ECOLOGICAL, TAXONOMIC, AND PHYSIOLOGICAL CORRELATES OF

CAVE USE BY MEXICAN BATS

Author(s): Rafael Avila-Flores and Rodrigo A. Medellín
Source: Journal of Mammalogy, 85(4):675-687. 2004.
Published By: American Society of Mammalogists
DOI: http://dx.doi.org/10.1644/BOS-127
URL: http://www.bioone.org/doi/full/10.1644/BOS-127

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, 85(4):675–687, 2004

ECOLOGICAL, TAXONOMIC, AND PHYSIOLOGICAL

CORRELATES OF CAVE USE BY MEXICAN BATS
RAFAEL AVILA-FLORES AND RODRIGO A. MEDELLÍN*
Instituto de Ecologı́a, Universidad Nacional Autónoma de México, Apartado Postal 70–275, México,
Distrito Federal 04510, México

Roost use by bats is likely affected by their water balance and thermoregulatory abilities. To test this hypothe-
sis, we explored the relationship between 4 traits of different species of bats (body size, general food habits,
taxonomic group, and thermoregulatory pattern) and microclimates at roosts (temperature and humidity). We
recorded roost variables and presence of bats in 18 caves from 5 contrasting biomes in central Mexico. There was
little evidence of microclimatic specificity among the 23 species studied, but maternity colonies used warmer
roosts and hibernating bats used cooler roosts. Heterothermic species (Vespertilionidae) used colder caves with
the widest temperature range (1.6–29.88C), whereas homeothermic species (Emballonuridae, Mormoopidae,
Phyllostomidae, and Natalidae) occupied warmer roosts (14.5–37.58C). Within these caves, precise (narrow body
temperature range) homeotherms occupied slightly cooler roosts than more labile homeotherms. Body size alone
was not associated with cave use patterns. However, when homeotherms and heterotherms were examined
separately, body size and temperature were negatively correlated. The smallest homeothermic insectivorous
species (,10 g) consistently occupied roosts with temperatures .208C (more often .258C), whereas only the
largest homeothermic insectivores were found as low as 168C. Frugivorous, nectarivorous, and sanguivorous bats
were found in a wide range of temperatures (14.5–378C), but often at ,208C. Humidity in roosts was highly
variable for most species and we did not detect any trend regarding this factor. Our data suggest that the
thermoregulatory ability resulting from the complex interaction of body size, type of food, and taxonomic
affiliation constrains species with respect to types of roosts that they can successfully exploit. Our results support
the hypothesis that temperature is the most important physical factor influencing roost selection in bats.

Key words: bats, caves, Mexico, microclimate, thermoregulation, use of roosts

Roosts are key resources for bats because bats spend over
half their lives subjected to selective pressures of roost
environments (Kunz 1982; Vaughan and O’Shea 1976). Bats
occupy a variety of roosts, but many species use caves as
roosting sites (Hill and Smith 1984; Kunz 1982). In Mexico,
almost half of the 138 species use caves as primary or
alternative roosts (Arita 1993).
Roost selection by bats depends on many factors, including
temperature, humidity, air flow, light intensity, safety from
predators, proximity to foraging areas, and takeoff height (Gaur
1980; Hill and Smith 1984; Kunz 1982; McCracken 1989;
Morrison 1980; Tuttle and Stevenson 1981). Within a given
species, individual roost requirements vary according to
season, sex, reproductive status, and age (Humphrey 1975;
Stebbings 1995). Roost selection can be important for
* Correspondent: medellin@miranda.ecologia.unam.mx
Ó 2004 American Society of Mammalogists
www.mammalogy.org
675
optimizing several physiological processes. For example,
temperature, humidity, and airflow could be the most important
physical factors influencing roost selection (Hill and Smith
1984; Tuttle and Stevenson 1981). Occupation of roost sites
with appropriate microclimate can minimize energetic costs
related to thermoregulation, food digestion and assimilation,
maintenance of a permanent state of alertness (which allow
bats to avoid predation and to interact socially), gestation,
embryonic development, parental care, lactation, and sper-
matogenesis (Bonaccorso et al. 1992; Hamilton and Barclay
1994; Hill and Smith 1984; Humphrey 1975; Kunz 1973;
McNab 1982; Tuttle and Stevenson 1981; Twente 1955).
Given the benefits obtained from selecting optimal roosting
sites, a high specificity in roost use can be expected among
species of bats. Although several bat species have highly
specific roost requirements (Baudinette et al. 2000; Humphrey
1975; McCracken 1989; McNab 1974), the opposite trend can
be expected for wide ranging, common species considering that
optimal roosts are limited in number (e.g., Tuttle and Stevenson
1981). Most bats are opportunistic with respect to roosting
habits, using a variety of roosts, including man-made structures
676 JOURNAL OF MAMMALOGY
(Altringham 1996; Fenton 2001; Kunz 1982). Anecdotal
observations suggest that many species can tolerate a wide
range of roost conditions.
If optimal roost conditions are not frequently available and
high specificity is not common among bats, what factors
ultimately limit the type of roosts that bats can use? Diversity
of roost use by bats may largely reflect different physiological
responses to microenvironmental conditions (Silva 1979).
Thermoregulatory and water-balance abilities seem to influence
the ability of species to use different kinds of roosts (Baudinette
et al. 2000). If this is true, factors that determine such
physiological traits should also influence the degree of
plasticity shown by bats regarding use of roosts.
Thermoregulatory abilities of mammals (including bats)
depend on factors such as body size, basal metabolic rate,
thermal conductance, food habits (Degen 1997; Elgar and
Harvey 1987; Hayssen and Lacy 1985; Lyman 1970; McNab
1982, 1989, 1992), and probably phylogeny (Elgar and Harvey
1987; Hayssen and Lacy 1985; McNab 1997). The problem of
water balance is more complex because it is linked to osmo-
regulation and excretory systems. For example, although food
habits are known to affect water intake, some mammals have
adapted to eat dry foods by modifying kidney structure (Degen
1997). However, body size is a primary factor influencing water
balance in small mammals (Degen 1997; Studier 1970).
In this study we explore the hypothesis that variation in
roost use exhibited by bats strongly reflects their thermoregu-
latory and water-balance abilities. We evaluated the rela-
tionship between some features associated with physiological
performance and the microclimate of sites used as roosts. Par-
ticularly, we related body mass, food habits, taxonomic affinity,
and thermoregulatory pattern of 23 species of bats to levels
of temperature recorded in natural roosts of central Mexico.
Because of the complexity of factors affecting balance of water
in mammals, we restricted correlates of roost humidity ex-
clusively to body size.
MATERIALS AND METHODS
Study sites.—We studied 18 caves and mines in the states of Puebla
and Mexico, in central Mexico. This area is complex and
heterogeneous in topography, vegetation, and climate (Ortega and
Arita 1998; Rzedowski 1978). We studied caves in 5 physiographic
regions (Balsas River Basin, Gulf Coastal Plain, Sierra Madre
Oriental, Tehuacan Valley, and Trans-volcanic Belt—Rzedowski
1978) and 5 habitats (tropical evergreen forest, tropical deciduous
forest, subtropical cloud forest, temperate coniferous forest, and
xerophytic scrub; Appendix I).
Roost characteristics.—From December 1997 to October 1998, we
conducted four 2-day-long seasonal (winter, spring, summer, and
autumn) visits to each cave. At each visit we measured air temperature
and relative humidity using a digital thermohygrometer (precision 6
2% relative humidity and 6 0.48C; Hanna Instruments, Inc., Ann
Arbor, Michigan) as close as possible to the roosting surfaces.
Microclimatic variables were measured at 0400, 1000, 1600, and
2200 h each day. Microclimatic stations were established exclusively
at sites occupied by bat colonies, regardless of colony size. We started
each visit by mapping the specific roosting sites of each species. The
number of stations in each cave ranged from 2 to 13.
Vol. 85, No. 4
We determined the type of rock by direct observation and made
a topographic map of each cave following methods of Nuñez-Jiménez
et al. (1988). We also noted if bats roosted inside or outside cavities
(crevices, cracks, fissures) with a diameter or width , 50 cm.
Biological data.—We captured bats nightly and daily at entrances or
inside caves using mist nets and hand nets, and identified species
following Medellı́n et al. (1997). We obtained information on size,
sex, age (juvenile, adult), reproductive status (only for females:
inactive, pregnant, lactating), and apparent thermoregulatory state
(normothermic, hypothermic, torpid), partly to determine roost use—
hibernacula, maternity roost, or nonbreeding male and/or female
roosts. Normothermic bats were individuals observed visibly awake
when encountered and/or found warm at touch (normothermia refers to
optimal body temperature, i.e., a temperature that allows common
physiological processes to occur); torpid bats (including hibernating)
were those individuals exhibiting no apparent movement and found cold
at touch; hypothermic bats were identified as above, but were arbitrarily
placed in this category when corresponding to typical homeothermic
species (e.g., McNab 1969, 1982) that can decrease body temperature
under unusual conditions (Audet and Thomas 1997). Because roost use
can reflect different individual physiological requirements, we com-
pared the observed physiological state of bats to roost temperature and
humidity, regardless of species. We compared roosts used by basal state
bats (not pregnant or lactating, normothermic bats) to those used by
reproductive (pregnant or lactating) females, torpid bats, and hypother-
mic bats.
We calculated mean body size of species based on body mass of
captured specimens; we obtained additional measurements from
museum specimens (Colección Nacional de Mamı́feros, Instituto de
Biologı́a, UNAM) when ,10 individuals were measured in the field.
We combined masses for males and females in calculating means.
Information about diet, taxonomy, and physiological responses to
changes in environmental humidity and temperature (both free living
and captive) was obtained from the literature.
Main food type, taxonomic affiliation and thermoregulatory pattern
were separated into categories, and each species was assigned to 1 of
these categories. Food categories were insects, nectar and pollen, fruit,
and blood. Taxonomic groups included families Emballonuridae,
Mormoopidae, Natalidae, and Vespertilionidae, and subfamilies
Phyllostominae, Glossophaginae, Carolliinae, Stenodermatinae, and
Desmodontinae (Phyllostomidae). We classified thermoregulatory
patterns as follows (modified from Dwyer 1971 and McNab 1982).
Type 0, species that can maintain body temperature (Tb) a few degrees
above ambient temperature (Ta), capable of entering deep torpor and
arousing from it at very low Tas; basal metabolic rate lower than that
expected from Kleiber function. Type 1, species that behave like Type
0 but do not enter deep torpor nor arouse from low Tbs (at low Tbs
they become hypothermic). Type 2, labile homeotherms, or species
that maintain Tb within a moderately wide range when exposed to
a variety of Tas; basal metabolic rate less than or equal to that
expected. Type 3, precise homeotherms, species that maintain Tb
within a narrow range under a wide range of Tas, although some
species can relax Tb at moderate Tas (Audet and Thomas 1997;
McNab 1969; Studier and Wilson 1970); basal metabolic rate is equal
or higher to that expected from the Kleiber function.
Species were classified into types 0, 2, and 3. We classified no
species as type 1. Although individuals of N. stramineus can become
lethargic in nature under unusual environmental conditions (McNab
1982; Nowak 1994), we always observed this species alert at Tas
ranging between 178C and 258C, and never forming compact groups or
entering a state resembling hypothermia or torpor. Although vampire
bats (Desmodus rotundus and Diphylla ecaudata) can tolerate Ta as
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS 677

