabella Peard for their comments, suggestions and corrections that help to improve the manuscript. This research was partially funded by the Conservation Land Trust Argentina (CLT), the Fundación para el Estudio de Especies Invasivas (FuEDEI), and the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET).

LITERATURE CITED

1. Bolton, B. 2013. Ants of Bolton World Catalog. AntWeb. <http://www.antweb.org>         [ Links ].

2. Calcaterra, L. A., F. Cuezzo, S. M. Cabrera, & J. A. Briano. 2010a. Ground ant diversity (Hymenoptera: Formicidae) in the Iberá Nature Reserve, the largest wetland of Argentina. Annals of the Entomological Society of America 103:71-83.         [ Links ]

3. Calcaterra, L. A., S. M. Cabrera, F. Cuezzo, I.Jiménez Peréz, & J. A. Briano. 2010b. Habitat and grazing influence on terrestrial ants in subtropical grasslands and savannas of Argentina. Annals of the Entomological Society of America 103:635-646.         [ Links ]

4. Calcaterra, L. A., Y. Di Blanco, M. Srur, & J. A. Briano. 2014. Fire effect on ground-foraging ant assemblages in northeastern Argentina. Journal of Insect Conservation, 18:339-352.         [ Links ]

5. Camilo-Alves, C. 2003. Adaptações dos tamanduás-bandeira à variação da temperatura ambiente no Pantanal da Nhecolândia, MS. Tesis de Maestría. Universidad Federal de Mato Grosso do Sul, Mato Grosso do Sul, Brazil.         [ Links ]

6. Camilo-Alves, C., & G. Mourão. 2006. Responses of a specialized insectivorous mammal (Myrmecophaga tridactyla) to variation in ambient temperature. Biotropica 38:52-56.         [ Links ]

7. Canziani, G., C. Rossi, S. Loiselle, & R. Ferrati. 2003. Los esteros del Iberá. El manejo sustentable de humedales en el Mercosur. Fundación Vida Silvestre Argentina, Buenos Aires.         [ Links ]

8. Cardoso Da Silva, J. M., & J. M. Bates. 2002. Biogeographic patterns and conservation in the South American Cerrado: a tropical savanna hotspot: the Cerrado, which includes both forest and savanna habitats, is the second largest South American biome, and among the most threatened on the continent. AIBS Bulletin 52:225-234.         [ Links ]

9. Chebez, J. C., & S. Cirignoli. 2008. Los que se van: fauna argentina amenazada: Mamíferos. Yurumí (J.C. Chebez, ed.). Albatros, Buenos Aires.         [ Links ]

10. Clarke, K. R., & R. H. Green. 1988. Statistical design and analysis for a biological effects study. Marine Ecology Progress Series 46:213-226.         [ Links ]

11. Colwell, R. K. 2013. EstimateS: Statistical estimation of species richness and shared species from samples. Version 9. <http://www.purl.oclc.org/estimates>         [ Links ].

12. Cuezzo, F. 1998. Formicidae. Biodiversidad de artrópodos Argentinos: una perspectiva biotaxonómica. Ediciones Sur, La Plata.         [ Links ]

13. Cunha, H. F. et al. 2015. Food preference of giant anteater and collared anteater (Pilosa: Myrmecophagidae) regarding the termite defense strategies. Bioscience Journal 31:234-241.         [ Links ]

14. Desbiez, A. L. J., & Í. M. Medri (2010). Density and habitat use by giant anteaters (Myrmecophaga tridactyla) and southern tamanduas (Tamandua tetradactyla) in the Pantanal wetland, Brazil. Edentata 11:4-10.         [ Links ]

15. Di Bitetti, M. S., C. D. De Angelo, Y. E. Di Blanco, & A. Paviolo. 2010. Niche partitioning and species coexistence in a Neotropical felid assemblage. Acta Oecologica. 36:403-412.         [ Links ]

16. Di Blanco, Y. E., I. Jiménez Pérez, & K. Spørring. 2012. Cinco años de radiomarcaje de osos hormigueros (Myrmecophaga tridactyla): mejoras implementadas y lecciones aprendidas. Edentata 13:49-55.         [ Links ]

17. Di Blanco, Y. E. 2015. Patrones de actividad y uso de hábitat de osos hormigueros (Myrmecophaga tridactyla) reintroducidos en Iberá, Corrientes, Argentina. Tesis de Doctorado. Universidad Nacional de Córdoba, Córdoba, Argentina.         [ Links ]

18. Di Blanco, Y. E., I. Jiménez Pérez, & M. S. Di Bitetti. 2015. Habitat selection in reintroduced giant anteaters: the critical role of conservation areas. Journal of Mammalogy 96:1024-1035.         [ Links ]

