Responses of a Specialized Insectivorous Mammal (Myrmecophaga tridactyla) to Variation in Ambient Temperature1

Abstract

This study reports the effect of daily mean ambient temperature on daily activity and habitat use by the giant anteater (Myrmerophaga tridactyla) in the Pantanal wetlands of Brazil. We equipped I I giant anteaters with a global position system collar programmed to acquire a location every 10 min for 15 d. A temperature data logger left in the meteorological station at the study site registered data on ambient temperature. Giant anteaters were mainly active in open fields except during the coldest days, when they, sought protection inside forests. Giant anteaters also used open habitats for resting but during the hottest or chilliest days the), sheltered in forests. In the Pantanal, where temperatures are usually high, giant anteaters were active all night long, but as daily average ambient temperature decreased anteaters began and finished activity progressively earlier and reduced total activity As a consequence, time spent active it daylight increased progressively and time spent active during the night decreased progressively, probably to allow the anteaters to expose themselves to solar radiation and to avoid heat loss during the night.

Full text

BIOTROPICA 38(1): 52–56 2006 10.1111/j.1744-7429.2006.00106.x
Responses of a Specialized Insectivorous Mammal (
Myrmecophaga tridactyla
)
to Variation in Ambient Temperature1
Constanc¸a de Sampaio e Paiva Camilo-Alves and Guilherme de Miranda Mour˜
ao2
Embrapa Pantanal C.P. 109. Corumb´
a, Mato Grosso do Sul 79320-900, Brazil
ABSTRACT
This study reports the effect of daily mean ambient temperature on daily activity and habitat use by the giant anteater (Myrmecophaga tridactyla) in the Pantanal
wetlands of Brazil. We equipped 11 giant anteaters with a global position system collar programmed to acquire a location every 10 min for 15 d. A temperature
data logger left in the meteorological station at the study site registered data on ambient temperature. Giant anteaters were mainly active in open fields except
during the coldest days, when they sought protection inside forests. Giant anteaters also used open habitats for resting but during the hottest or chilliest days they
sheltered in forests. In the Pantanal, where temperatures are usually high, giant anteaters were active all night long, but as daily average ambient temperature decreased
anteaters began and finished activity progressively earlier and reduced total activity. As a consequence, time spent active at daylight increased progressively and
time spent active during the night decreased progressively, probably to allow the anteaters to expose themselves to solar radiation and to avoid heat loss during the
night.
RESUMO
O objetivo deste trabalho foi relacionar a actividade e uso do habitat do tamandu´
a-bandeira (Myrmecophaga tridactyla) com as temperaturas m´
edias di´
arias do
Pantanal Brasileiro. Equipamos 11 tamandu´
as-bandeira com um sistema de posicionamento global (GPS) previamente programado para registrar uma localizac¸˜
ao
cada 10 minutos durante 15 dias. Colocamos um registrador de temperatura na estac¸˜
ao meteorol´
ogica da ´
area de campo para registrar a temperatura ambiente
a cada 10 minutos. A atividade dos tamandu´
as-bandeira ocorreu essencialmente em habitat descobertos, excepto durante os dias mais frios, onde tamb´
em es-
tiveram ativos em florestas. Para repouso, usaram tamb´
em habitat descobertos, mas nos extremos de temperatura optaram por abrigo em ´
areas mais cobertas.
Os tamandu´
as tˆ
em apenas um per´
ıodo de atividade por dia, cuja durac¸˜
ao ´
e reduzida com a diminuic¸˜
ao da temperatura m´
edia di´
arias do ambiente. No Pan-
tanal, geralmente com clima quente, os tamandu´
as-bandeira s˜
ao noturnos mas, `
a medida que a temperatura m´
edia di´
aria diminui, a sua atividade inicia-se e
termina progressivamente mais cedo. Como conseq¨
uˆ
encia, `
a medida que os dias v˜
ao esfriando, os tamandu´
as-bandeira v˜
ao se tornando cada vez mais ativos
durante o dia e menos ativos durante a noite, aumentando assim a sua exposic¸˜
ao `
a radiac¸˜
ao solar e evitando que emanem muito calor corporal durante a
noite.
Key words: activity; giant anteater; habitat use; insectivore; Myrmecophaga tridactyla; Pantanal; seasonal flood plain.