TABLE 1.—Body size and thermoregulatory traits of 23 species of bats that use caves in central Mexico. Patterns of thermoregulation: 0,
heterotherms which can enter in deep torpor; 2, labile homeotherms (moderate range of Tbs); 3, precise homeotherms (narrow range of Tbs). Tb ¼
body temperature, Ta ¼ ambient temperature.

Mean basal Mean Tb in Mean Tb

Mean body metabolic rate thermoneutrality at Ta ; 108C Thermoregulatory
Species mass (g) (ml O2 g/h) (8C) (8C) pattern Source
a
Balantiopteryx plicata 5.3 1.86
2.65 2 Genoud and Bonaccorso 1986
Mormoops megalophylla 15.1 1.48 36.9 37.0 3 Bonaccorso et al. 1992
Pteronotus davyi 7.3 1.63 38.8 33.0 2 Bonaccorso et al. 1992
P. parnellii 16.0 1.60 36.4 34.5
35.3 3 Bonaccorso et al. 1992
P. personatus 5.5 1.64 37.5 32.5 2 Bonaccorso et al. 1992
Macrotus waterhousii 15.1 1.25b 33.5
37.5 32.0 2 Bell et al. 1986; Riedesel and Williams 1976
Micronycteris megalotis 6.7 38.6 17.5 2 McManus 1977; Studier and Wilson 1970
Choeronycteris mexicana 15.2 3
Glossophaga soricina 9.8 2.25 35.5 11.5
36.5 3 McManus 1977; McNab 1969, 1989;
Studier and Wilson 1970
Leptonycteris curasoae 22.5 1.42, 1.66, 2.00 35.3
35.7 35.0
37.0 3 Arends et al. 1995; Carpenter and
Graham 1967; McNab 1969, 1989
L. nivalis 30.2 3
Carollia brevicauda 15.9 3
C. perspicillata 18.3 2.11 36.4 12.0
37.0 3 McManus 1977; McNab 1969,
1989; Studier and Wilson 1970
Artibeus jamaicensis 41.5 1.25 36.4 14.0
38.0 3 McManus 1977; McNab 1969,
1989; Studier and Wilson 1970
Desmodus rotundus 34.9 0.91 35.0 32.0
36.0 2 McManus 1977; McNab 1969, 1989
Diphylla ecaudata 31.1 1.22 32.4 31.0
32.5 2 McManus 1977; McNab 1969, 1989
Natalus stramineus 5.9 2c
Corynorhinus mexicanus 7.2 0
C. townsendii 8.8 0
Eptesicus fuscus 16.2 0.80
1.20 0 Herreid and Schmidt
Nielsen 1966
Myotis keaysi 3.8 1.30d 0 McNab 1989
M. velifer 9.7 0.65
1.55 36.0
36.2 11.8
12.1 0 Riedesel and Williams 1976
M. volans 5.8 0
a
Calculated for other emballonurids (Saccopteryx leptura, S. bilineata, and Peropteryx macrotis).
b
Calculated for Macrotus californicus.
c
McNab (1982) places this species in thermoregulatory pattern 1.
d
Calculated for Myotis nigricans.

low as 108C (McNab 1969, 1982; Villa-R. 1966), and could be
placed in category 3, we assigned these species to category 2
because of their very low basal metabolic rates (McNab 1989), and
because some individuals maintain body temperatures lower than
338C at ambient temperatures of 108C (McNab 1969).
Statistical analysis.—We calculated mean and range of temperature
and humidity for each species using the measurements taken when at
least some individuals were present at the roost site. In order to group
species that use microclimatically similar roosts, we performed
a cluster analysis using Ward’s hierarchical method (SAS Institute
Inc. 1995) and obtained corresponding dendrograms. We computed
2-cluster analyses, 1 for temperature and 1 for humidity, comparing
among species mean roost temperature, range of temperature, mean
roost relative humidity, and range of relative humidity.
Further statistical analyses were performed using a database
containing all temperature and humidity site-specific averages. We
calculated these averages using both nightly and daily measurements of
temperature and humidity taken at a given site during each 2-day visit,
each average representing 1 data point. Differences in roost use
according to observed physiological state of bats were evaluated by
nonparametric analysis of variance (Kruskall–Wallis test). Body mass
was the only numerical and continuous variable examined. We used
correlation analyses (Pearson coefficients) to explore the relationships
between mean body mass and 2-day site-specific mean roost temper-
atures, and between mean body mass and 2-day site-specific mean
relative humidity. We evaluated differences in roost temperature
among categories or groups of species separately for each of the other 3
variables. Our data did not follow a normal distribution and variances
were not homogeneous, thus we used nonparametric, 1-way analysis of
variance (Kruskall–Wallis test) to look for differences among groups.
Paired differences between groups were evaluated by means of Mann–
Whitney tests. For relative humidity, we only explored correlations
with body size. In all cases we used a critical alpha value of 0.05. We
used computational software from SAS Institute Inc. to conduct the
cluster analysis and SPSS software for other analyses.
RESULTS
We found 23 species of bats of 5 families (number of spe-
cies in parentheses): Emballonuridae (1), Mormoopidae (4),
Phyllostomidae (11), Natalidae (1), and Vespertilionidae (6).
Five subfamilies were represented among phyllostomids:
Phyllostominae (2), Glossophaginae (4), Carolliinae (2),
Stenodermatinae (1), and Desmodontinae (2). Information
about each species is summarized in Tables 1–3.
Physiological state, regardless of species, was significantly
correlated with roost temperature (v2 ¼ 108.9, d.f. ¼ 3, P ,
0.001; Fig. 1), but not with roost humidity (v2 ¼ 0.280, d.f. ¼
3, P ¼ 0.964). Torpid bats were observed at significantly cooler
roosts (U ¼ 2,264, P , 0.001), whereas reproductive females
occupied warmer places (U ¼ 5,414.5, P ¼ 0.015) compared to
678 JOURNAL OF MAMMALOGY Vol. 85, No. 4

TABLE 2.—Characteristics of roosting places detected for 23 species of bats found in 18 caves in central Mexico (see Table 1 for generic
names). For roost and temperature humidity, the mean (6 1 SD), lowest, and highest values both from daily and nightly measurements are shown.
Size of caves: 1, ,20 m; 2, 20–39 m; 3, 40–79 m; 4, 80–159 m; 5, 160–319 m; 6, 320–640 m. Type of rock: L, limestone; C, clastic. Degree of
darkness: L, lighted; I, intermediate; D, dark. Cavities refers to all cavities ,1 m in diameter (I ¼ bats inside cavity, O ¼ bats outside cavity).

Other characteristics of roosts

Roost temperature (8C) Roost humidity (%)
Number of Size of Type of Degree Inside/Outside
Species caves used 
X SD (n) Min Max 
X SD (n) Min Max caves rock of darkness cavities
B. plicata 6 26.7 3.1 (319) 18.6 37.5 50.2 20.2 (273) 10.0 100 1,2 C L,I I,O
M. megalophylla 4 32.9 4.2 (91) 16.2 37.0 64.2 13.8 (75) 46.6 100 3,4,5 C,L D O
P. davyi 1 34.7 0.7 (56) 33.4 36.1 65.4 8.6 (48) 51.5 91.0 5 C D O
P. parnellii 4 26.1 7.3 (178) 16.0 37.0 82.9 19.8 (159) 46.6 100 3,5,6 C,L D I,O
P. personatus 1 34.7 0.7 (58) 33.4 36.1 65.7 8.7 (50) 51.5 91.0 5 C D O
M. waterhousii 2 30.2 2.8 (174) 24.5 37.0 52.9 13.6 (151) 25.4 99.0 2,5 C I,D I,O
M. megalotis 2 28.0 2.8 (60) 22.0 32.2 62.2 18.6 (46) 20.5 100 1 C I,D O
C. mexicana 2 22.0 2.9 (74) 15.0 29.5 59.6 13.0 (64) 33.0 91.8 1 C L,I,D I,O
G. soricina 2 20.6 2.1 (237) 16.8 27.2 98.9 2.5 (201) 86.5 100 5,6 C,L D O
L. curasoae 5 29.5 4.9 (91) 21.1 37.0 70.6 19.1 (81) 44.3 100 1,3,5 C,L L,I,D I,O
L. nivalis 1 17.0 1.6 (5) 14.5 18.6 82.4 11.1 (5) 71.0 100 4 C D O
C. brevicauda 2 19.2 1.6 (124) 15.0 21.5 97.8 3.9 (111) 84.4 100 3,6 L D O
C. perspicillata 2 21.6 2.7 (108) 15.2 28.5 97.9 3.4 (93) 86.5 100 5,6 C,L D I
A. jamaicensis 1 21.9 2.6 (139) 15.7 27.2 98.0 4.5 (118) 69.9 100 5 C L,I,D I
D. rotundus 5 21.7 3.1 (327) 15.9 33.2 88.9 19.6 (277) 20.5 100 1,2,3,5,6 C,L I,D I,O
D. ecaudata 4 20.6 1.6 (180) 16.8 25.0 98.1 3.7 (151) 81.8 100 2,3,5,6 C,L D O
N. stramineus 2 22.4 2.1 (61) 17.0 25.4 98.6 3.0 (50) 88.3 100 3,5 C,L I,D I
C. mexicanus 4 7.5 1.8 (146) 1.6 11.6 83.0 13.5 (138) 45.4 100 2,3 L I,D O
C. townsendii 2 14.3 3.3 (82) 4.9 20.0 80.8 10.5 (76) 55.4 100 4 L D I,O
E. fuscus 2 12.6 4.1 (4) 7.8 16.2 84.8 10.8 (5) 75.0 100 3,4 L
M. keaysi 3 20.8 3.5 (147) 12.5 29.8 96.6 7.3 (133) 63.3 100 2,5,6 C,L I,D I
M. velifer 8 11.6 6.6 (236) 1.6 27.2 86.6 14.3 (215) 45.4 100 2,3,4,5,6 C,L I,D I,O
M. volans 1 8.0 1.3 (33) 5.3 10.7 92.3 7.6 (27) 75.6 100 3 C D I

nonreproductive, normothermic bats. No significant differences
were observed between roosts of hypothermic bats and those of
nonreproductive bats (U ¼ 273, P ¼ 0.542) and reproductive
females (U ¼ 20, P ¼ 0.250), probably because of the few
records (2) from hypothermic bats.
Cluster analysis using temperature data (Fig. 2) yielded 2
main groups: species that were observed in a wide roost
temperature range (.108C), and species that were found in
a narrower range of roost temperatures (,108C). Within the
1st group, 5 homeothermic insectivores belonging to 3 taxa
(Balantiopteryx plicata, Pteronotus parnellii, Mormoops
megalophylla, Macrotus waterhousii, and Micronycteris
megalotis) and Leptonycteris curasoae, a pollen-nectarivore
(species feeding primarily on nectar and secondarily on
pollen), formed a unique subgroup distinguished by the use
of hot roosts (X  temperatures between 26 and 328C).
Frugivores (Carollia perspicillata and Artibeus jamaicensis),
other pollen-nectarivores (Choeronycteris mexicana and
Glossophaga soricina), 1 sanguivore (Desmodus rotundus),
and 2 heterothermic insectivores (Myotis keaysi and Coryno-
rhinus townsendii) formed another subgroup using cooler
roosts (X ,228C); Myotis velifer joins this group, but
similarity to other species is low because it uses relatively
 : 11.68C) and was observed in the widest range
cold roosts (X
of roost temperatures (25.68C). Within the group of species
observed in a narrow range of roost temperatures, 2 of the
smallest homeothermic insectivorous species (Pteronotus
davyi and P. personatus) formed a subgroup at the highest
roost temperatures (X ; 358C); another subgroup was formed
by a large nectarivore (Leptonycteris nivalis), a frugivore
(Carollia brevicauda), a vampire (Diphylla ecaudata), and an
insectivore (Natalus stramineus), all of which roosted at
intermediate temperatures (X  ; 208C). The 3rd subgroup was
formed by 3 Nearctic heterothermic insectivores (Coryno-
rhinus mexicanus, Eptesicus fuscus, and Myotis volans)
observed at the coolest roost temperatures (X  : 7–138C).
Log body mass and log roost temperature were positively
correlated (r ¼ 0.132, P ¼ 0.009; Fig. 3). Correlation between
these variables was significant and negative as expected when
homeotherms and heterotherms were evaluated separately (r ¼

0.344, P , 0.001 for homeotherms; r ¼
0.426, P , 0.001
for heterotherms; Fig. 3). Within the homeothermic group, type
2 species, hypothesized to be the most sensitive to roost
temperature, showed a significant correlation between roost
temperature and body size (r ¼
0.559, P , 0.001); type 3
species, hypothesized to be less sensitive to changes in ambient
temperature, showed no correlation (r ¼
0.132, P ¼ 0.125;
Fig. 3). This indicates that body mass is not strongly related to
roost temperature for all species of bats, but is an important
factor for labile homeotherms. Roost temperature trends were
also evaluated within trophic, thermoregulatory, and taxonomic
groups of species (Fig. 4). Kruskall–Wallis tests on each
variable yielded significant differences at least between 1 of the
pairs of groups (for general food habit categories: v2 ¼ 13.2,
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS 679

TABLE 3.—Types of roosts reported for species of bats using caves in central Mexico, with additional data on night roosts from this study (see
Table 1 for generic names). Letters in parentheses refer to occasional roosts. D, day roosts; N, night roosts.

Species Caves Mines Man-made structures Trees Source

a
B. plicata D,N D D D Arroyo-Cabrales and Jones 1988
M. megalophylla D, Na D (D) Reid 1997; Rezsutek and Cameron 1993
P. davyi D, Na D,N Adams 1989; Bateman and Vaughan 1974
P. parnellii D, Na D (D) Herd 1983; Reid 1997
P. personatus D, Na D,N Bateman and Vaughan 1974; Reid 1997
M. waterhousii D, Na D,N (D),N Anderson 1969; Barbour and Davis 1969; Reid 1997
M. megalotis D D D D Alonso-Mejı́a and Medellı́n 1991
C. mexicana D D, Na (D),N (D) Arroyo-Cabrales et al. 1987; Davis and Russell 1952
G. soricina D, Na D D D Alvarez et al. 1991
L. curasoae D, Na D N Cockrum 1991; Reid 1997
L. nivalis D, Na D D D Pfrimmer
Hensley and Wilkins 1988
C. brevicauda D, Na D D D Linares 1987; Reid 1997
C. perspicillata D, Na D D D,N Cloutier and Thomas 1992
A. jamaicensis D, Na D,N D Reid 1997
D. rotundus D, Na D D,N D,N Greenhall et al. 1983; Reid 1997
D. ecaudata D, Na D (D) (D) Greenhall et al. 1984; Reid 1997; Villa
R. 1966
N. stramineus D, Na D D Moreno 1996
C. mexicanus D D, Na Tumlison 1992
C. townsendii D,N D D,N Barbour and Davis 1969; Kunz and Martin 1982
E. fuscus D, Na D,N D,N D Barbour and Davis 1969; Kurta and Baker 1990; Reid 1997
M. keaysi D, Na (D) (D) Reid 1997
M. velifer D,N D,N D,N Barbour and Davis 1969; Reid 1997
M. volans D D, Na D D Warner and Czaplewski 1984
a
Information from this study.

d.f. ¼ 3, P ¼ 0.004; for thermoregulatory groups: v2 ¼ 150.9, DISCUSSION

d.f. ¼ 2, P , 0.001; for taxonomic groups: v2 ¼ 249.1, d.f. ¼
Specificity in roost requirements.— Our results support the
8, P , 0.001). Paired multiple comparisons detected signifi-
hypothesis that most species of bats in caves of Mexico have
cant differences (P , 0.05) among all thermoregulatory cate-
little specificity (i.e., they are not obligated to exploit a narrow
gories, and in most cases for trophic and taxonomic groups,
range of environmental conditions) in roost requirements.
except between the following pairs: insectivorous–nectarivo-
However, we did not measure preference (i.e., optimal condi-
rous and sanguivorous–frugivorous trophic groups; and
tions sought by individuals regardless of degree of tolerance).
Glossophaginae–Stenodermatinae, Glossophaginae–Natalidae,
Only 8 of the 23 species studied used roosts showing tem-
Carolliinae–Stenodermatinae, Carolliinae–Desmodontinae,
perature variations less than 108C, whereas only 5 species were
Carolliinae–Natalidae, Stenodermatinae–Desmodontinae, Sten-
observed within ranges of relative humidity less than 20%.
odermatinae–Natalidae, and Desmodontinae–Natalidae. These
These relatively narrow ranges of roost temperature and
results indicate that fruit, pollen-nectar, and blood feeders used
humidity were found in species from which we obtained
very similar roosting places.
few data, such as Leptonycteris nivalis and Eptesicus fuscus.
In general (Figs. 2–4), small heterothermic insectivores oc-
Only Pteronotus davyi and P. personatus showed some degree
cupy a wide range of temperatures and the coldest sites; small
of thermal specificity. The small Pteronotus (P. davyi, P.
homeothermic insectivores usually occupy the warmest roosts;
personatus, and P. quadridens) usually are reported from very
and nectarivorous, frugivorous, and sanguivorous species usu-
hot caves (33.0–36.08C—Aguilar and Ruiz 1995; Bonaccorso
ally occupy sites with cool to moderate temperatures, but also
et al. 1992; Rodriguez–Durán 1995; this study), with 278C
use warm roosts; and nectarivores used both very hot and very
being the lowest recorded temperature (Arita and Vargas 1995;
cold roosts.
McNab 1969).
Species also showed high variability in relative humidity of
Torpid and reproductive bats tend to have a relatively nar-
roosts (Table 2). Although there was a significant correlation
row microclimatic specificity. The significant association ob-
between roost humidity and body mass (r ¼ 0.302, P , 0.001;
and r ¼ 0.301, P , 0.001, arcsin and log transformations of served between roost temperature and physiological state of
humidity), the slope was positive and the trend was contrary to individuals, regardless of species, suggests that this spe-
our predictions. This pattern and the resulting cluster analysis cificity is a reflection of individual energetic demands rather
for roost humidity (Fig. 5) seem to reflect the humidity level of than species-specific requirements. Pteronotus parnellii, M.
the external environment. Thus, small species living in hot, dry megalophylla, and L. curasoae seem to depend on hot caves
habitats occupied a wide range of humidity levels, compared to during lactation; they were observed in wide ranges of
the narrower humidity range of large bat species found mainly temperature (.158C) and humidity (.50%), but only used
in humid environments. the hottest cave as maternity roost. In all 3 cases, maternity
680 JOURNAL OF MAMMALOGY
FIG. 1.—Box plots of roost temperatures at colonies of 23 cave-
dwelling bat species from central Mexico grouped according to
apparent physiological state of individuals as observed in the field.
Horizontal line within box represents median score, and box encloses
50% of observations around the median. Upper and lower vertical
lines represent the highest and lowest values that are not outliers or
extreme scores. Sample size is indicated above each box. Re-
productive bats include pregnant and lactating females; basal-state bats
are nonreproductive, normothermic bats (active, warm-bodied bats);
torpid bats are bats observed inactive and cold at touch (include
hibernating bats); hypothermic bats defined as above but correspond-
ing to typical homeothermic species.
roost temperature was 33.5–37.08C and humidity ranged
between 45% and 90%, consistent with data from other studies
(Aguilar and Ruiz 1995; Arends et al. 1995; Bonaccorso et al.
1992; Fleming et al. 1998; Riechers et al. 1998; C. Chávez-
Tovar and G. Ceballos pers. comm.; Program for Conservation
of Mexican Bats [PCMM], in litt.).
Temporal specificity in roost use has been reported for many
Holarctic species, mainly vespertilionids, which typically show
2 well defined stages: the winter phase, when both males and
females roost in cold sites to hibernate, and summer phase,
when females use warm roosts to rear young, while most males
roost in cooler places (Kunz 1982). In this study, only M.
velifer differentially used both kinds of roost, similar to reports
from more temperate regions (Fitch et al. 1981; Kunz 1973;
Twente 1955). However, we found a nonhibernating, year-
round resident colony using the same cave as a maternity roost,
suggesting that the 2-phase pattern is not typical of the species.
Rock type, perch surfaces, cave size, and degree of lighting
seem to have minimal direct effects on roost site selection
(Table 2), as has been described elsewhere (see Table 3).
However, these parameters can become very important as they
affect microclimate (Tuttle and Stevenson 1981). Microstruc-
tural heterogeneity probably favors the establishment of
multispecific assemblages in single caves, as has been recently
documented in dissolved sandstone caves from northern
Puebla, Mexico (Brunet and Medellı́n 2001).
Vol. 85, No. 4
FIG. 2.—Similarity in roost temperature among 23 cave-dwelling
bat species from central Mexico. Parameters included in the analysis
are mean and range of roost temperature. Species are coded according
to ecophysiological guilds: HoI ¼ homeothermic insectivores; HtI ¼
heterothermic insectivores; NP ¼ pollen–nectarivores F ¼ frugivores;
S ¼ sanguivores.
Most cases of roost specificity are related to optimization of
several physiological processes for which temperature plays
a fundamental role. Parturition, lactation, and hibernation are
critical phases for bats (Gittleman 1988; Kunz 1973; Kurta et
al. 1990; McNab 1974; Richter et al. 1993; Speakman and
Racey 1987; Studier et al. 1973; Tuttle 1975). Because high-
latitude species face both the coldest temperatures and most
pronounced food shortage during winter, and because warm
roosts necessary for successful reproduction are not common
during summer in many of these regions (Tuttle and Stevenson
1981), Holartic species should display some degree of seasonal
specificity. This might be why temperature has been identified
as the single most important physical factor governing roost
selection by bats in temperate and cold regions (e.g., Betts
1997; Callahan et al. 1997; Hamilton and Barclay 1994;
Kalcounis and Brigham 1998; Kunz 1973; McNab 1974;
Ormsbee and McComb 1998; Rabe et al. 1998; Richter et al.
1993; Rydell 1990; Vonhof and Barclay 1996). Roost tem-
perature also might be critical for subtropical and tropical spe-
cies (Arita and Vargas 1995; Bonaccorso et al. 1992; Fleming
et al. 1998; McCracken and Gustin 1991; Rodrı́guez-Durán
1995; this study).
Correlates to body size, diet, taxonomic affiliation, and
thermoregulatory pattern.— Strict specificity in use of cave
roosts seems to be uncommon among bats, and only some
species show some degree of seasonal specificity, particularly
in relation to roost temperature. Within an annual cycle, the
ranges of temperature and humidity tolerated by each species in
roosts are usually quite broad. Although bat microclimatic
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS 681

FIG. 3.—Correlation between mean body mass and roost temper-

ature for 23 bat species from central Mexico. Points represent average
seasonal temperatures for sites occupied by bat colonies. Three fitted
lines were obtained for roost temperature according to thermoregu-
latory patterns of species: a) type 0 species (heterotherms), b) type 2
species (labile homeotherms), and c) type 3 species (more precise
homeotherms).

plasticity must have limits, factors determining natural limits of

roost temperature for any given species are unclear. Our data
show that body size, diet, taxonomic affinity, and thermoreg-
ulatory pattern (traits associated with thermoregulatory ability)
are related in some degree to roost temperature. Although each
of these traits shows statistically significant differences among
categories, no single factor determined strongly the temperature
range that any species tolerates within roosts. Considering that
such characters are not independent from one another (McNab
1992), we combined characteristics of each species to explain
the observed patterns. Three of these characteristics (size,
feeding habits, and taxonomic group) interact through
metabolic rate to set the thermoregulatory capability of a species
(Elgar and Harvey 1987; Hayssen and Lacy 1985; McNab

!
FIG. 4.—Box plots of roost temperatures at colonies of 23 cave-
dwelling bat species from central Mexico, grouped according to
general food habits, pattern of thermoregulation, and taxonomic
group. Box plots as in Fig. 1. Sample size indicated above each box.
Categories for thermoregulatory patterns: 0, heterotherms capable of
entering torpor; 2, homeotherms that maintain relatively wide ranges
of body temperature under a variety of air temperatures (labile
homeotherms); 3, homeotherms that maintain their body temperature
in a narrow range under a variety of air temperatures (precise
homeotherms). Categories for main source of food: insects, pollen–
nectar, fruit, and blood. Taxonomic groups: Emb, Emballonuridae;
Mor, Mormoopidae; Phy, Phyllostominae; Glo, Glossophaginae; Car,
Carolliinae; Ste, Stenodermatinae; Des, Desmodontinae; Nat, Natali-
dae, Ves, Vespertilionidae.
682 JOURNAL OF MAMMALOGY
FIG. 5.—Similarity in roost humidity among 23 cave-dwelling bat
species from central Mexico. Parameters included in the analysis are
mean and range of roost humidity. Abbreviations as for Fig. 2.
1969, 1982, 1992, 1997), whereas thermoregulatory pattern is
only an arbitrary classification of this ability.
Species able to enter torpor (pattern type 0) occupied the
widest temperature range and the coldest roosts, from near 08C
to almost 308C. Adding some extreme temperature data for type
0 species from the literature (e.g., Gaisler 1970; Harmata 1973;
Henshaw and Folk 1966; Licht and Leitner 1967; Vaughan and
O’Shea 1976), that range would increase at least by 208C (from
under
58C to at least 458C), which comprises the entire range
of temperatures observed among bat species. Within homeo-
therms, those species with precise thermoregulatory capabilities
(pattern type 3) were observed in the lowest temperatures and in
a wider temperature range (14.58C–378C). Species with a more
labile thermoregulatory capacity (pattern type 2) occupied
a slightly narrower range of temperatures (15.9–37.58C). The
difference between these 2 groups was statistically significant,
perhaps because type 3 species tend to use colder roosts (below
208C) than type 2 species. If vampires (that indeed seem to be
precise thermoregulators) were excluded from group 2, the
difference between these groups would be even greater. Thus,
the general pattern exhibited by each group would be much
clearer, possibly as follows: species in group 0 use the more
diverse roosts, from extreme temperatures below 08C to near
408C; group 1 species only use warm sites (between 308C and
408C, comprising the thermoneutral zone of many species);
group 2 occupies sites between 208 and 408C; and group 3 uses
roosts with temperatures ranging from 108C to 408C. The lower
limit of 108C for homeothermic species is suggested by many
studies in captivity (see data compiled by Lyman 1970;
McManus 1977; McNab 1969) and by the scarce empirical
data collected in the field (López-Wilchis 1999; McNab 1969;
Vargas 1998; Villa-R. 1966).
Vol. 85, No. 4
Because thermoregulatory patterns at the species level are an
outcome of the interaction of characteristics such as body size,
type of diet, and taxonomic affinity (which are known or can be
inferred for virtually all species), the combination of these
variables could have practical predictive value. Such variables
yielded more diffuse and complex patterns when they were
treated separately. Body size alone, the primary factor
determining basal metabolic rate (McNab 1969, 1992), was
not correlated with roost temperature, but when diet, taxonomic
affinity, and thermoregulatory performance are incorporated
into this relationship, the pattern changes. The almost null
correlation between body size and roost temperature is
accentuated by the high dispersion of points along the y axis,
which results from the great variation of roost temperatures for
each species (Fig. 3). Assuming that maximum roost temper-
atures tolerated are similar for all species (slightly above the
thermoneutral zone, i.e., around 408C), it would be more
appropriate to fit a line employing only the minimum roost
temperatures used by species. This approach is more useful for
homeothermic species, because heterotherms could theoreti-
cally occupy very similar minimum roost temperatures irre-
spective of size (but see McNab 1974 for preferred minimum
roost temperatures as a function of size in vespertilionids).
The slope of this line could approximate a line that sets the
lower limit of temperature, as a function of size, that any homeo-
thermic bat can tolerate while roosting; similar to the relation-
ship between body size and basal metabolic rate (McNab 1969,
1982), the residual variation could be explained mainly by feed-
ing habits and taxonomic affiliation. Although our 1st approach
to this relationship showed a negative relationship, where in gen-
eral large species were found at lower minimum temperatures,
more data are needed.
Relationships between feeding habits per se with roost
temperature point out that frugivores, pollen–nectarivores, and
sanguivores frequently can occupy minimum temperatures
close to 158C (lowest records: 15.08C, 14.58C and 15.98C,
respectively). Our records of medium-sized homeothermic
insectivorous P. parnelli and M. megalophylla at temperatures
near 168C are noteworthy. In the literature, roost temperatures
for these and other mormoopids generally have been reported
as above 258C (Arita and Vargas 1995; Bateman and Vaughan
1974; Bonaccorso et al. 1992; Rodrı́guez-Durán 1995), and
a minimum of 228C has been reported for P. parnelli (Vargas
1998). These 2 species and Macrotus waterhousii were the
largest insectivorous bats evaluated in our study. Data from
other sources for neotropical species confirm the trends
observed here for each trophic group. Minimum roost temper-
atures reported for each group are: homeothermic insectivores,
18.5–278C (Arita and Vargas 1995; McNab 1969; Vargas
1998); frugivores, 11.5–198C (López-Wilchis 1999; McNab
1969; Vargas 1998); nectarivores, 12–198C (McNab 1969;
Vargas 1998; PCMM, in litt.); sanguivores, 13–16.58C (McNab
1969; Vargas 1998); and carnivores, 148C (McNab 1969).
Diets of fruit, pollen–nectar, vertebrates, and perhaps to
a lesser degree blood allow their consumers to use colder
roosts, sometimes dipping to 108C for short periods. Our data
on minimum roost temperature for homeothermic insectivores
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS
(B. plicata, M. megalophylla, P. parnelli and N. stramineus)
are lower than previous records, probably because the study
area is close to the northernmost limit for many tropical
species. In stable environments, a diet of insects probably
allows only medium and large species (such as M. mega-
lophylla and P. parnellii) to use roosts where temperature
sometimes drops to 158C. The smallest insectivores can use
roosts with temperatures as low as 208C (there are only a few
records slightly below this for B. plicata and N. stramineus),
although most species prefer temperatures above 258C (Table
2). Strict microclimatic specialization occurs among the small
homeothermic insectivores, the least efficient thermoregulators,
which select roosts with temperatures within their thermoneu-
tral zone (30–408C); the 2 smallest mormoopids are restricted
to very hot roosts (Table 2; Fig. 4).
In general, species that can acquire large quantities of
energy during feeding can employ a larger proportion of avail-
able energy to face a great variety of thermal environ-
ments, including those that are energetically expensive (roost
temperature below thermoneutral zone). Thus, fruit, nectar–
pollen, and vertebrates, available throughout the year in many
tropical regions, provide an abundant and constant energy
supply (McManus 1977; McNab 1969). High caloric value of
fruits (Dinerstein 1986) and nectar (Howell 1974) facilitate the
acquisition of a surplus of energy. Moreover, large sizes
associated with carnivory and frugivory can increase thermo-
regulatory ability by reducing surface-to-volume ratio (hence
reducing heat loss) and by making the turnover of energy stores
more efficient (Dwyer 1971; McNab 1982). Thus, species of the
latter trophic guilds show high rates of metabolic heat
production, which permit them to exploit otherwise unfavorable
roosts. Insectivores face at least 2 major problems: their size is
generally small, which makes them more vulnerable to low
temperatures; and their food source exhibits a strong seasonality
even in tropical environments, which reduces energy acquisition
(McManus 1977; McNab 1969). Torpor-capable insectivores
can exploit the coldest roosts, whereas other species must use
almost exclusively hot roosts. The case of sanguivores is more
complex. Their prey can be locally abundant, but their inability
to transport large quantities of food limits their daily energy
intake (McNab 1973); also, the usable caloric value of blood is
very low (Wimsatt 1969). Both factors explain the low meta-
bolic rates of vampires (McNab 1989), but they do not explain
the observed ability of vampires to tolerate ambient temper-
atures in roosts close to 108C (McNab 1969; PCMM, in litt.).
Microchiroptera exhibit strong relationships between taxo-
nomic affiliation and several morphological, physiological and
ecological traits of species (Cruz-Neto and Jones in press;
Freeman 2000; McNab 1986, 1992). Although phylogeny
might have a strong influence on basal metabolic rate of bats
(Cruz-Neto and Jones in press), it is difficult to separate its
individual effects upon other traits such as thermoregulatory
pattern or roost use. However, vespertilionids showed the only
pattern clearly determined by phylogenetic inertia. In the
Americas, there are many small insectivorous species, but
vespertilionids are the only group that can enter deep torpor
(Hill and Smith 1984; McNab 1982), including at least some of
683
its tropical members (e.g., Studier and Wilson 1970); our
observations on Myotis keaysi in the field and laboratory
support this view (it is capable of entering and arousing from
deep torpor at 38C). The use of torpor allows vespertilionids to
use a variety of roosting conditions, and to range around the
world, exploiting nearly all terrestrial biomes and environments.
No clear pattern regarding roost humidity was found. The
observed trend in roost humidity can reflect the aridity of
the external environment. Species observed in zones of xero-
phytic scrub and/or tropical deciduous forest were also those
observed in the driest roosts and in the widest ranges of
relative humidity, whereas species found in humid roosts in-
habit tropical evergreen forests, subtropical cloud forests, and
coniferous forests. Alternatively, bats roosting in hot caves
could select sites with low to moderate levels of humidity to
facilitate pulmocutaneous water loss (Baudinette et al. 2000;
Bonaccorso et al. 1992). Variation in roost humidity is
positively correlated with species richness within some tropical
caves, which suggests that water balance might be a significant
factor for bats in warm environments and could partially
determine species richness of cave bats (Arita 1993; Brunet and
Medellı́n 2001). Our results, however, suggest that roost use
is limited more by temperature than by relative humidity.
Structural characteristics of roosts might be important because
they influence microclimatic parameters. Our findings support
the hypothesis that temperature is the most important and
restrictive roost parameter for bats.
RESUMEN
El uso de refugios por los murciélagos probablemente refleja
las capacidades de termorregulación y de balance hı́drico
propias de cada especie. Para probar esta hipótesis, exploramos
la relación que guardan cuatro atributos de las especies
asociados con tales capacidades fisiológicas (tamaño corporal,
tipo general de alimentación, grupo taxonómico y patrón de
termorregulación) con el microclima (temperatura y humedad
relativa) de los sitios utilizados como refugio. Recopilamos
datos de microclima y registramos la presencia de especies en
18 cuevas del centro de México situadas en 5 diferentes tipos de
ecosistema. Las 23 especies encontradas mostraron poca
evidencia de especificidad microclimática, aunque en general
utilizaron refugios más calientes durante los perı́odos de
maternidad, y refugios más frı́os para hibernar. Los hetero-
termos en su conjunto (Vespertilionidae) ocuparon las cuevas
más frı́as y el más amplio intervalo de temperatura (1.6–
29.88C), mientras que los homeotermos (Emballonuridae,
Mormoopidae, Phyllostomidae, y Natalidae) ocuparon refugios
mas cálidos (14.5–37.58C). Dentro de este grupo, los homeo-
termos más precisos (temperatura corporal más estable)
ocuparon sitios con temperaturas ligeramente más bajas que
los homeotermos menos eficientes (temperatura corporal más
variable). El tamaño corporal por sı́ solo no definió patrones
claros de uso de cuevas, pero cuando analizamos por separado
homeotermos y heterotermos encontramos una relación
negativa entre tamaño corporal y temperatura del refugio.
Las especies insectı́voras más pequeñas (,10 g) ocuparon
684 JOURNAL OF MAMMALOGY
consistentemente refugios .208C (generalmente .258C); sólo
los insectı́voros homeotermos más grandes fueron observados
a 168C. Los frugı́voros, nectarı́voros, y hematófagos fueron
observados en un amplio intervalo de temperatura (14.5–378C),
frecuentemente ,208C. La humedad de los refugios fue
altamente variable para la mayorı́a de las especies, y no se
observó ninguna tendencia respecto a este factor. La capacidad
termorregulatoria (resultante de la compleja interacción de
tamaño corporal, tipo de alimento y afinidad taxonómica)
parece limitar el tipo de refugios que cada especie puede
explotar. Nuestros resultados apoyan la hipótesis de que la
temperatura es el factor fı́sico más importante para los
murciélagos en la selección de refugios.
ACKNOWLEDGMENTS
We thank Fondo Mexicano para la Conservación de la Naturaleza
(FMCN), United States Fish and Wildlife Service, National Fish and
Wildlife Foundation, CONACYT, and Bat Conservation International,
for financial support. We thank all those who helped in the field,
especially M. Gutiérrez, G. Téllez, and C. Rodrı́guez. C. Chávez and
M. Friendly provided statistical advice, and J. Vargas, J. Ortega, M. B.
Fenton, T. H. Kunz, and 3 anonymous reviewers provided helpful
comments on the manuscript. This paper was submitted while RAM
was on sabbatical leave at Center for Environmental Research and
Conservation, Columbia University, and is a contribution of the
Program for Conservation of Mexican Bats (PCMM).
LITERATURE CITED
ADAMS, J. K. 1989. Pteronotus davyi. Mammalian Species 346:1–5.
AGUILAR, S., AND A. RUIZ. 1995. Una comunidad de murciélagos en
una ‘‘cueva de calor’’ como factor determinante del sostenimiento
de la diversidad animal cavernı́cola. B.Sc. Thesis, Facultad de
Ciencias, Universidad Nacional Autónoma de México, Mexico
City, Mexico.
ALONSO-MEJÍA, A., AND R. A. MEDELLÍN. 1991. Micronycteris
megalotis. Mammalian Species 376:1–6.
ALTRINGHAM, J. D. 1996. Bats. Biology and behavior. Oxford
University Press, Oxford, United Kingdom.
ALVAREZ, J., M. R. WILLIG, J. K. JONES, JR., AND W. D. WEBSTER.
1991. Glossophaga soricina. Mammalian Species 379:1–7.
ANDERSON, S. 1969. Macrotus waterhousii. Mammalian Species
1:1–4.
ARENDS, A., F. J. BONACCORSO, AND M. GENOUD. 1995. Basal rates of
metabolism of nectarivorous bats (Phyllostomidae) from a semiarid
thorn forest in Venezuela. Journal of Mammalogy 76:947–956.
ARITA, H. T. 1993. Conservation biology of the cave bats of Mexico.
Journal of Mammalogy 74:693–702.
ARITA, H. T., AND J. A. VARGAS. 1995. Natural history, interspecific
association, and incidence of the cave bats of Yucatan, Mexico.
Southwestern Naturalist 40:29–37.
ARROYO-CABRALES, J., R. H. HOLLANDER, AND J. K. JONES, JR. 1987.
Choeronycteris mexicana. Mammalian Species 291:1–5.
ARROYO-CABRALES, J., AND J. K. JONES, JR. 1988. Balantiopteryx
plicata. Mammalian Species 301:1–4.
AUDET, D., AND D. W. THOMAS. 1997. Facultative hypothermia as
a thermoregulatory strategy in the phyllostomid bats, Carollia
perspicillata and Sturnira lilium. Journal of Comparative Physiol-
ogy, B: Biochemical, Systematic, and Environmental Physiology
167:146–152.
Vol. 85, No. 4
BARBOUR, R. W., AND W. H. DAVIS. 1969. Bats of America. The
University Press of Kentucky, Lexington.
BATEMAN, G. C., AND T. A. VAUGHAN. 1974. Nightly activities of
mormoopid bats. Journal of Mammalogy 55:45–65.
BAUDINETTE, R. V., S. K. CHURCHILL, K. A. CHRISTIAN, J. E. NELSON,
AND P. J. HUDSON. 2000. Energy, water balance and the roost
microenvironment in three Australian cave-dwelling bats (Micro-
chiroptera). Journal of Comparative Physiology, B: Biochemical,
Systematic, and Environmental Physiology 170:439–446.
BELL, G. P., G. A. BARTHOLOMEW, AND K. A. NAGY. 1986. The roles of
energetics, water economy, foraging behavior, and geothermal
refugia in the distribution of the bat, Macrotus californicus. Journal
of Comparative Physiology, B: Biochemical, Systematic, and
Environmental Physiology 156:441–450.
BETTS, B. J. 1997. Microclimate in Hell’s Canyon mines used by
maternity colonies of Myotis yumanensis. Journal of Mammalogy
78:1240–1250.
BONACCORSO, F. J., A. ARENDS, M. GENOUD, D. CANTONI, AND T.
MORTON. 1992. Thermal ecology of moustached and ghost-faced
bats (Mormoopidae) in Venezuela. Journal of Mammalogy 73:
365–378.
BRUNET, A. K., AND MEDELLÍN, R. A. 2001. The species-area
relationship in bat assemblages of tropical caves. Journal of
Mammalogy 82:1114–1122.
CALLAHAN, E. V., R. D. DROBNEY, AND R. L. CLAWSON. 1997.
Selection of summer roosting sites by Indiana bats (Myotis sodalis)
in Missouri. Journal of Mammalogy 78:818–825.
CARPENTER, R. E., AND J. B. GRAHAM. 1967. Physiological responses
to temperature in the long-nosed bat, Leptonycteris sanborni.
Comparative Biochemistry and Physiology 22:709–722.
CLOUTIER, D., AND D. W. THOMAS. 1992. Carollia perspicillata.
Mammalian Species, 417:1–9.
COCKRUM, E. L. 1991. Seasonal distribution of northwestern
populations of the long-nosed bats, Leptonycteris sanborni family
Phyllostomidae. Anales del Instituto de Biologı́a, Universidad
Nacional Autónoma de México, Serie Zoologı́a 62:181–202.
CRUZ-NETO, A. P., AND K. E. JONES. In press. Exploring the evolution
of the basal rate of metabolism in bats. In: Functional and
evolutionary ecology of bats (A. Zubaid, G. F. McCracken and T. H.
Kunz, eds.). Oxford University Press, Oxford, United Kingdom.
DAVIS, W. B., AND R. J. RUSSELL, JR. 1952. Bats of the Mexican State
of Morelos. Journal of Mammalogy 33:234–239.
DEGEN, A. A. 1997. Ecophysiology of small desert mammals.
Springer-Verlag, Berlin, Germany.
DINERSTEIN, E. 1986. Reproductive ecology of fruit bats and the
seasonality of fruit production in a Costa Rican cloud forest.
Biotropica 18:307–318.
DWYER, P. D. 1971. Temperature regulation and cave-dwelling in bats:
an evolutionary perspective. Mammalia 35:424–455.
ELGAR, M. A., AND P. H. HARVEY. 1987. Basal metabolic rates in
mammals: allometry, phylogeny and ecology. Functional Ecology
1:25–36.
FENTON, M. B. 2001. Bats. 2nd ed. Checkmark Books, New York.
FITCH, J. H., K. A. SHUMP, JR., AND A. U. SHUMP. 1981. Myotis velifer.
Mammalian Species 149:1–5.
FLEMING, T. H., A. A. NELSON, AND V. M. DALTON. 1998. Roosting
behavior of the lesser long-nosed bat, Leptonycteris curasoae.
Journal of Mammalogy 79:147–155.
FREEMAN, P. W. 2000. Macroevolution in Microchiroptera: recoupling
morphology and ecology with phylogeny. Evolutionary Ecology
Research 2:317–335.
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS
GAISLER, J. 1970. Remarks on the thermopreferendum of Palearctic
bats in their natural habitats. Bijdragen Tot De Dierkunde, 40:
33–35.
GAUR, B. S. 1980. Roosting ecology of the Indian desert rat-tailed bat,
Rhinopoma kinneari Wroughton. Pp. 125–128 in Proceedings of the
Fifth International Bat Research Conference (D. E. Wilson and A.
L. Gardner, eds.). Texas Tech Press, Lubbock.
GENOUD, M., AND F. J. BONACCORSO. 1986. Temperature regulation,
rate of metabolism, and roost temperature in the greater white-lined
bat, Saccopteryx bilineata (Emballonuridae). Physiological Zoology
59:49–64.
GITTLEMAN, J. L. 1988. Energy allocation in mammalian reproduction.
American Zoologist 28:863–875.
GREENHALL, A. M., G. JOERMANN, AND U. SCHMIDT. 1983. Desmodus
rotundus. Mammalian Species 202:1–6.
GREENHALL, A. M., U. SCHMIDT, AND G. JOERMANN. 1984. Diphylla
ecaudata. Mammalian Species 227:1–3.
HAMILTON, I. M., AND R. M. R. BARCLAY. 1994. Patterns of daily
torpor and day-roost selection by male and female big brown bats
(Eptesicus fuscus). Canadian Journal of Zoology 72:744–749.
HARMATA, W. 1973. The thermopreferendum of some species of bats
(Chiroptera) in natural conditions. Zeszyty Naukowe Uniwersytetu
Jagiellonskiego, Prace Zoologiczne 127–141.
HAYSSEN, V., AND R. C. LACY. 1985. Basal metabolic rates in
mammals: taxonomic differences in the allometry of BMR and
body mass. Comparative Biochemistry and Physiology 81A:741–
754.
HENSHAW, R. E., AND G. E. FOLK, JR. 1966. Relation of thermoreg-
ulation to seasonally changing micoclimate in 2 species of bats
(Myotis lucifugus and M. sodalis). Physiological Zoology 39:223–
236.
HERD, R. M. 1983. Pteronotus parnelli. Mammalian Species 209:1–5.
HERREID, C. F., II, AND K. SCHMIDT-NIELSEN. 1966. Oxygen con-
sumption, temperature, and water loss in bats from different environ-
ments. American Journal of Physiology 211:1108–1112.
HILL, J. E., AND J. D. SMITH. 1984. Bats: a natural history. University
of Texas Press, Austin.
HOWELL, D. J. 1974. Bats and pollen: physiological aspects of the
syndrome of chiropterophily. Comparative Biochemistry and
Physiology 48A:263–276.
HUMPHREY, S. H. 1975. Nursery roosts and community diversity of
Nearctic bats. Journal of Mammalogy 56:321–346.
KALCOUNIS, M. C., AND R. M. BRIGHAM. 1998. Secondary use of aspen
cavities by tree-roosting big brown bats. Journal of Wildlife
Management 62:603–611.
KUNZ, T. H. 1973. Population studies of the cave bat (Myotis velifer):
reproduction, growth, and development. Occasional Papers of the
Museum of Natural History, University of Kansas 15:1–43.
KUNZ, T. H. 1982. Roosting ecology of bats. Pp. 1–55 in Ecology of
bats (T. H. Kunz, ed.). Plenum Press, New York.
KUNZ, T. H., AND R. A. MARTIN. 1982. Plecotus townsendii.
Mammalian Species 175:1–6.
KURTA, A., AND R. H. BAKER. 1990. Eptesicus fuscus. Mammalian
Species 356:1–10.
KURTA, A., T. H. KUNZ, AND K. A. NAGY. 1990. Energetics and water
flux of free-ranging big brown bats (Eptesicus fuscus) during
pregnancy and lactation. Journal of Mammalogy 71:59–65.
LICHT, P., AND P. LEITNER. 1967. Behavioral responses to high
temperatures in three species of California bats. Journal of
Mammalogy 48:52–61.
LINARES, O. J. 1987. Murciélagos de Venezuela. Cuadernos Lagoven,
Caracas, Venezuela.
685
LÓPEZ-WILCHIS, R. 1999. Murciélagos asociados a una colonia de
Corynorhinus mexicanus G. M. Allen, 1916 (Chiroptera: Vesperti-
lionidae). Vertebrata Mexicana 5:9–16.
LYMAN, C. P. 1970. Thermoregulation and metabolism in bats. Pp.
301–330 in Biology of bats. Vol. I (W. A. Wimsatt, ed.). Academic
Press, New York.
MCCRACKEN, G. F. 1989. Cave conservation: special problems of bats.
National Speleological Society Bulletin 51:49–51.
MCCRACKEN, G. F., AND M. K. GUSTIN. 1991. Nursing behavior
in Mexican free-tailed bat maternity colonies. Ethology 89:
305–321.
MCMANUS, J. J. 1977. Thermoregulation. Pp. 281–292 in Biology of
bats of the New World family Phyllostomatidae. Part II (R. J.
Baker, J. K. Jones, Jr., and D. C. Carter, eds.). No. 13, Special
Publications, The Museum, Texas Tech University, Lubbock.
MCNAB, B. K. 1969. The economics of temperature regulation in
Neotropical bats. Comparative Biochemistry and Physiology
31:227–268.
MCNAB, B. K. 1973. Energetics and the distribution of vampires.
Journal of Mammalogy 54:131–144.
MCNAB, B. K. 1974. The behavior of temperate cave bats in
a subtropical environment. Ecology 55:943–958.
MCNAB, B. K. 1982. Evolutionary alternatives in the physiological
ecology of bats. Pp. 151–200 in Ecology of bats (T. H. Kunz, ed.).
Plenum Press, New York.
MCNAB, B. K. 1986. The influence of food habits on the energetics of
eutherian mammals. Ecological Monographs 56:1–19.
MCNAB, B. K. 1989. Temperature regulation and rate of metabolism in
three Bornean bats. Journal of Mammalogy 70:153–161.
MCNAB, B. K. 1992. A statistical analysis of mammalian rates of
metabolism. Functional Ecology 6:672–679.
MCNAB, B. K. 1997. On the utility of uniformity in the definition of
basal rate of metabolism. Physiological Zoology 70:718–720.
MEDELLÍN, R. A., H. T. ARITA, AND O. SÁNCHEZ-H. 1997. Identificación
de los murciélagos de México: clave de campo. Publicaciones
Especiales de la Asociación Mexicana de Mastozoologı́a, Mexico
City, Mexico.
MORENO, A. 1996. Murciélagos de Nuevo León. Impresora Monterrey,
Monterrey, Mexico.
MORRISON, D. W. 1980. Foraging and day-roosting dynamics of
canopy fruit bats in Panama. Journal of Mammalogy 61:20–29.
NOWAK, R. M. 1994. Walker’s bats of the world. The Johns Hopkins
University Press, Baltimore, Maryland.
NÚÑEZ-JIMÉNEZ, A., N. VIÑA, M. ACEVEDO, J. MATEO, M. ITURRALDE,
AND A. GRAÑA. 1988. Cuevas y carsos. Editorial Cientı́fico–
Técnica, La Habana, Cuba.
ORMSBEE, P. C., AND W. C. MCCOMB. 1998. Selection of day roosts by
female long-legged myotis in the central Oregon Cascade Range.
Journal of Wildlife Management 62:596–603.
ORTEGA, J., AND H. T. ARITA. 1998. Neotropical–Nearctic limits in
Middle America as determined by distribution of bats. Journal of
Mammalogy 79:772–783.
PFRIMMER-HENSLEY, A., AND K. T. WILKINS. 1988. Leptonycteris
nivalis. Mammalian Species 307:1–4.
RABE, M. J., T. E. MORRELL, H. GREEN, J. C. DEVOS, JR., AND C. R.
MILLER. 1998. Characteristics of ponderosa pine snag roosts used
by reproductive bats in northern Arizona. Journal of Wildlife
Management 62:612–621.
REID, F. A. 1997. A field guide to the mammals of Central America
and southeast Mexico. Oxford University Press, New York.
REZSUTEK, M. AND G. N. CAMERON. 1993. Mormoops megalophylla.
Mammalian Species 448:1–5.
686 JOURNAL OF MAMMALOGY
RICHTER, A. R., S. R. HUMPHREY, J. B. COPE, AND V. BRACK, JR. 1993.
Modified cave entrances: thermal effect on body mass and resulting
decline of endangered Indiana bats (Myotis sodalis). Conservation
Biology 7:407–415.
RIECHERS, A., M. MARTÍNEZ-CORONEL, AND S. GAONA. 1998. Hábitos
alimentarios de una colonia de maternidad de Leptonycteris
curasoae de la Cueva ‘‘Los Laguitos,’’ Chiapas. Memoires, IV
Congreso Nacional de Mastozoologı́a, Asociación Mexicana de
Mastozoologı́a, Xalapa, Mexico.
RIEDESEL, M. L., AND B. A. WILLIAMS. 1976. Continuous 24-hour
oxygen consumption studies of Myotis velifer. Comparative
Biochemistry and Physiology 54A:95–99.
RODRÍGUEZ-DURÁN, A. 1995. Metabolic rates and thermal conductance
in 4 species of Neotropical bats roosting in hot caves. Comparative
Biochemistry and Physiology, A: Sensory, Neural, and Behavioral
Physiology 110:347–355.
RYDELL, J. 1990. Ecology of the northern bat Eptesicus nilssoni during
pregnancy and lactation. Ph.D. Dissertation, Lund University,
Lund, Sweden.
RZEDOWSKI, J. 1978. Vegetación de México. Limusa, Mexico City,
Mexico.
SAS INSTITUTE INC. 1995. JMP statistics and graphics guide. Version
3.1 of JMP. SAS Institute Inc., Cary, North Carolina.
SILVA, G. 1979. Los murciélagos de Cuba. Editorial Academia, La
Habana, Cuba.
SPEAKMAN, J. R., AND P. A. RACEY. 1987. The energetics of pregnancy
and lactation in the brown long-eared bat, Plecotus auritus. Pp.
367–393 in Recent advances on the study of bats (M. B. Fenton, P.
A. Racey, and J. M. V. Rayner, eds.). Cambridge University Press,
Cambridge, United Kingdom.
STEBBINGS, R. E. 1995. Why should bats be protected? A challenge for
conservation. Biological Journal of the Linnean Society 56:103–
118.
STUDIER, E. H. 1970. Evaporative water loss in bats. Comparative
Biochemistry and Physiology 35:935–943.
STUDIER, E. H., V. L. LYSENGEN, AND M. J. O’FARRELL. 1973. Biology
of Myotis thysanodes and M. lucifugus (Chiroptera: Vespertilioni-
dae)—II. Bioenergetics of pregnancy and lactation. Comparative
Biochemistry and Physiology 44A:467–471.
STUDIER, E. H., AND D. E. WILSON. 1970. Thermoregulation in some
Neotropical bats. Comparative Biochemistry and Physiology
34:251–262.
TUMLISON, R. 1992. Plecotus mexicanus. Mammalian Species 401:
1–3.
TUTTLE, M. D. 1975. Population ecology of the gray bat (Myotis
grisescens): factors influencing early growth and development.
Occasional Papers of the Museum of Natural History, University of
Kansas 36:1–24.
TUTTLE, M. D., AND D. E. STEVENSON. 1981. Variation in the cave
environment and its biological implications. Pp. 46–59 in Cave
gating, a handbook. 2nd ed. (R. Stitt, ed.). National Speleological
Society, Huntsville, Texas.
TWENTE, J. W., JR. 1955. Some aspects of habitat selection and other
behavior of cavern-dwelling bats. Ecology 36:706–732.
VARGAS, J. A. 1998. Factores microclimáticos y selección del refugio
diurno por murciélagos cavernı́colas en Gómez Farı́as, Tamaulipas.
M.Sc. Thesis, Facultad de Ciencias, Universidad Nacional Autón-
oma de México, Mexico City.
VAUGHAN, T. A., AND T. J. O’SHEA. 1976. Roosting ecology of the
pallid bat, Antrozous pallidus. Journal of Mammalogy 57:19–42.
VILLA-R., B. 1966. Los murciélagos de Mexico. Instituto de Biologı́a,
Universidad Nacional Autónoma de México, Mexico City.
Vol. 85, No. 4
VONHOF, M. J., AND R. M. R. BARCLAY. 1996. Roost-site selection and
roosting ecology of forest-dwelling bats in southern British
Columbia. Canadian Journal of Zoology 74:1797–1805.
WARNER, R. M., AND N. J. CZAPLEWSKI. 1984. Myotis volans.
Mammalian Species 224:1–4.
WIMSATT, W. A. 1969. Transient behavior, nocturnal activity patterns,
and feeding efficiency of vampire bats (Desmodus rotundus) under
nocturnal conditions. Journal of Mammalogy 50:233–245.
Submitted 28 February 2003. Accepted 23 September 2003.
Associate Editor was Thomas J. O’Shea.
APPENDIX I
Geographical, physical and ecological values for 18 caves examined
in this study. All caves are in central Mexico, in the states of Mexico
and Puebla. Data for each cave include (in the following order) name,
location (includes township or municipio), coordinates, altitude,
physiographic region, surrounding vegetation, type of rock, total
length (all tunnels), annual mean temperature, and annual mean
humidity. Caves are numbered 1 to 18.
Caves examined: 1), 2), and 3) Caidenı́ I, Caidenı́ II, and Caidenı́
III (entrances located within a radius of 300 m). 1 km NE Presa de
Iturbide, township of Isidro Fabela, Estado de Mexico; 198339020N,
998289110W, 3,480 m; Transversal Volcanic Belt; coniferous forest;
sandstone; 41 m, 44 m, 30 m, respectively; 8.3 6 1.18C, 5.9 6 2.38C,
8.0 6 0.98C, respectively; 87.2 6 8.5%, 65.6 6 14.8%, 86.6 6 7.2%,
respectively. 4) El Charro. 0.2 km E Zapotitlán de Méndez, township
of Zapotitlán de Méndez, Puebla; 208009000N, 978429260W, 670 m;
Sierra Madre Oriental; subtropical cloud forest; limestone; 30 m;
21.0 6 4.58C; 93.8 6 9.7%. 5) El Jardı́n. Rancho El Jardı́n, 4 km
NW Huehuetlán El Chico, township of Huehuetlán El Chico, Puebla;
188239520N, 988439100W, 850 m; Balsas River Basin; tropical
deciduous forest; conglomerate; 14 m; 26.4 6 3.08C; 56.1 6
21.5%. 6) Huiluapan. 7 km W Rı́o Frı́o, township of Ixtapaluca,
Estado de Mexico; 198219350N, 988449360W, 3,385 m; Transversal
Volcanic Belt; coniferous forests; sandstone; 76 m; 7.7 6 1.88C; 91.5
6 8.3%. 7) Karmida. 0.3 km N Zapotitlán de Méndez, township of
Zapotitlán de Méndez, Puebla; 208009190N, 978429530W, 670 m;
Sierra Madre Oriental; subtropical cloud forest; sludge stone and
limestone; 67 m; 21.9 6 1.38C; 95.3 6 6.0%. 8) La Ahorcada.
1.5 km E Coxcatlán, township of Coxcatlán, Puebla; 188169290N,
978089350W, 1,400 m; Tehuacan Valley; xerophitic scrub; sandstone;
25 m; 26.4 6 2.38C; 48.1 6 13.9%. 9) La Garita. Rancho La Garita,
5 km NW Hueytamalco, township of Hueytamalco, Puebla;
198579580N, 978199010W, 700 m; Sierra Madre Oriental; subtropical
cloud forest; sandstone and limestone; 500 m; 19.8 6 0.88C; 98.7 6
3.0%. 10) La Organera. 4 km SE Zapotitlán Salinas, township of
Zapotitlán Salinas, Puebla; 188189450N, 978259580W, 1,400 m;
Tehuacan Valley; xerophitic scrub; sandstone and metamorphic;
15 m; 21.5 6 2.48C; 59.6 6 13.0%. 11) Las Vegas. 2.5 km S Tenam-
pulco, township of Tenampulco, Puebla; 208089540N, 978249390W,
183 m; Gulf Coastal Plain; tropical evergreen forest; sandstone; 254 m;
21.6 6 2.58C; 98.5 6 3.8%. 12) Lencho Diego I. 1.25 km NW
San José Tilapa, township of Coxcatlán, Puebla; 188109490N,
978079100W, 1,200 m; Tehuacan Valley; xerophitic scrub; sandstone
and mica; 14 m; 26.3 6 2.18C; 48.0 6 19.9%. 13) Lencho Diego II.
2.7 km NW San José Tilapa, township of Coxcatlán, Puebla;
188109420N, 978079040W, 1,250 m; Tehuacan Valley; xerophitic
scrub; sandstone and mica; 35 m; 27.9 6 1.78C; 47.5 6 14.3%. 14)
Piedra Encampanada. Rancho El Terrero, 4.5 km NW Huehuetlán El
August 2004 AVILA-FLORES AND MEDELLÍN—PATTERNS OF CAVE USE IN MEXICAN BATS 687