19. Di Blanco, Y. E., K. S. Spørring, & M. S. Di Bitetti. 2016. Daily activity pattern of reintroduced giant anteaters (Myrmecophaga tridactyla): effects of seasonality and experience. Mammalia 81:11-21.         [ Links ]

20. Di Blanco, Y. E. et al. 2017. Habitat selection and home range use by resident and reintroduced giant anteaters in two South American wetlands. Journal of Mammalogy 98:1118-1128.         [ Links ]

21. Di Rienzo, J. A., F. Casanoves, M. G. Balzarini, L.González, M. Tablada, & C.W. Robledo. 2015. Infostat. Grupo InfoStat, Universidad Nacional de Córdoba, Córdoba, Argentina. <http://www.infostat.com.ar>         [ Links ].

22. Diniz, M. F., & D. Brito. 2013. Threats to and viability of the giant anteater, Myrmecophaga tridactyla (Pilosa: Myrmecophagidae), in a protected Cerrado remnant encroached by urban expansion in central Brazil. Zoología 30:151-156.         [ Links ]

23. Drumond, M. A. 1992. Padrões de forrageamento do tamanduá-bandeira (Myrmecophaga tridactyla) no Parque Nacional da Serra da Canastra: dieta, comportamento alimentar e efeito de queimadas. Tesis de Maestría. Instituto de Ciencias Biológicas, Belo Horizonte, Brazil.         [ Links ]

24. Eisenmann, E. 1961. Favorite foods of Neotropical birds: flying termites and Cecropia catkins. The Auk 78:636-638.         [ Links ]

25. Fernández, F. (Ed.). 2003. Introducción a las hormigas de la región Neotropical. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt, Bogotá         [ Links ].

26. Gallo, J., A. Abba, L. Elizalde, D. Nucci, & M.Superina. 2017. First study on food habits of anteaters, Myrmecophaga tridactyla and Tamandua tetradactyla, at the southern limit of their distribution. Mammalia 81:601-604.         [ Links ]

27. Hammer, O., D. A. T. Harper, & P. D. Ryan. 2001. PAST: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4:9.         [ Links ]

28. International Union for Conservation of Nature. Red List of Threatened Species. Version 2013. <http://www.iucnredlist.org>         [ Links ].

29. Jiménez Pérez, I. (Ed.). 2013. Lessons learnt, experience gained. Giant anteater: a homecoming to Corrientes. The Conservation Land Trust, Buenos Aires.         [ Links ]

30. Jiménez-Pérez, I. et al. 2016. Re-introduction of the giant anteater in Iberá Nature Reserve, Corrientes, Argentina; Global Re-introduction Perspectives: 2016. Case-studies from around the globe (P. S. Soorae, ed.). Gland.         [ Links ]

31. Kusnezov, N. 1978. Hormigas argentinas: claves para su identificación. Miscelánea 61.         [ Links ]

32. Manly, B. F. J., L. L. Mcdonald, D. L. Thomas, T.L. Mcdonald, & W. P. Erickson. 2002. Resource selection by animals: statistical analysis and design for field studies, 2nd edition. Kluwer, Boston.         [ Links ]

33. Mcnab, B. K. 1984. Physiological convergence amongst ant‐eating and termite‐eating mammals. Journal of Zoology 203:485-510.         [ Links ]

34. Medri, I. M., & G. M. Mourâo. 2005. Home range of giant anteaters (Myrmecophaga tridactyla) in the Pantanal wetland, Pantanal, Brazil. Journal Zoology London 266:365-375.         [ Links ]

35. Medri, I. M., G. Mourão, & A. Y. Harada. 2003. Dieta do tamanduá-bandeira (Myrmecophaga tridactyla) no Pantanal da Nhecolândia, Brasil. Edentata 5:29-34.         [ Links ]

36. Miranda, F., R. Veloso, M. Superina, & F. J. Zara. 2009. Food habits of wild silky anteaters (Cyclopes didactylus) of São Luis do Maranhão, Brazil. Edentata 8:1-5.         [ Links ]

37. Montgomery, G. G. (Ed.).1985. Movements, foraging and food habits of the four extant species of neotropical vermilinguas (Mammalia: Myrmecophagidae). The evolution and ecology of armadillos, slots, and vermilinguas. Smithsonian Institution Press, Washington and London.         [ Links ]

38. Mourão, G., & I. G. Medri. 2007. Activity of a specialized insectivorous mammal (Myrmecophaga tridactyla) in the Pantanal of Brazil. Journal of Zoology 27:187-192.         [ Links ]

39. Neiff, J., & A. Poi de Neiff. 2006. Situación ambiental en la ecorregión Iberá. La situación ambiental argentina 2005. Fundación Vida Silvestre Argentina, Buenos Aires.         [ Links ]

40. Palacio, E. E., & F. Fernández. 2003. Claves para las subfamilias y géneros. Introducción a las hormigas de la región Neotropical (F. Fernández ed.). Instituto de Investigación de Recursos Biológicos Alexander von Humboldt. Bogotá         [ Links ].