GIANT ANTEATERS (MYRMECOPHAGA TRIDACTYLA)HAVE A NEOTROPI-
CAL DISTRIBUTION, which includes forests and savannas from Belize
to southern Brazil and northern Argentina (Wetzel 1985). They are
limited in energy expenditure because of their food habits based
strictly on prey with low calorific content (McNab 2000), so they
are restricted in their distribution to tropical environments (McNab
1985). Giant anteaters have a basal metabolic rate about 34 per-
cent of what is expected for their body size based on Kleibers’ rule
and have a low body temperature for a mammal (33◦C vs. 36–
38◦C for most mammals), but their long coarse fur coat and large
hairy tail provide them effective insulation so the minimal ther-
mal conductance is 94 percent of what is expected for their body
mass (McNab 1984). This low value of minimal thermal conduc-
tance, together with large body mass, allows giant anteaters to have
a very wide region of thermoneutrality, between 15.0◦Cand36
◦C,
while most of studied ant eating specialists have a thermoneutral
range of about 4–12◦C (McNab 1984). At temperatures below
15◦C, their body temperature can drop 2◦C because of their lim-
ited capacity to maintain high rates of metabolism (McNab 1984).
1Received 11 March 2005; revision accepted 26 April 2005.
2Corresponding author; e-mail: gui@cpap.embrapa.br
Nevertheless, anteaters live in the Pantanal, where ambient air tem-
perature can fall down to 0◦C (Calheiros & Fonseca 1996). Summer
temperatures may exceed 36◦C, which is the upper limit of anteater
thermoneutrality. Therefore, giant anteaters are likely to change
their activity and habitat use in response to the adverse climatic
conditions.
Activity periods recorded varied among studies. Shaw et al.
(1987) reported significant diurnal activity, which peaked in the
late evening, Medri and Mour˜
ao (2005) reported evening and noc-
turnal activity, and Montgomery and Lubin (1977) reported activity
throughout the night. There are also differences between studies in
relation to habitat use. Shaw et al. (1987) reported that anteaters
preferred to rest in scrub savannas and were more active in forests
in a region that also had grasslands. Medri and Mour˜
ao (2005) re-
ported more activity in grasslands and resting in forests in a region
where scrub savannas were present. Montgomery and Lubin (1977)
reported greater activity in forest patches in Venezuelan savannas.
Owing to the anteaters’ physiological characteristics, our hypothesis
was that the variation found in activity and habitat use would be
related to climatic variation, especially in temperature. Therefore,
our goals were to determine whether changes in average daily ambi-
ent temperature could explain daily variation in activity and habitat
use.
52 C2005 The Author(s)
Journal compilation C2005 by The Association for Tropical Biology and Conservation

Insectivore Responses to Temperature 53
MATERIAL AND METHODS
STUDY SITE.—The Pantanal is a seasonal flood plain located in
the west of Brazil, near the geographic center of South America at
about 100 m elevation. The weather is partly humid tropical with
25◦C average annual temperature (Calheiros & Fonseca 1996).
Summers (November to March) are hot and rainy with highest
air temperature reaching more than 40◦C and winters (April to
October) are warm and dry except for occasional cold fronts from
the South, which cause abrupt drops in air temperature. Average
minimum air temperatures are less than 20◦C, and the absolute
minimum is near 0◦C (Calheiros & Fonseca 1996).
This study was conducted in an area of about 94 km2on Nhu-
mirim and Porto Alegre Ranches (18◦59S, 56◦37W; Corumb ´
a,
MS). These ranches are covered by a mosaic of seasonally flooded
grasslands, savannas, scrub savannas, forests, and several permanent
and temporary ponds (Alho et al. 1987).
We captured the giant anteaters during the dry season, from
March to September 2002, usually during daylight and sometimes
at night until 2200 h. We located the anteaters visually, by car or by
horse. After that, we walked near them, which was possible because
anteaters are slow and have poor vision, lassoed and immobilized
them with a ketch all pole. We injected 2 mg/kg of Zoletil(Virbac
S. A., Carros Cedex, Franc¸a; association of tiletamine and zolazepan)
to sedate the anteaters. We used a modified global position system
(GPS) unit developed by Mour˜
ao and Medri (2002) to track the
animals. We fixed collars with the modified GPS unit on the scapular
waist of all anteaters that weighed more than 27 kg and had no
offspring. We took all appropriate precautions in regard to the
anteater’s welfare and safety.
After 15 d, we located the anteaters’ VHF signals and recap-
tured them to recover the GPS unit. No sedation was used at this
time. The GPS units were then reused.
The modified GPS unit consisted in a handheld Garmin E-
trex Venture GPS coupled to a radio-transmitter VHF collar (Collar
2/5920, advanced telemetry system—ATS). The GPS track-log was
programmed to store data every 10 min. Data field records included
date, time, and locations. If two adjacent locations were more than
30 m from each other, the anteater was considered to be active.
If the GPS did not receive enough satellite signals to fix a certain
location, resulting in 20 min or more between two locations, but
the animal was still in the same place, we considered that the animal
was inactive. With the records, we calculated the onset, termination,
and duration of daily activity.
We recorded the ambient temperature every 10 min using a
StowAwayTM Tidbit external temperature data logger at the me-
teorological station in the field study site. We used these data to
calculate daily average ambient temperature, the factor used to ex-
plain daily activity and habitat use.
Using Idrisi 32 for Windows (Eastman 2001) Geographic
Information System (SIG), we identified and mapped five-habitat
categories: forests, scrub savannas, savannas, seasonally flooded
grassland, and lagoons, according to Abdon et al. (1998). Forests
included semideciduous canopies and scrub forests with canopy
coverage between 70 and 100 percent. Scrub savannas had a lower
tree density, with canopy covering 25 to 69 percent. Savannas had
only sparse scrubs, with canopy covering less than 25 percent. Sea-
sonally flooded grasslands had almost no shrubs. Lagoons had fresh
or brackish water. The overlay of the GPS track-log with the geo-
referenced habitat map of the study area allowed us to analyze the
habitat type associated with each location and the percentage of
habitat use.
HABITAT USE.—We used the Friedman test to test for significant
differences between habitat use and habitat availability (Zar 1999).
To compare the time anteaters spent in one habitat to that habitat’s
availability in the home range, we estimated the electivity index
(EI) =(Ui−Di)/(Ui+Di), where Uiis the use of a given resource
proportion and Dithe resource availability proportion (Ivlev 1961).
The index varies between −1 (avoiding) and 1 (selecting), and 0
means that there are no preferences in habitat use. We analyzed
graphically the EI of all habitats.
To compare habitat use along the rank of daily average tem-
perature, we categorized temperature at 2◦C intervals. For each
temperature class, we estimated the percentage of time spent in
each habitat. We analyzed the results graphically.
ACTIVITY.—Giant anteater daily activity was described by its du-
ration and by times of onset and termination. We ignored periods
of inactivity that were less than 1 h during the activity period, and
periods of activity less than 1 h long during the extended resting pe-
riod. We analyzed the effect of the independent factor “daily average
temperature” on the daily activity duration using linear regression.
Although the time animals begin and finish activity is a circular fac-
tor, if data are concentrated in a short-time interval, we can apply
linear models using only that interval (Fisher 1993). Therefore, we
did a linear regression to analyze the effect of the independent factor
“daily average temperature” on times of onset and termination of
activity.
We did all graphical and statistical analysis with SYSTAT 9
for windows software (Wilkinson 1998). We excluded data from
the day of capture due to possible influence of the anesthesia and
handling stress on behavior.
RESULTS
We captured 17 giant anteaters. We affixed the modified GPS unit to
seven males (33–41 kg) and four females (27–39 kg). Temperature
measured at the meteorological station ranged from 8.0◦C (0445 h,
14 July) to 37.8◦C (1407 h, 10 September).
HABITAT USE.—Anteaters used habitat in proportions different to
habitat availability, both during activity (χ2=15.911, df =4,
P=0.003) and for resting (χ2=12.622, df =4, P=0.013). In
general, habitat preferences of active giant anteaters were inversely
related to the order of the canopy coverage gradient (Fig. 1a) and this
pattern was almost the opposite when the anteaters were inactive
(Fig. 1b), though this pattern varied with daily mean temperature.

54 Camilo-Alves and Mour˜
ao
FIGURE 1. Electivity index for habitat use by giant anteaters (Myrmecophaga
tridactyla) for (a) activity and for (b) resting in Pantanal wetlands, Brazil, ordered
by increase in coverage canopy. Scrub =scrub savanna; Positive values indicate
preference for and negative values avoidance of particular habitats.
When active, they used grasslands and savannas at frequencies higher
than predicted by availability, and apparently avoided forests, except
when air temperature was less than 17◦C (Fig. 2a). At intermediate
temperatures (17–27◦C), anteaters were inactive in grasslands and
savannas and during hot or chilly days they used forests for resting
(Fig. 2b).
ACTIVITY.—We observed only one activity period and one rest pe-
riod per day, with one exception. One anteater rested twice in 1 d,
but we considered this result an outlier and did not include it in
analyses. The activity period lasted on average 7 h 45 min ±2h
22 min SD and decreased with declining daily average ambient
temperatures (F(1,47) =18.15, P<0.001, R2=0.26; “activity
period” =−0.24 +0.34 ∗“daily average temperature”). Since
all animals started their activity between 0900 h and 0100 h, and
since all animals finished their activity between 1700 h and 0900 h,
we applied these intervals for linear modeling (Fig. 3). The time
FIGURE 2. Habitat used by giant anteaters for (a) activity and for (b) resting
in Pantanal wetlands, in each class of daily mean temperature. Use is expressed as
the percentage of all locations recorded at particular temperatures that occurred
in each habitat type.
anteaters began activity was positively related to daily average am-
bient temperature (F(1,51) =9.71, P=0.003, R2=0.14; “onset
time” =9.36 +0.37 ∗“daily average temperature”) (Fig. 3a) and
the time they finished activity was positively related to daily aver-
age ambient temperature (F(1,50) =33.63, P<0.001, R2=0.39;
“termination time” =12.95 +0.56 ∗“daily average temperature”;
Fig. 3b). Thus, during hot days (28–30◦C), anteaters were active
until sunrise. When daily average ambient temperature was about
25◦C, anteaters were active between sunset and late night, but not
at dawn. As daily average ambient temperature decreased, anteaters
tended to begin and finish activity earlier and reduce total activity.
During cold days (15◦C), they became essentially diurnal.
DISCUSSION
We found a sex ratio of two males for one female, less than the
sex ratio of three males for one female (N=12) found by Medri

Insectivore Responses to Temperature 55
FIGURE 3. (a) Onset and (b) termination of giant anteaters activity in relation
to daily mean temperature. We present the Y-axis 48 h sealed so as to simplify
plotting of periods that spanned midnight. (“Onset time” =9.36 +0.37 ∗“daily
average temperature”; R2=0.14); (“Termination time” =12.95 +0.56 ∗“daily
average temperature”; R2=0.39).
and Mour˜
ao (2005) in the same study field. Shaw et al. (1987)
found a 1:1 sex ratio (N=29) in the Canastra Mountains, Brazil,
and De Miranda (2004) found 1.5 male per female (N=32) in
a savanna area of central Brazil. Male-biased sex ratios are usual
in giant anteater populations, but it could also be an artifact due
to the capture procedures. Although we did not have enough an-
imals to analyze males and females separately, exploratory graphs
suggested that females might walk less during daylight. It is possible
that the bias found in sex ratio is a consequence of capture efforts
concentrated at daylight.
HABITAT USE.—When it was hot, anteaters used relatively more cov-
ered habitats for resting. On cold days, anteaters used less covered
habitat for both activity and resting. Giant anteaters have been ob-
served resting in open fields during cold days, exposing themselves to
solar heat (Medri 2002; this study). This basking behavior presum-
ably allows them to warm up before and during activity. However,
when daily ambient temperature fell below 17◦C, anteaters used
more forests, probably to shelter from the winds that are typical
during cold days in the Pantanal. During activity, the relationship
between habitat use and daily average ambient temperature was
weaker than when at rest, possibly because habitat choice during
activity is mainly related to prey availability.
ACTIVITY.—Giant anteaters feed on average 40 sec on each ant
or termite nest, then they move on looking for another (Redford
1985, Drumond 1992). Therefore, during foraging, anteaters move
constantly. This foraging behavior helps identify activity periods be-
cause animals are basically walking when active. Alternatively, long
pauses suggest true resting. Nevertheless, anteaters presented small
pauses up to 1-h long. Montgomery and Lubin (1977) considered
that movement patterns were related to prey distribution and abun-
dance, such that anteaters tend to move less when facing clustered
nests and we may not have detected movement in this situation.
However, anteaters may stop to protect themselves from potential
predators or adverse climatic conditions. Resting time was nearly
uninterrupted, although we did register small active periods of up to
1 h. This could be due to anteaters walking to change their shelter
due to disturbance, or it could be due to location errors by the GPS
unit.
Both times at which anteaters began or finished activity and
anteaters’ total activity were related to daily average ambient tem-
perature such that they could avoid insolation during hot days or
could absorb solar radiation during colder days. The study was con-
ducted during the dry season when prey availability is relatively
constant, thus reducing its effects on activity.
In conclusion, during the dry season, daily average ambient
temperature explained much of the variation in giant anteater ac-
tivity patterns in the Central Pantanal, Brazil. Similar patterns were
observed for the nine-banded armadillo (Dasypus novemcinctus)in
Florida (Layne & Glover 1985), where animals adjusted their ac-
tivity to avoid low temperatures in winter and high temperatures
in summer. Ant-eating specialists are known to have relatively low
metabolism and are sensitive to ambient temperature. They have
to minimize metabolic costs of heat production when the ambi-
ent temperature is low and they have to avoid overheating when
temperature is high. The strategies used by the giant anteaters to
reduce metabolic costs in cold days included decreasing heat loss by
diminishing total activity and by avoiding being active during the
coldest hours of the days and maximizing possible opportunities for
gain of heat through exposure to solar radiation during diurnal ac-
tivity in open areas. Giant anteaters apparently avoided direct solar
radiation in hot days through nocturnal activity and use of more
covered habitats for resting.
CONSERVATION IMPLICATIONS.—Historically, Giant anteaters were
abundant in the Cerrado and Pantanal regions. However, the
Cerrado has been greatly altered, and most of the savannas and
scrub savannas are being replaced by soybean cultures (Klink &
Moreira 2002). Many large- and medium-sized mammals from the
Cerrado still have healthy populations in the Pantanal. Therefore,
conservationists consider the Pantanal as an important refuge for

56 Camilo-Alves and Mour˜
ao
wildlife species, even though it is currently suffering from human
impact. In fact, more than 40 percent of the Pantanal’s forests
and scrub savannas have already been altered for cattle ranching
(Harris et al. 2005). The destruction of the forest will obviously
have a serious impact on species closely associated with them; fur-
thermore, it might also have an impact on a species not typical
of forested habitats. Giant anteaters are usually associated with
open vegetation formations (Eisenberg & Redford 1999); how-
ever, we have found that even when inhabiting open flooded sa-
vannas, like the Pantanal, they still require covered habitat for
thermoregulation. This is particularly critical during periods of
extreme temperatures, as giant anteaters have limited abilities to
control their body temperatures through metabolism. Therefore,
conservation measures for this species must also take into account
the protection of habitats providing sufficient coverage for their
thermoregulation.
ACKNOWLEDGMENTS
We thank Embrapa Pantanal and Conservation International for
logistic support and Fundac¸˜
ao para a Ciˆ
encia e Tecnologia (FCT),
Portugal e Fundo Social Europeu (FSE) no ˆ
ambito do III Quadro
Comunit´
ario de apoio for the grant #3990/2001 to Ms. Camilo-
Alves. We are grateful to Henrique de Jesus and Armindo Gonc¸alves,
who helped us in the field. IBAMA provided a license (#017/2002)
to capture and sedate the giant anteaters. Arnauld Debiez and
William Magnusson improved the draft manuscript.
LITERATURE CITED
ABDON,M.M.,J.S.V.SILVA,V.J.POTT,A.POT T,AND M. P. SILVA. 1998.
Utilizac¸˜
ao de dados anal´
ogicos do Landsat-TM na discriminac¸˜
ao da
vegetac¸˜
ao de parte da sub-regi˜
ao da Nhecolˆ
andia no Pantanal. Pesq.
Agropec. Bras. 33: 1799–1813.
ALHO,C.J.R.,Z.M.S.CAMPOS,AND H. C. GONC¸ALVES. 1987. Ecologia de
capivara (Hydrochaeris hydrochaeris, Rodentia) do Pantanal: -I Habitats,
densidades e tamanho di grupo. Revista Brasileira de Biologia 47: 87–97.
CALHEIROS,D.F.,AND W. C. D AFONSECA,JR. 1996. Perspectivas de estudos
ecol´
ogicos sobre o Pantanal, p. 41. EMBRAPA-CPAP, Corumb´
a, MS,
Brazil. (EMBRAPA-CPAP. Documentos, 18).
DEMIRANDA, G. H. B. 2004. Ecologia e conservac¸˜
ao do tamandu´
a-bandeira
no parque nacional das Emas, Goi´
as. Ph.D. dissertation, Universidade
de Bras´
ılia, Bras´
ılia, DF, Brazil.
DRUMOND, M. A. 1992. Padr˜
oes de forrageamento do tamandu´
a-bandeira
(Myrmecophaga tridactyla) no Parque Nacional da Serra da Canastra:
dieta, comportamento alimentar e efeito de queimadas. Master thesis,
Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil.
EASTMAN, J. R. 2001. Idrisi 32. Guide to GIS and Image Processing. Manual
Version 32.21. Clark Labs. http://www.clarklabs.org.
EISENBERG,J.F.,AND K. H. REDFORD. 1999. Mammals of the neotrop-
ics, volume 3, The central neotropics: Ecuador, Peru, Bolivia, Brazil,
pp. 92–93. The University of Chicago Press, Chicago and London.
FISHER, N. I. 1993. Statistical analysis of circular data. Cambridge University
Press, Cambridge, United Kingdom.
HARRIS,M.,W.TOM´
AS,G.MOUR˜
AO,C.J.DASILVA,E.GUIMAR˜
AES,F.SONODA,
AND E. FACHIM. 2005. Safeguarding the Pantanal wetlands: Threats and
conservation initiatives. Conserv. Biol. 19: 714–720.
IVLEV, V. S. 1961. Experimental ecology of the feeding of fishes. Yale University
Press, New Haven, Connecticut.
KLINK,C.A.,AND A. G. MOREIRA. 2002. Past and current human occupation,
and land use. In P. S. Oliveira and R. J. Marquis (Eds.). The Cerrados
of Brazil, pp. 69–88. Columbia University Press, New York.
LAYNE , J. N., AND D. GLOVER. 1985. Activity patterns of the common long-
nosed armadillo Dasyspus novemcinctus in South-Central Florida. In
G. G. Montgomery (Ed.). The evolution and ecology of armadillos,
sloth, and vermilinguas, pp. 407–417. Smithsonian Institution Press,
Washington, DC.
MCNAB, B. K. 1984. Physiological convergence amongst ant-eating and termite-
eating mammals. J. Zool. 203: 485–510.
———. 1985. Energetics, population biology, and distribution of Xenarthrans,
living and extinct. In G. G. Montgomery (Ed.). The evolution and ecol-
ogy of armadillos, sloth, and vermilinguas, pp. 219–232. Smithsonian
Institution Press, Washington, DC.
———. 2000. Metabolic scaling: energy constraints on carnivore diet. Nature
407: 584
MEDRI, I. M. 2002. ´
Area de vida e uso de habitat de tamandu´
a-bandeira—
Myrmecophaga tridactyla Linnaeus, 1758—nas Fazendas Nhumirim e
Porto Alegre,Pantanal da Nhecolˆ
andia, MS. Master Thesis, Universidade
Federal de Mato Grosso do Sul, Campo Grande, MS, Brasil.
———, AND G. MOUR˜
AO. 2005. Home range of the giant anteaters (Myrme-
cophaga tridactyla). J. Zool. 266: 365–375.
MONTGOMERY,G.G.,AND Y. D. LUBIN. 1977. Prey influences on movements of
neotropical anteaters. In R. L. Phillips and C. Jonkel (Eds.). Proceedings
of the 1975 predator symposium, pp. 103–131. Montana forest and
conservation experiment station, University of Montana, Missoula.
MOUR˜
AO,G.,AND I. M. MEDRI. 2002. A new way of using inexpensive large-
scale assembled GPS to monitor giant anteaters in short time intervals.
Wildl. Soc. Bull. 30: 1029–1032.
REDFORD, K. H. 1985. Feeding and food preference in captive and wild
giant anteaters (Myrmecophaga tridactyla). J. Zool. Lond. 205: 559–
572.
SHAW,J.H.,J.MACHADO-NETO,AND E. T. S. CARTER. 1987. Behaviour of free-
living giant anteaters (Myrmecophaga tridactyla). Biotropica 19: 255–
259.
WETZEL, R. M. 1985. The identification and distribution of recent Xenarthra
(=Edentata). In G. G. Montgomery (Ed.). The evolution and ecology of
armadillos, sloth, and vermilinguas, pp. 5–21. Smithsonian Institution
Press, Washington, DC.
WILKINSON, L. 1998. SYSTAT—Systems for Statistics. Version 9. Systat Soft-
ware Inc., Chicago.
ZAR, J. H. 1999. Biostatistical analysis, pp. 263–267. Prentice-Hall, Inc., Upper
Saddle River, New Jersey.