Chico, township of Huehuetlán El Chico, Puebla; 188239310N,
988439380W, 850 m; Balsas River Basin; tropical deciduous forest;
conglomerate; 8 m; 29.9 6 1.78C; 69.4 6 17.4%. 15) San Gabriel. 4.5
km ESE Zapotitlán Salinas, township of Zapotitlán Salinas, Puebla;
188199310N, 978259020W, 1,400 m; Tehuacan Valley; xerophitic
scrub; conglomerate; 9.5 m; 25.8 6 3.78C; 47.0 6 20.5%. 16, 17) San
Lorenzo I and San Lorenzo II (entrances located within the same large
hollow). 1.5 km WNW San Lorenzo, township of Tehuacán, Puebla;
188279510N, 978269520W, 1,740 m; Tehuacan Valley; xerophitic
scrub; sandstone; 150 m, 100 m, respectively; 13.2 6 2.78C, 17.6 6
1.68C, respectively; 79.7 6 10.8%, 86.0 6 8.1%, respectively. 18)
Tzinacanostoc. 1 km NW Jolalpan, township of Jolalpan, Puebla;
188199570N, 988539170W, 1,037 m; Balsas River Basin; tropical
deciduous forest; sandstone; 220 m; 33.2 6 2.38C; 60.3 6 12.8%.