41. Prada, M. & J. Marinho‐Filho. 2004. Effects of fire on the abundance of Xenarthrans in Mato Grosso, Brazil. Austral Ecology 29:568-573.         [ Links ]

42. Quiroga, V. A., Y. E. Di Blanco, A. Noss, A. J. Paviolo, & M. S. Di Bitetti. 2017. The giant armadillo (Priodontes maximus) in the Argentine Chaco. Mastozoología Neotropical 24:163-175.         [ Links ]

43. Redford, K. H. 1985. Feeding and food preference in captive and wild giant anteaters (Myrmecophaga tridactyla). Journal of Zoology of London 205:559-572.         [ Links ]

44. Rodrigues, F. H. G., Í. M. Medri, G. H. B. Miranda, C. Camilo-Alves, & G. Mourão. 2008. Anteater behavior and ecology. The biology of the Xenarthra (S. F. Vizcaíno & W. J. Loughry, eds.). University Press of Florida, Gainesville.         [ Links ]

45. Sandoval-Gomez, V. E., H. E. Ramírez-Chaves, & D. Marín. 2012. Registros de hormigas y termitas presentes en la dieta de osos hormigueros (Mammalia: Myrmecophagidae) en tres localidades de Colombia. Edentata 13:1-9.         [ Links ]

46. Scholtz, O. I., N. Macleod, & P. Eggleton. 2008. Termite soldier defense strategies: a reassessment of Prestwich’s classification and an examination of the evolution of defense morphology using extended eigenshape analyses of head morphology. Zoological Journal of the Linnean Society Londres 153:631-650.

47. Shaw, J. H., J. Machado-Neto, & T. S. Carter. 1987. Behavior of free-living giant anteaters (Myrmecophaga tridactyla). Biotropica 19:255-259.         [ Links ]

48. Silvestre, R., C. R. F. Brandão, & R. Rosa da Silva. 2003. Grupos funcionales de hormigas: el caso de los gremios del Cerrado. Introducción a las hormigas de la región Neotropical (F. Fernández, ed.). Instituto de Investigación de Recursos Biológicos Alexander von Humboldt, Bogotá         [ Links ].

49. Superina, M., & L. Elizalde. 2011. Estimación de la tasa de hormigas Acromyrmex lobicornis por piches Zaedyus pichiy (Mammalia: Dasypodidae): aportes preliminares. XXIV Jornadas Argentinas de Mastozoología, La Plata, Buenos Aires.< http://www.fcnym.unlp.edu.ar/uploads/docs/rmlp_zoo_2012_t18_n172.pdf>         [ Links ]

50. Superina, M. A., M. Abba, & S. F. Vizcaíno. 2012. Orden Pilosa. Libro rojo de los mamíferos amenazados de la Argentina (R. A. Ojeda, V. Chillo & G. Díaz Isenrath, eds.). SAREM, Mendoza.         [ Links ]

51. The Conservation Land Trust. 2017. Proyecto Iberá: oso hormiguero gigante. <http://www.proyectoibera.org/especiesamenazadas_osohormiguero.htm>         [ Links ]

52. Torales, G. J. et al. 2005. Composición faunística y distribución de Isoptera (Insecta) del Litoral. Miscelánea 14:259-280.         [ Links ]

53. Torales, G. J., J. M. Coronel, E. R. Laffont, J. L. Fontana, & M. C. Godoy. 2009. Termite associations. (Insecta, Isoptera) in natural or semi-natural plant communities in Argentina. Sociobiology 53:1-55.         [ Links ]

54. Tschinkel, W. R. 2006. The fire ants. Harvard University Press, Belknap.         [ Links ]

55. Vynne, C., et al. 2011. Resource selection and its implications for wide-ranging mammals of the Brazilian Cerrado. PLoS One 6:e28939.         [ Links ]

56. Zamboni, T., S. Di Martino, & I. Jiménez-Pérez. 2017. A review of a multispecies reintroduction to restore a large ecosystem: The Iberá Rewilding Program (Argentina). Perspectives in Ecology and Conservation 15:248-256.         [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons