Habitat use by colonial tuco-tucos (Ctenomys sociabilis): specialization, variation,

and sociality
Author(s): Mauro N. Tammone , Eileen A. Lacey , and Maria A. Relva
Source: Journal of Mammalogy, 93(6):1409-1419. 2012.
Published By: American Society of Mammalogists
DOI: http://dx.doi.org/10.1644/11-MAMM-A-266.1
URL: http://www.bioone.org/doi/full/10.1644/11-MAMM-A-266.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.
Journal of Mammalogy, 93(6):1409–1419, 2012

Habitat use by colonial tuco-tucos (Ctenomys sociabilis): specialization,

variation, and sociality
MAURO N. TAMMONE*, EILEEN A. LACEY, AND MARIA A. RELVA
Laboratorio Ecotono, CRUB, Universidad Nacional del Comahue, INIBIOMA-CONICET, Bariloche, Argentina (MNT, MAR).
Museum of Vertebrate Zoology, University of California, Berkeley, CA 94720 USA (EAL).
* Corresponding author: mtammone@gmail.com

Understanding habitat requirements has implications for numerous aspects of a species’ biology, including where
individuals live and how they behave. Specialization for mesic, resource-rich habitats known as mallines is
thought to have favored group living in the colonial tuco-tuco (Ctenomys sociabilis), a subterranean rodent that is
endemic to Neuquén Province in southwestern Argentina. To explore the proposed relationship between mallines
and sociality in this species in greater detail, we characterized the habitats occupied by C. sociabilis at 3 locations
representing the extremes of this species’ geographic range. Specifically, plant composition and vegetative
structure were characterized for 57 occupied burrow systems distributed across the 3 sampling localities. Our
data indicate that C. sociabilis is not restricted to mallines. Although significant variation in vegetation was
detected among the 3 study sites, the majority of active burrow systems surveyed at each site occurred in
nonmallı́n habitats. In addition to providing the first species-wide survey of habitat use by C. sociabilis, our data
yield new insights into the role of habitat specialization in promoting sociality in this behaviorally unusual
species of ctenomyid rodent.

Key words: Ctenomys sociabilis, habitat, mallı́n, sociality, subterranean rodents

Ó 2012 American Society of Mammalogists
DOI: 10.1644/11-MAMM-A-266.1

Knowledge of the habitats in which a species occurs can (1985) noted that the animals were associated with wet
provide valuable insights into its biology, including aspects of meadows known locally as ‘‘mallines.’’ Subsequently, using
its ecology, behavior, and conservation. With regard to social comparative data from C. sociabilis and the syntopic but
behavior, the distribution of critical resources—including solitary C. haigi, Lacey and Wieczorek (2003) postulated that
suitable habitats—is thought to be an important factor favoring the patchy distribution of mallı́n habitat favors philopatry and
group living in numerous species (Emlen 1982; Koenig et al. group living in the former species.
1992). Specifically, if resources such as suitable habitat are rare Although intriguing, this analysis was based upon data from
or patchily distributed, the costs of dispersing may be high a limited portion of the geographic distribution of C. sociabilis.
enough that individuals benefit by being philopatric and Mallines are found throughout the range of this species but the
remaining in their natal group (Emlen 1982, 1991). As a relationship between mallı́n habitats and the occurrence of
result, for social species, understanding patterns of habitat use colonial tuco-tucos has not been assessed at this scale. Thus,
may yield critical insights into the reasons for group living. the primary objective of this study was to determine whether C.
Specialization for mesic habitat patches has been suggested sociabilis is associated with mallı́n habitats throughout its
to be an important ecological factor favoring group living in range. Accordingly, we quantified the plant species and
the colonial tuco-tuco (Ctenomys sociabilis). Although more vegetative structure associated with currently occupied burrow
than 50 species of tuco-tucos (Rodentia: Ctenomyidae) occur systems of colonial tuco-tucos at 3 localities representing the
in sub-Amazonian South America (Reig et al. 1990; Wilson extremes of this species’ geographic distribution. In addition to
and Reeder 2005), C. sociabilis is the only ctenomyid for providing the first general characterization of the habitats used
which quantitative evidence of group living is available (Lacey by this species, our data yield potentially important insights
et al. 1997; Lacey and Wieczorek 2004). This species is
endemic to Neuquen Province, Argentina, where it occupies
open habitats in the eastern foothills of the Andes mountains.
In their initial description of C. sociabilis, Pearson and Christie www.mammalogy.org
1409
1410 JOURNAL OF MAMMALOGY Vol. 93, No. 6

FIG. 1.—Maps depicting the location of the study area in Nahuel Huapi National Park, Neuquén Province, Argentina. In A), the location of the
park in southwestern Argentina is indicated. In B), the location of the study area (rectangle) within the park (hatched area) is denoted. In C), the
known geographic distribution of C. sociabilis is indicated by the dashed line. Within this area, the locations of the 3 focal study sites are denoted
by numbers, with 1 ¼ Paso Coihue, 2 ¼ La Lipela, 3 ¼ Rincón Grande.

into the ecological bases for group living in this behaviorally geographic distribution of the colonial tuco-tuco (Fig. 1). The
unusual species of ctenomyid rodent. area is bounded to the east by the Limay River, to the south by
Lake Nahuel Huapi, to the north by the Traful River and Lake
Traful, and to the west by dense montane forest (Pearson and
MATERIALS AND METHODS Christie 1985). The habitat in this region consists primarily of
Study area.—The study was conducted in Nahuel Huapi precordilleran steppe, which extends from the high-
National Park, Neuquén Province, Argentina. Data were precipitation (1,000 mm per year) eastern edge of the
collected in November–January of 2007 and 2008. This Andean–Patagonian forest to the low-precipitation (300 mm
corresponds to the austral spring–summer, which is the per year) western edge of the Patagonian steppe; the mean
portion of the year when animal activity and food resources annual isotherm for this region does not exceed 108C (Bran
are greatest (E.A. Lacey, in litt). The study area (40847 0 to 2000; Cabrera and Willink 1980). Precipitation occurs
40859 0 S and 71803 0 to 71820 0 W) comprises the current known primarily in winter, largely as snow, with a sharp west–east
December 2012 TAMMONE ET AL.—HABITAT RELATIONS OF COLONIAL TUCO-TUCO 1411

gradient caused by the rain shadow of the Andes (Bran 2000; examined provided a relatively unbiased sample of the other
Cabrera and Willink 1980). Vegetation in the area generally is habitats present at each focal locality.
consistent with precordilleran steppe habitat and is dominated The presence of C. sociabilis was determined on the basis of
by small shrubs and bunch grasses such as Festuca pallescens the size and configuration of burrow entrances (burrow
(coirón dulce), Stipa speciosa (coirón amargo), Mulinum openings ~7.5 cm in diameter, with multiple open entrances
spinosum (neneo), and Acaena splendens (cepa caballo; Bran located in proximity to one another—Lacey et al. 1997). No
2000; León et al. 1998). Toward the western end of the study other species of tuco-tuco is known to occur within the study
area isolated groves of Austrocedrus chilensis (ciprés de la area and no other rodents in this region construct burrow
cordillera) occur and, at higher altitudes, stands of deciduous entrances with the same configuration (Pearson 1995; Pearson
Nothofagus pumilio (lenga) are common (Bran 2000; León et and Christie 1985). As a result, it is unlikely that these criteria
al. 1998). resulted in the incorrect assignment of burrow systems to C.
Within the study area, 3 localities were chosen for detailed sociabilis. To determine if a burrow system was currently
investigation. These sites formed the apices of the roughly occupied, we looked for fresh (moist) mounds of dirt around
triangular geographic distribution of C. sociabilis (Fig. 1) and burrow openings or the presence of fresh (moist) soil plugs
thus they encompassed the variation in topography and climate within burrow entrances. Fresh mounds and plugs typically are
present within the range of this species. As a result, we expect generated daily; because exposed soil dries quickly, the
that these sites also encompassed much of the variation in presence of moist mounds or plugs is indicative of animal
habitat experienced by the study species. Because our data activity within the last 24 h (Pearson and Pearson 1993). In
were collected as part of the 1st quantitative assessment of the addition, because C. sociabilis is diurnal and produces a
geographic distribution of C. sociabilis, sampling focused on distinctive, birdlike alarm call that is emitted above ground
the extremes of this species’ suspected range. Given the limited (Pearson and Christie 1985), we typically waited at each area
area (~700 km2) in which C. sociabilis occurs (Fig. 1) and of apparent tuco-tuco activity for a period of 20–30 min to
obtain visual and auditory confirmation that animals were
given that our sampling regime encompassed the west–east
present. Burrow systems that did not provide evidence of
rainfall and habitat gradient that predominates in this region
current activity were recorded as unoccupied; long-term field
(Ezcurra and Brion 2005; Martı́n and Mermoz 2005; Mermoz
studies of C. sociabilis (e.g., Lacey and Wieczorek 2003, 2004)
et al. 2009), it is unlikely that this emphasis on the
indicate that unused burrow entrances quickly deteriorate and
distributional limits of the study species caused us to overlook
thus unoccupied burrow systems must have contained tuco-
significant variation in the types of habitats occupied by
tucos within the past 1–2 years.
colonial tuco-tucos.
For each location at which evidence of C. sociabilis
With regard to the 3 sites selected for detailed study, the
(occupied and unoccupied burrow systems) was detected, we
distance between adjacent sites was ~18 km. Paso Coihue
recorded a global positioning system (GPS) coordinate (WGS
(southwestern site) and La Lipela (northern site) were more
84, accuracy 5–7 m) using a hand-held Garmin ETrex GPS
mesic (.1,100 mm annual precipitation) and consisted of hilly unit. On the basis of the spatial distribution of burrow systems
cordilleran terrain characterized by deep valleys separated by monitored as part of long-term behavioral studies of this
steep, heavily vegetated slopes. Although Nothofagus forest species (Lacey and Wieczorek 2004), clusters of burrow
was present at both sites, Paso Coihue was characterized by a entrances located within a 10-m radius of one another were
dense understory of bamboo (Chusquea), whereas forested counted as part of the same burrow system and were assigned
areas at La Lipela were more open and brushy. In contrast, to the same GPS locality. In contrast, clusters of active burrow
Rincón Grande (southeastern site) was more arid (,800 mm entrances separated by .10 m were considered distinct and
annual precipitation) and was characterized by rolling hills and were recorded as separate GPS coordinates.
open, level plains. Microhabitat characterization.—To characterize the habitats
Identifying burrow systems.—At each focal locality, we occupied by C. sociabilis, we selected 57 burrow systems at
searched for burrow systems occupied by colonial tuco-tucos which tuco-tuco activity was detected for detailed analysis. At
by walking through the habitat; the minimum distance Rincón Grande, this included a subset (N ¼ 21) of the burrow
traversed at each locality was 20 km. Surveys did not follow systems monitored as part of long-term studies of the
a predetermined pathway and, although we attempted to behavioral ecology of C. sociabilis (Lacey and Wieczorek
include all major vegetation types present in the region (Martı́n 2004). At Paso Coihue and La Lipela, preliminary surveys
and Mermoz 2005), no quantitative assessment of the were used to identify the locations of occupied burrow systems,
prevalence of different habitat types was conducted. Because after which a subsample of localities was randomly selected for
previously published studies suggested that C. sociabilis is analysis; the total number of localities characterized at each of
associated with mallı́n habitat (Lacey and Wieczorek 2003; these sites (Paso Coihue N ¼ 21; La Lipela N ¼ 15) was
Pearson and Christie 1985), our surveys targeted mallı́n areas comparable with that examined at Rincón Grande.
but included the intervening habitat traversed while walking At each occupied burrow system, we established a 200-m2
from one mallı́n to the next. Thus, although mallı́n habitat rectangular plot centered on the point at which the GPS fix for
likely is overrepresented in our samples, the intervening areas that locality was taken. Plot size was based on the estimated
1412 JOURNAL OF MAMMALOGY Vol. 93, No. 6

area occupied by a single adult female C. sociabilis (E. A.

Lacey, in litt.). We then assigned the area within the plot to 1 of
3 habitat types previously defined for this part of Argentina
(Bran 2000). The 3 habitat types (Fig. 2) to which these
localities were assigned were: mallı́n–herbaceous vegetation
and hydrophilic grasses such as Carex spp., Juncus spp., and
Poa pratensis, with ,20% bare soil; shrub–woody vegetation,
primarily of Berberis spp. and Senecio spp., with .20% bare
soil; and mixed grasses–herbaceous and shrubby vegetation,
often including numerous grasses, with equal proportions of
vegetation and bare soil.
To provide a more quantitative assessment of the habitat at
each locality sampled, within each 200-m2 plot we established
5 2-m2 subplots; 1 subplot was placed at the center of the 200-
m2 plot, with the remaining 4 subplots placed 8 m away along
each of the 4 cardinal compass directions. Within each subplot,
percent cover was estimated for shrubs (e.g., Senecio
bracteolatus), woody herbs (e.g., M. spinosum), herbs (e.g.,
Trifolium), and bare soil. Because grasses (e.g., Poa) and
graminoids (e.g., Carex) are smaller and more cryptic, percent
cover for these vegetation types was estimated using 2 0.25-m2
quadrats nested within each 2-m2 subplot. Percent cover for all
vegetation types was estimated visually following the method
of Braun-Blanquet (Newton 2007). Values for each vegetation
type were averaged across subplots to yield a mean percent
cover for each 200-m2 plot sampled.
Digital habitat data.—Digital data layers are now routinely
used to assess environmental parameters, including patterns of
habitat use by species of conservation concern (Breininger et
al. 1991; Danks and Klein 2002). To compare habitat types
identified by digital environmental data with our field measures
of habitats occupied by C. sociabilis, the GPS locations of all
burrow systems detected (occupied and unoccupied) were used
to construct a digital layer depicting the occurrence of this
species at the 3 focal sampling localities. By superimposing
maps of burrow system localities onto existing environmental
data layers for this region obtained from the Delegación
Regional Patagonia (Administración de Parques Nacionales, S.
C. de Bariloche, Argentina), we were able to determine values
for annual precipitation (isohyet), elevation, and major
vegetation type (e.g., steppe grassland; Martı́n and Mermoz
2005) for each of the 3 study sites. Comparisons of digital
layers and quantification of environmental parameters were
completed using the Geoprocessing techniques available in
ArcView GIS 3.2 (ESRI 1998).
Data analysis.—To determine if the vegetative structure of
mallı́n and nonmallı́n (shrub and mixed grass habitats pooled)
habitats differed, we used a generalized linear model (GLZ)
with a normal distribution and a log link function (McCullagh
and Nelder 1989) to compare these habitat types with respect to

FIG. 2.—Photos of the 3 habitat types to which occupied burrow

systems of C. sociabilis were assigned during field surveys. For each plot assessed. Assignments of habitat types were based on the
burrow system examined, a 200-m2 rectangular plot was established identities of the plant species present and the overall percent cover by
(axes denoted by red strings in each photo) and vegetation within that vegetation versus bare soil, as described by Bran (2000).
December 2012 TAMMONE ET AL.—HABITAT RELATIONS OF COLONIAL TUCO-TUCO 1413

TABLE 1.—Number of burrow systems sampled and associated environmental conditions at each focal study site. Data on elevation and
precipitation were obtained by plotting the location of each burrow system (occupied and unoccupied) detected on environmental layer data
obtained from the Administracion de Parques Nacionales Argentinas. Numbers in parentheses are ranges.

Number of Number (%) of active

Site burrow systems burrow systems Elevation (m above sea level) Precipitation (mm/year)
Paso Coihue 128 45 (35) 1,428 6 197 (1,095–1,813) 1,553 6 50 (1,500–1,600)
La Lipela 90 29 (32) 1,389 6 84 (1,146–1,511) 1,193 6 25 (1,100–1,200)
Rincón Grande 147 59 (40) 793 6 39 (747–887) 670 6 61 (600–800)

mean percent cover by each of the 5 vegetation types Rincón Grande (Table 1). Burrow systems were found at
considered (shrubs, woody herbs, herbs, grasses, graminoids) elevations ranging from 750 to 1,800 m and in areas
plus mean percent bare soil. These analyses were completed characterized by precipitation ranging from 600 to 1,600 mm
using study site and habitat type (mallı́n versus non-mallı́n) as per year (Table 1). Among the focal sites, both elevation and
categorical variables under a full factorial model. Significant precipitation differed significantly (elevation: Kruskal–Wallis
differences in vegetative structure revealed by the GLZ H ¼ 262.93, P , 0.0001; Dunn post hoc analyses: Paso Coihue
analyses were then examined in greater detail by comparing versus La Lipela P . 0.05, all other comparisons P , 0.0001;
vegetation types or study sites directly using Kruskal–Wallis or precipitation: Kruskal–Wallis H ¼ 331.46, P , 0.0001; Dunn
Mann–Whitney U-tests. GLZ and post hoc comparisons were post hoc analyses: all P , 0.0001). Within each focal site,
completed using Statistica 5.1 (StatSoft 1997). however, the locations of occupied and unoccupied burrow
To determine if plant species composition differed among systems did not differ with respect to these parameters
study sites (locations of occupied burrow systems) or between (elevation: Rincón Grande Mann–Whitney U ¼ 2461.50, P
mallı́n versus nonmallı́n habitats, we used the nonparametric . 0.05; Paso Coihue Mann–Whitney U ¼ 1633.00, P . 0.05;
multivariate analysis of similarities (ANOSIM; Clarke 1993). La Lipela Mann–Whitney U ¼ 830.50, P . 0.05; precipitation:
ANOSIM is based on the Bray–Curtis similarity index and Rincón Grande Mann–Whitney U ¼ 2535.00, P . 0.05; Paso
produces a global R statistic that provides an absolute measure Coihue Mann–Whitney U ¼ 1717.50, P . 0.05; La Lipela
of the distance between data sets. R values approaching 1 Mann–Whitney U ¼ 881.50, P . 0.05).
indicate that the data sets are strongly differentiated, whereas Characterization of habitat types.—On the basis of
values approaching 0 indicate that groups cannot be distin- comparisons of the locations of burrow systems (occupied
guished. The significance of R was determined on the basis of and unoccupied) with existing digital environmental data
10,000 permutations of the data set among localities and layers, C. sociabilis occurred in 5 major habitat types (Table
between habitat types. Differences among the 3 focal sites and 2). At each study site, the majority of burrow systems
between mallı́n and nonmallı́n habitats were depicted using encountered occurred in herbaceous and woody steppe
nonmetric multidimensional scaling (NMDS) ordination on the habitat (Rincón Grande, La Lipela) or in Nothofagus forest
basis of Bray–Curtis similarity values. This analysis generates (Paso Coihue); no burrow systems were encountered in steppe
a stress value that reflects the degree of fit between the data grassland. At all sites, mallines (wet meadows) accounted for
matrix and the graphical, 2-dimensional NMDS output. A high 11% of burrow systems encountered (Table 2). For
stress value (stress ¼ 1) indicates a poor fit, suggesting that the occupied burrow systems (N ¼ 133), 3% occurred in mallı́n
graphical NMDS output distorts the underlying data set (Clarke habitats. Within each focal study site, the locations of active
and Warwick 2001). SIMPER analysis (Clarke 1993) was used and inactive burrow systems did not differ with respect to
to determine the contribution of each species to the mean habitat type (chi-square tests, Rincón Grande v2 ¼ 1.71, d.f. ¼
Bray–Curtis index values used in the NMDS analyses. 1, P . 0.05; Paso Coihue v2 ¼ 1.95, d.f. ¼ 2, P . 0.05; La
ANOSIM, NMSD, and SIMPER analyses were conducted Lipela v2 ¼ 2.31, d.f. ¼ 2, P . 0.05).
using the computer package PAST 1.81 (Hammer et al. 2001). Habitat characterizations conducted in the field for the subset
Throughout the text, means are reported 6 1 SD. Unless of 57 occupied burrow systems indicated that the occurrence of
otherwise indicated, a ¼ 0.05. When the same statistical test C. sociabilis in mallı́n habitats (28% of burrow systems) was
was used multiple times, a was adjusted using the Bonferroni significantly greater than expected given the occurrence of
correction procedure (Rice 1989). these animals in mallı́n habitats indicated by the digital
environmental data (3% of active burrow systems; v2 ¼ 120.3,
d.f. ¼ 1, P , 0.05). Field characterizations, however, were
RESULTS consistent with the digital data in revealing that the majority of
A total of 365 burrow systems (occupied and unoccupied) occupied burrow systems characterized did not occur in mallı́n
attributed to C. sociabilis was identified and their locations habitat (Fig. 3). Although nearly half (48%) of the 21 burrow
recorded as GPS coordinates. The number of burrow systems systems sampled at Rincón Grande were located in mallı́n
per study site varied, ranging from 90 at La Lipela to 147 at habitats, the occurrence of occupied burrow systems in
Rincón Grande (Table 1). The percent of active burrow mallines was significantly less at the other 2 study sites (Paso
systems per site ranged from 32% at La Lipela to 40% at Coihue: 19% of 21 burrow systems; La Lipela: 13% of 15
1414 JOURNAL OF MAMMALOGY Vol. 93, No. 6

TABLE 2.—Occurrence of burrow systems (occupied and unoccupied) by vegetation type. For each focal study site, the locations of occupied
burrow systems were compared with digital maps of vegetation types. Numbers in parentheses are percentages.

Number of burrow systems (% occurrence)

Vegetation type Paso Coihue La Lipela Rincón Grande Overall
a
Nothofagus pumilio forest 75 (58.6) 17 (18.9) 0 (0.0) 92 (25.2)
Herbaceous and woody steppeb 0 (0.0) 64 (71.1) 131 (89.2) 195 (53.4)
Transitional forestc 34 (26.0) 0 (0.0) 0 (0.0) 34 (9.3)
High-elevation desertsd 19 (14.8) 9 (10.0) 0 (0.0) 28 (7.7)
Mallı́ne 0 (0.0) 0 (0.0) 16 (10.8) 16 (4.4)
Total 128 (100) 90 (100) 147 (100) 365 (100)
a
Translated from bosque de lenga, b estepa herbácea y arbustiva enana, c mosaico de bosques bajos y/o matorrales con herbáceas, d semidesiertos de altura, and e vegas in Martı́n and
Mermoz (2005).

burrow systems; chi-square tests, frequency of occupied analyses revealed that occupied burrow systems at Rincón
burrows in mallines at Rincón Grande used to determine Grande had significantly higher mean percent cover by grasses
expected values; Paso Coihue v2 ¼ 52.85, d.f. ¼ 2, P , 0.0001; than occupied burrow systems at either Paso Coihue or La
La Lipela v2 ¼ 41.90, d.f. ¼ 2, P , 0.0001). Lipela (Kruskal–Wallis H ¼ 24.62, P , 0.0001; Dunn post hoc
Characterization of vegetation structure.—Generalized analyses: Rincón Grande versus Paso Coihue P , 0.01; Rincón
linear analyses (log-normal GLZ) indicated that the mean Grande versus La Lipela P , 0.0001, Paso Coihue versus La
percent cover by shrubs, woody herbs, and graminoids did not Lipela P . 0.05; Fig. 4a). For herbs, mean percent cover at La
differ with habitat type (mallı́n versus nonmallı́n), study site, or Lipela was significantly less than that at the other 2 study sites;
the interaction of these factors (shrubs: habitat Wald stat ¼ mean percent cover by herbs at Rincón Grande and Paso
0.36, P ¼ 0.54, site Wald stat ¼ 1.03, P ¼ 0.59, habitat 3 site Coihue did not differ (Kruskal–Wallis H ¼ 18.82, P ¼ 0.0001;
Wald stat ¼ 5.58, P ¼ 0.06; woody herbs: habitat Wald stat ¼
Dunn post hoc analyses Paso Coihue versus La Lipela P ,
1.20, P ¼ 0.27; site Wald stat ¼ 5.92, P ¼ 0.05; habitat 3 site
0.0001; Paso Coihue versus Rincón Grande P . 0.05; Rincón
Wald stat ¼ 1.44, P ¼ 0.48; graminoids: habitat Wald stat ¼
Grande versus La Lipela P , 0.001).
1.16, P ¼ 0.27; site Wald stat ¼ 3.20, P ¼ 0.20; habitat 3 site
Wald stat ¼ 1.09, P ¼ 0.57). Similarly, we found no effects of Characterization of plant species composition.—Analyses
these factors on the mean percentage of bare soil observed of similarities revealed significant differences between sites
(habitat Wald stat ¼ 0.94, P ¼ 0.33; site Wald stat ¼ 0.74, P ¼ with regard to plant species composition at the locations of
0.69; habitat 3 site Wald stat ¼ 0.57, P ¼ 0.75). In contrast, the occupied burrow systems (Rglobal ¼ 0.514, P , 0.0001; Fig. 5).
mean percent cover by herbs and grasses differed significantly Post hoc pair-wise comparisons indicated that although Paso
between habitats and between study sites (herbs: habitat Wald Coihue and La Lipela did not differ in terms of plant species
stat ¼ 20.86, P , 0.0001; site Wald stat ¼ 28.79, P , 0.0001; composition (R ¼ 0.005, P ¼ 0.3800; Fig. 5), Rincón Grande
grasses: habitat Wald stat ¼ 13.23, P , 0.0001; site Wald stat ¼ differed significantly from both of these sites (Rincón Grande
26.28, P , 0.0001); in neither case was the interaction between versus La Lipela R ¼ 0.778, P , 0.0001; Rincón Grande versus
habitat type and site significant (both P . 0.05). Post hoc Paso Coihue R ¼ 0.624, P , 0.0001).

FIG. 3.—Distribution of occupied burrow systems by habitat type. At each focal study site, field surveys were used to determine the percentage
of occupied burrow systems occurring in each of 3 habitat types (shrub, mallı́n, mixed grasses). For Rincón Grande, this represents all (N = 15 per
site) active burrow systems detected; for Paso Coihue and La Lipela, this represents a randomly selected subset (N = 21 per site) of the occupied
burrow systems at these sites.
December 2012 TAMMONE ET AL.—HABITAT RELATIONS OF COLONIAL TUCO-TUCO 1415

FIG. 5.—Analyses of plant species composition. Shown is the

nonmetric multidimensional scaling (NMDS) ordination output from
these analyses, which uses the Bray–Curtis dissimilarity index to
characterize differences in plant species composition among the study
sites. Analyses are based on the percent cover by plant species (N ¼
28) identified during analyses of 57 burrow systems occupied by C.
FIG. 4.—Comparisons of mean percent cover by herbs and grasses. sociabilis (Rincón Grande: N ¼ 21; Paso Coihue: N ¼ 21; La Lipela: N
In a), data are from the same 57 occupied burrow systems ¼ 15). Ellipses denote the 95% confidence interval of each site. The
characterized in Fig. 3. In b), data are from 10 mallı́n and 11 stress value indicates the fit between matrix data and the 2-
nonmallı́n localities at Rincón Grande. In c), data are from 16 mallı́n dimensional plane. High stress indicates a poor fit and that the MDS
localities (pooled across study sites) and 16 randomly selected representation distorts the underlying data.
nonmallı́n localities. * denotes significant (P , 0.05) contrasts
(statistical results presented in text). versus Paso Coihue P , 0.0001, Rincón Grande versus La
Lipela P ¼ 0.0003, La Lipela 3 Paso Coihue P . 0.05). Mean
The SIMPER analysis revealed that, of the 28 plant taxa percent cover by the remaining 5 of the 9 species identified by
included in our analyses, 9 accounted for .66% of the overall the SIMPER analysis did not differ significantly among study
dissimilarity between sites (Table 3). In particular, mean sites (Kruskal–Wallis tests, all P . 0.05).
percent cover by both the shrub Adesmia boronioides and the Comparisons of mallı́n and nonmallı́n habitats.—Given the
herb Acaena splendens was significantly higher at Paso Coihue predicted importance of mallı́n habitat to C. sociabilis and
and La Lipela (Adesmia boronioides: Kruskal–Wallis H ¼ given the results of GLZ and SIMPER analyses of the focal
24.46, P , 0.0001, Dunn post hoc analyses: Rincón Grande study sites, we compared the mean percent cover of herbs and
versus Paso Coihue P ¼ 0.019, Rincón Grande versus La grasses in mallı́n versus nonmallı́n habitats. Because the
Lipela P , 0.0001, La Lipela versus Paso Coihue P . 0.05; percentage of occupied burrow systems occurring in mallı́n
Acaena splendens: Kruskal–Wallis H ¼ 32.57, P , 0.0001, habitat was significantly greater at Rincón Grande (Fig. 3), we
Dunn post hoc analyses: Rincón Grande versus Paso Coihue P first compared mallı́n (N ¼ 10) and nonmallı́n (N ¼ 11) habitats
¼ 0.0014, Rincón Grande versus La Lipela P , 0.0001, La for this study site alone. This analysis revealed that mean
Lipela versus Paso Coihue P . 0.05). In contrast, mean percent cover by both herbs and grasses was significantly
percent cover by the grasses Stipa speciosa and Vulpia greater in mallı́n habitat (herbs: Mann–Whitney U ¼ 20.50, N ¼
australis was significantly higher at Rincón Grande (S. 10, 11, P ¼ 0.012; grasses: Mann–Whitney U ¼ 30.0, N ¼ 10,
speciosa: Kruskal–Wallis H ¼ 21.83, P , 0.0001, Dunn post 11, P , 0.05; Fig. 4b). We then pooled data for occupied
hoc analyses: Rincón Grande versus Paso Coihue P , 0.0001, burrow systems in mallı́n habitats at all 3 study sites (N ¼ 16)
Rincón Grande versus La Lipela P ¼ 0.001, La Lipela versus and compared that information with data from an equal number
Paso Coihue P . 0.05; V. australis: Kruskal–Wallis H ¼ of nonmallı́n burrow systems (N ¼ 16) that had been randomly
36.23, P , 0.0001, Dunn post hoc analyses: Rincón Grande selected from across the study sites. The results of this analysis
1416 JOURNAL OF MAMMALOGY Vol. 93, No. 6

TABLE 3.—Comparisons of plant species compositions of the 3 focal study sites. For each plant species identified, the mean percent cover at
each study site is indicated. Also indicated for each species is its contribution to the overall Bray–Curtis dissimilarity index for the study sites and
the cumulative percentage of species composition added by each taxon.

Mean cover (%)

Plant taxa Contribution to dissimilarity Cumulative % Paso Coihue La Lipela Rincón Grande
Acaena splendens 8.53 10.92 10.70 18.30 0.02
Adesmia boronioides 7.89 21.03 8.14 17.00 0.00
Stipa speciosa 6.37 29.20 3.39 1.43 13.30
Berberis microphylla 6.31 37.28 8.90 6.07 11.00
Poa ssp. 6.24 45.28 10.00 9.62 10.00
Mulinum spinosum 5.15 51.88 6.93 8.90 4.02
Senecio bracteolatus 4.78 58.00 0.07 2.66 10.80
Vulpia australis 4.69 64.00 0.00 0.35 10.90
Chiliotrichum rosmarinifolium 3.33 68.27 7.50 1.90 0.00
Festuca ssp. 3.11 72.26 5.17 4.47 0.14
Discaria trinervis 2.90 75.98 0.00 0.60 7.38
Trifolium repens 2.68 79.41 5.87 0.65 1.52
Carex/Juncus 2.37 82.45 0.63 0.53 5.83
Baccharis magellanica 2.23 85.30 3.62 1.90 0.00
Taraxacum officinale 2.00 87.86 3.77 0.90 1.51
Rytidosperma virescens 1.65 89.98 3.64 1.08 0.00
Bromus ssp. 1.63 92.06 1.26 1.70 1.54
Holcus lanatus 1.18 93.57 0.00 0.10 2.93
Agrostis ssp. 1.09 94.97 1.37 1.02 0.77
Hordeum ssp. 0.67 95.82 1.02 0.92 0.04
Carduus thoermeri 0.57 96.55 0.00 0.00 1.35
Maytenus chubutensis 0.55 97.26 1.50 0.00 0.00
Ribes cucullatum 0.52 97.93 1.10 0.53 0.00
Rosa rubiginosa 0.42 98.47 0.00 0.00 1.07
Nothofagus pumilio 0.41 98.99 1.14 0.00 0.00
Apera interrupta 0.36 99.45 0.00 1.03 0.07
Schinus patagonicus 0.27 99.79 0.02 0.00 0.71
Gaultheria pumila 0.16 100.00 0.33 0.00 0.00

were similar in that percent cover by grasses and herbs was such habitat. Instead, C. sociabilis also occurs in areas
significantly greater for mallı́n versus nonmallı́n localities dominated by woody shrubs that are characterized by a lower
(grasses: Mann–Whitney U ¼ 46.00, N ¼ 16, 16, P ¼ 0.001; percent cover of herbs and grasses than is typical of mallines.
herbs: Mann–Whitney U ¼ 35.00, N ¼ 16, 16, P , 0.0001; Fig. Although our sampling regime was not completely system-
4c). Finally, ANOSIM of these data indicated that plant species atic, 2 lines of evidence suggest that our findings provide a
composition differed significantly between mallı́n and reasonable representation of the habitats occupied by C.
nonmallı́n habitats (ANOSIM R ¼ 0.431, P , 0.0001); this sociabilis. First, given our sampling protocol at Paso Coihue
difference was due primarily to the greater abundance of and La Lipela, our data set should have favored burrow
grasses (genus Poa) in mallines at all 3 study sites. systems located in mallı́n areas yet, at both sites, the majority
of the active systems encountered occurred in nonmallı́n
DISCUSSION habitats. Second, because Rincón Grande is the site of a long-
term study of the behavioral ecology of this species (Lacey
Our analyses revealed that C. sociabilis occurred in mallines
2001; Lacey et al. 1997; Lacey and Wieczorek 2003, 2004)
at each of our focal study sites, suggesting that these animals
are associated with mallı́n habitat throughout their geographic that includes annual censuses of occupied burrow systems,
range. At each site, however, .50% of the occupied burrow sampling of active burrow systems at Rincón Grande was
systems examined occurred in nonmallı́n habitats; these effectively complete. Even under this intensive sampling
habitats differed significantly from mallines in terms of both regime, ~50% of active burrow systems were located in
plant species composition and the mean percent cover by nonmallı́n areas. Collectively, these observations suggest that
grasses and herbs, indicating that mallı́n and nonmallı́n areas despite potential biases in our sampling protocol, our data are
were quantitatively distinct. Thus, although our data are robust in suggesting that C. sociabilis is not restricted to mallı́n
consistent with those of Pearson and Christie (1985) and habitats.
Lacey and Wieczorek (2003) in confirming that C. sociabilis Digital versus field-based analyses of habitat use.—Our
occurs in mallines, our findings indicate that, contrary to analyses revealed a marked contrast between the percentages of
suggestions by these authors, the animals are not limited to burrow systems assigned to mallı́n habitat by digital
December 2012 TAMMONE ET AL.—HABITAT RELATIONS OF COLONIAL TUCO-TUCO 1417

environmental data layers versus direct, field-based study was conducted (Bran 2000; León et al. 1998). Seasonal
assessments of habitat type. Although multiple authors have variation in habitat use, however, cannot explain the
questioned the accuracy of digitally based analyses of species differences between our findings and those of Lacey and
distributions (e.g., Araújo et al. 2005; Gogol-Prokurat 2011; Wieczorek (2003). Data for both studies were collected during
Pearson and Dawson 2003), few studies appear to have the late austral spring and early austral summer and thus reflect
compared directly the results of digital and field-based the habitats occupied by C. sociabilis during the same portion
assessments of habitat availability. We suspect that the of the year. Further, long-term studies of this species indicate
greater use of mallines revealed by our field surveys reflects that a social group occupies the same burrow system in
the greater precision of these habitat data; given the resolution consecutive years, with no evidence of changes in the area
of the digital data layers used (scale 1:250.000; minimum occupied by a given set of individuals either within or between
mappable unit ¼ 12 ha—Martı́n and Mermoz 2005), it seems years (Lacey and Wieczorek 2004).
likely that smaller mallı́n patches recorded during field surveys Habitat use may also vary geographically. Our analyses
were not represented in the digital data, leading to revealed that the prevalence of C. sociabilis in mallı́n versus
underestimation of use of this habitat type in our digitally other habitat types differed among the 3 study sites.
based analyses. Although perhaps not surprising, this outcome Specifically, active burrow systems in mallı́n areas were most
has potentially important implications for studies that use common at Rincón Grande and least common at La Lipela.
digital data layers to assess habitat use, particularly for species This difference may reflect variation in the prevalence of mallı́n
that favor habitats distributed in small patches. habitats across study sites; the digital environmental layers
Habitat use and sociality.—Our characterization of the examined indicated that mallines were more prevalent at
habitats in which C. sociabilis occurs also has potential Rincón Grande. At the same time, habitat use by C. sociabilis
implications for understanding the distinctive social behavior may vary across the species’ range. Both elevation and
of this species. C. sociabilis is unusual among tuco-tucos in precipitation decline sharply along a west–east gradient across
that burrow systems are regularly occupied by multiple adult the study area; this variation is thought to substantially
females and, in some cases, a single adult male (Lacey 2000). influence the vegetation in the region (Ezcurra and Brion
Burrow sharing arises due to natal philopatry by females and 2005; Martı́n and Mermoz 2005) and may contribute to
thus social groups consist primarily of close female kin (Lacey differences among study sites regarding the relative use of
and Wieczorek 2004). Previous descriptions of habitat use by different habitat types by C. sociabilis. For example, our field
C. sociabilis suggested that mallı́n habitats are important for observations suggest that mallines at the western (higher
the social behavior of this species (Lacey and Wieczorek 2003; elevation, more mesic) end of the species’ distribution may be
Pearson and Christie 1985). In particular, Lacey and Wieczorek too wet to be occupied by these animals during much of the
(2003) hypothesized that the patchy spatial distribution of year. In contrast, at the eastern (lower elevation, more arid) end
mallines renders dispersal particularly costly for C. sociabilis, of the species’ distribution, mallines may be among the few
which might favor natal philopatry and the formation of social habitat types that are mesic enough to support C. sociabilis.
groups. To understand fully patterns of habitat use by C. sociabilis,
Our data indicate that specialization for mallı́n habitats per future studies of this species should undertake systematic
se is less important than previously assumed but do not negate surveys of habitat use versus availability, and should examine
the idea that a patchy distribution of suitable habitats favors the effects of geographic variation in environmental parameters
natal philopatry in this species. C. sociabilis occurs in the on the specific habitats used by these animals. Dietary analyses
precordilleran interface between the arid steppe grasslands of indicate that the primary foods consumed by C. sociabilis
eastern Patagonia and the temperate mesic forests of the Andes living in mallines are grasses and shrubs associated with mallı́n
(Mermoz et al. 2009). At the landscape scale, this area consists habitats (Bonvissuto et al. 2008; Lacey and Wieczorek 2003);
of a mosaic of steppe, forest, and mallı́n habitats (Mermoz et al. further research is needed to determine how food resources
2009), with patches of the latter 2 habitat types scattered differ in nonmallı́n areas. Variation in the habitats occupied by
throughout the prevailing steppe grassland. In our analyses, no conspecifics may also be associated with geographic differ-
burrow systems were detected in steppe habitat, despite the ences in behavior (Busch et al. 2000). Although some species
occurrence of this habitat type at all 3 study sites. Thus, even of Ctenomys are known to occur in a variety of habitats
the expanded range of vegetation types reported here may be (Lizarralde et al. 2001; Rosi et al. 2002), studies of these taxa
patchily distributed, with the result that dispersal between have not typically explored correlated differences in behavior.
suitable areas of habitat is difficult for C. sociabilis. Consequently, future studies that document social structure in
Quantitative analyses of the distributions of the different populations of C. sociabilis located throughout this species’
habitat types used by C. sociabilis are required to explore the range may reveal important relationships between habitat type,
relationship between habitat patchiness and sociality in this resource use, and social structure in this species.
species in greater detail. Implications for conservation.—Ctenomys sociabilis is
Variation in habitat use.—Patterns of habitat use may vary currently listed as ‘‘critically endangered’’ by the International
temporally, particularly for animals that inhabit temperate Union for Conservation of Nature and the Red Book of
environments such as the precordilleran region in which this Mammals of Argentina (Bidau et al. 2008; Dı́az and Ojeda
1418 JOURNAL OF MAMMALOGY Vol. 93, No. 6

2000), due primarily to its limited geographic range and the the project. This work was partially funded by a grant from Parque
generally disjunct distribution of local populations. Although Nacional Nahuel Huapi to MNT.
our findings regarding habitat use by C. sociabilis do not
contradict this conclusion, they do have implications for the LITERATURE CITED
conservation of this species. For example, because burrow ARAÚJO, M. B., R. G. PEARSON, W. THUILLER, AND M. ERHARD. 2005.
systems occurring in mallı́n and nonmallı́n habitats may Validation of species–climate impact models under climate change.
respond differently to conservation threats, use of a greater Global Change Biology 11:1504–1513.
range of habitat types may lead to greater resilience of C. BIDAU, C., E. LESSA, AND R. OJEDA. 2008. Ctenomys sociabilis. IUCN
sociabilis to deterioration or loss of mallı́n areas. Although our 2010. IUCN Red List of Threatened Species. Version 2010.2.
findings should be viewed positively by conservation biologists www.iucnredlist.org. Accessed 25 February 2011.
and regulatory agencies, further studies are required to BONVISSUTO, G. L., R. C. SOMLO, M. L. LANCIOTTI, A. GONZALES
CARTEAU, AND C. A. BUSSO. 2008. Guı́a de condición para pastizales
characterize directly relationships between habitat use and
naturales de ‘‘Precordillera’’, ‘‘Sierras y Mesetas’’ y ‘‘Monte
persistence of populations of colonial tuco-tucos in Nahuel Austral’’ de Patagonia (INTA, eds.). Buenos Aires, Argentina.
Huapi National Park. BRAN, D. 2000. Las regiones ecológicas de la Patagonia y sus
principales formaciones vegetales. Pp. 93–100 in Principios de
RESUMEN ecologı́a y conservación de los recursos naturales de la Patagonia
(INTA, eds.). Buenos Aires, Argentina.
Comprender los requerimientos de hábitat tiene implica- BREININGER, D. R., M. J. PROVANCHA, AND R. B. SMITH, 1991. Mapping
ciones para numerosos aspectos de la biologı́a de las especies, Florida scrub jay habitat for purposes of land-use management.
incluyendo donde viven los individuos y como estos se Photogrammetric Engineering & Remote Sensing 57:1467–1474.
comportan. La especialización en hábitat mésicos, ricos en BUSCH, C., ET AL. 2000. Population ecology of subterranian rodents.
recursos, conocidos como mallines se cree que ha favorecido la Pp. 183–226 in Life undergound: the biology of subterranean
vida en grupo en el tuco-tuco colonial (Ctenomys sociabilis), rodents (E. A. Lacey, J. L. Patton, and G. N. Cameron, eds.).
un roedor subterráneo endémico de la Provincia de Neuquén en University of Chicago Press, Chicago, Illinois.
CABRERA, A. L., AND A. WILLINK. 1980. Biogeografı́a de América
el suroeste de Argentina. Para explorar en mayor detalle la
Latina. Secretaria General de la Organización de los Estados
relación propuesta entre los mallines y la vida social en esta Americanos, Washington.
especie, se caracterizaron los hábitats ocupados por C. CLARKE, K. R. 1993. Non-parametric multivariate analyses of changes
sociabilis en 3 localidades que representan los extremos de la in community structure. Australian Journal of Ecology 18:117–143.
distribución geográfica de la especie. Especı́ficamente, se CLARKE, K. R., AND R. M. WARWICK. 2001. Change in marine
caracterizó la composición y estructura vegetal para 57 communities: an approach to statistical analysis and interpretation.
sistemas de madrigueras con animales, distribuidos en las 3 PRIMER-E, Plymouth, United Kingdom.
localidades de estudio. Los datos indican que C. sociabilis no DANKS, F. S., AND D. R. KLEIN. 2002. Using GIS to predict potential
se restringe a hábitats de mallı́n. A pesar de que se detectó wildlife habitat: a case study of muskoxen in northern Alaska.
variación significativa en la vegetación entre los 3 sitios de International Journal of Remote Sensing 23:4611–4632.
DÍAZ, G., AND R. OJEDA. 2000. Libro rojo de mamı́feros amenazados de
estudio, la mayorı́a de los sistemas de madrigueras activos
la Argentina. SAREM, Buenos Aires, Argentina.
muestreados en cada sitio se encontraron en hábitats no EMLEN, S. T. 1982. The evolution of helping. I. An ecological
mallı́noso. Además de proporcionar el primer muestreo amplio constraints model. American Naturalist 119:29–39.
de hábitat para C. sociabilis, nuestros datos podrı́an ofrecer EMLEN, S. T. 1991. Evolution of cooperative breeding in birds and
nuevas perspectivas sobre el papel de la especialización de mammals. Pp. 301–337 in Behavioral ecology (J. R. Krebs and N.
hábitat en la promoción del comportamiento social inusual de B. Davies, eds.). Blackwell Scientific, Oxford, United Kingdom.
esta especie de roedor ctenomido. ESRI. 1998. ArcView GIS 3.2. Environmental Systems Research
Institute, Inc. Redlands, California.
EZCURRA, C., AND C. BRION. 2005. Plantas del Nahuel Huapi: catálogo
ACKNOWLEDGMENTS de la flora vascular del Parque Nacional Nahuel Huapi, Argentina.
This work was completed in partial fulfillment of the requirements Universidad Nacional del Comahue-Red Latinoamericana de
for a Licenciatura in Biology from the Centro Regional Universitario Botánica, San Carlos de Bariloche.
Bariloche, Universidad Nacional del Comahue in S. C. de Bariloche, GOGOL-PROKURAT, M. 2011. Predicting habitat suitability for rare
Argentina. We thank the Delegación Técnica Regional Patagonia of plants at local spatial scales using a species distribution model.
Parques Nacionales Argentinas for permits and support, especially to Ecological Applications 21:33–47.
C. Chehebar. Thank to G. Iglesias for his assistence with the GIS layer HAMMER, O., D. A. P. HARPER, AND P. D. RYAN. 2001. PAST:
analyses. The staff of Parque Nacional Nahuel Huapi provided support paleontological statistics software package for education and data
in field. C. Ezcurra and L. Ghermandi helped with the plant species analysis. Palaeontologia Electronica 4:1–9.
identifications. N. Bacala provided advice regarding the statistical KOENING, W. D., F. A. PITELKA, W. J. CARMEN, R. L. MUMME, AND M.
analyses. We thank A. Thouyaret for permission to work at Rincón T. STANBACK. 1992. The evolution of delayed dispersal in
Grande and E. Obrum for permission to work at Paso Coihue. We cooperative breeders. Quarterly Review of Biology 67:111–150.
particularly thank J. Wieczorek for assistance in the field and for GIS LACEY, E. A. 2000. Spatial and social systems of subterranean rodents.
advice. M. Fernandez provided considerable personal support during Pp. 257–293 in Life undergound: the biology of subterranean
December 2012 TAMMONE ET AL.—HABITAT RELATIONS OF COLONIAL TUCO-TUCO 1419

rodents (E. A. Lacey, J. L. Patton, and G. N. Cameron, eds.). PEARSON, O. P. 1995. Annotated key for identifying small mammals
University of Chicago Press, Chicago, Illinois. living in or near Nahuel Huapi National Park or Lanin National
LACEY, E. A. 2001. Microsatellite variation in solitary and social tuco- Park, southern Argentina. Mastozoologı́a Neotropical 2:99–148.
tucos: molecular properties and population dynamics. Heredity PEARSON, O. P., AND M. I. CHRISTIE. 1985. Los tuco-tucos (género
86:628–637. Ctenomys) de los Parques Nacionales Lanı́n y Nahuel Huapi.
LACEY, E. A., S. H. BRAUDE, AND J. R. WIECZOREK. 1997. Borrow Historia Natural 5:337–343.
sharing by colonial tuco-tucos (Ctenomys sociabilis). Journal of PEARSON, R. G., AND T. P. DAWSON. 2003. Predicting the impacts of
climate change on the distribution of species: are bioclimate
Mammalogy 78:556–562.
envelope models useful? Global Ecology & Biogeography 12:361–
LACEY, E. A., AND J. R. WIECZOREK. 2003. Ecology of sociality in
371.
rodents: a ctenomyid perspective. Journal of Mammalogy 84:1198–
PEARSON, O. P., AND A. K. PEARSON. 1993. La fauna de mamı́feros
1211. pequeños de Cueva Traful I, Argentina: pasado y presente.
LACEY, E. A., AND J. R. WIECZOREK. 2004. Kinship in colonial tuco- Præhistoria 1:211–224.
tucos: evidence from group composition and population structure. REIG, O. A., C. BUSCH, M. O. ORTELLS, AND J. R. CONTRERAS. 1990. An
Behavioral Ecology 15:988–996. overview of evolution, systematics, molecular biology and
LEÓN, R. J. C., D. BRAN, M. COLLANTES, J. M. PARUELO, AND A. speciation in Ctenomys. Pp. 71–91 in Evolution of subterranian
SORIANO. 1998. Grandes unidades de vegetación de la Patagonia mammals at the organismal and molecular levels (E. Nevo and O.
extra andina. Ecologı́a Austral 8:125–144. A. Reig, eds.). Progress in Clinical and Biological Research 335:1–
LIZARRALDE, M., G. DEFERRARI, S. ALVAREZ, AND J. ESCOBAR. 2001. 422.
Diferenciación evolutiva en Ctenomys magellanicus: variación RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution
morfológica, alozı́mica y consideraciones biogeograficas de 2 43(1):223–225.
formas cromosómicas. Interciencia 26:13–17. ROSI, M. I., M. I. CONA, AND V. G. ROIG. 2002. Estado actual del
MARTÍN, C. E., AND M. A. MERMOZ. 2005. Mapa de vegetación del conocimiento del roedor fosorial Ctenomys mendocinus Philippi
1869 (Rodentia: Ctenomyidae). Mastozoologı́a Neotropical 9:277–
Parque y Reserva Nacional Nahuel Huapi. Anales de Parques
295.
Nacionales 7:51–62.
STATSOFT. 1997. Statistica. Version 5.1. StatSoft, Inc., Tulsa,
MCCULLAGH, M., AND J. A. NELDER. 1989. Generalized linear model.
Oklahoma.
2nd ed. Chapman and Hall, London, United Kingdom. WILSON, D. E., AND D. E. REEDER. 2005. Mammal species of the world:
MERMOZ, M., ET AL. 2009. El Parque Nacional Nahuel Huapi. Sus a taxonomic and geographic reference. 3rd ed. The Johns Hopkins
caracterı́sticas ecológicas y estado de conservación. APN, Parque University Press, Baltimore, Maryland.
Nacional Nahuel Huapi, San Carlos de Bariloche.
NEWTON, A. C. 2007. Forest as habitat. Pp. 285–331 in Forest ecology Submitted 27 July 2011. Accepted 13 March 2012.
and conservation: a handbook of techniques (A. C. Newton, eds.).
Oxford University Press, New York. Associate Editor was Douglas Kelt.