Energy Metabolism and Evaporative Water Loss in the European Free-Tailed Bat and Hemprich’S Long-Eared Bat (Microchiroptera): Species Sympatric in the Negev Desert

944

Energy Metabolism and Evaporative Water Loss in the European Free-Tailed Bat and Hemprich’s Long-Eared Bat (Microchiroptera): Species Sympatric in the Negev Desert

Sagi Marom1 for other bats, indicating efﬁcient water conservation mecha- Carmi Korine1 nisms in the study species. Comparing thermoregulatory abil- Michał S. Wojciechowski1,* ities suggests that O. hemprichii is better adapted to hot, arid Christopher R. Tracy2 environments than T. teniotis, which may explain its wider Berry Pinshow1,† desert distribution. By both standard and phylogenetically in- 1Mitrani Department of Desert Ecology, Jacob Blaustein formed ANCOVA, we found no differences in basal metabolic Institutes for Desert Research, Ben-Gurion University of the rate (BMR) between desert and nondesert species of insectiv- Negev, 84990 Midreshet Ben-Gurion, Israel; 2School of orous bats, substantiating previous studies suggesting that low Science and Primary Industry, Charles Darwin University, BMR is a characteristic common to insectivorous bats in Darwin, Northern Territory 0909, Australia general.

Accepted 3/15/2006; Electronically Published 8/17/2006

Introduction ABSTRACT It is generally accepted that small endothermic mammals with We compared the thermoregulatory abilities of two insectivo- high surface-to-volume ratios have physiological adaptations to rous bat species, Tadarida teniotis (mean body mass 32 g) and cope with problems of thermoregulation and evaporative water Otonycteris hemprichii (mean body mass 25 g), that are of dif- loss imposed by the desert environment (Schmidt-Nielsen 1964). ferent phylogenetic origins and zoogeographic distributions but These problems are exacerbated in bats because they have much are sympatric in the Negev Desert. At night, both were nor- higher surface-to-volume ratios than other mammals of equal mothermic. By day, both were torpid when exposed to ambient mass due to their membranous, vascularized wings (Poole 1936; Њ temperatures (Ta) below 25 C, with concomitant adjustments Hartman 1963). Bats, like other mammals, are endothermic in metabolic rate (MR). Otonycteris hemprichii entered torpor and when euthermic, during periods of activity, maintain body Њ Њ at higher Ta than T. teniotis, and, when torpid, their body tem- temperatures between 35 C and 39 C (Altringham 1996). If a Њ Њ Њ Њ peratures (Tb) were 1 –2 C and 5 –8 C above Ta, respectively; constant body temperature is to be maintained, heat gain must MR was correspondingly reduced. At night, the lower critical equal heat loss, and as in all animals, heat transfer between a temperature of T. teniotis was 31.5ЊC, and that of O. hemprichii bat and its environment occurs by four pathways: conduction, was 33ЊC. Mean nocturnal thermoneutral MR of T. teniotis was convection, radiation, and evaporation. These pathways are all

37% greater than that of O. hemprichii.AthighTa, evaporative surface area dependent. water loss (EWL) increased markedly in both species, but it Among the behavioral thermoregulatory mechanisms used was significantly higher in T. teniotis above 38ЊC. In both spe- by bats are spreading and waving of the wings by megachi- cies, the dry heat transfer coefficient (thermal conductance) ropterans in order to increase convection and radiative heat followed the expected pattern for small mammals, by day and loss (Reeder and Cowles 1951) and moving between cooler or by night. Total EWL was notably low in normothermic and warmer microclimates within their roost to fine-tune heat ex- torpid animals of both species, much lower than values reported change in mega- and microchiropterans (Licht and Leitner 1967; Riskin and Pybus 1998). In addition, when ambient tem- * Present address: Department of Animal Physiology, Institute of General and perature (Ta) is lower than their lower critical temperature (Tlc), Molecular Biology, Nicolaus Copernicus University, Torun´, Poland. bats may huddle in order to keep warm (McNab 1982). The †Corresponding author; e-mail: [email protected]. reverse phenomenon, that is, increasing distance among indi-

Physiological and Biochemical Zoology 79(5):944–956. 2006. ᭧ 2006 by The viduals when Ta increases to reduce heat gain from their neigh- University of Chicago. All rights reserved. 1522-2152/2006/7905-5086$15.00 bors, has been observed in several species, including the Yuma Thermoregulation in Desert-Dwelling Insectivorous Bats 945

Table 1: Similarities and differences between the European free-tailed bat and Hemprich’s long-eared bat Hemprich’s Long-Eared Bat Otonycteris hemprichii, European Free-Tailed Bat Tadarida teniotis, Chiroptera: Vespertilionidae Chiroptera: Molossidae Similarities: Body mass 26 g 32 g Maternal care Alone Alone Roost Crevices Crevices Differences: Food availability Inadequate during winter Adequate all year round Number of offspring Twins 1 Distribution Desert habitats: Saharo-Arabian and Broad central Palearctic (genus worldwide) Irano-Turanian Distribution in Israel Found in all desert habitats All over the country; in desert area, found mainly in the vicinity of human settlements Species in genus 1 8–10 Activity in the Negev Spring to end of autumn All year round Diet Surface-dwelling arthropods Aerial insects Flight Low and slow High and fast Foraging habitat Dense vegetation Open spaces Sources. Gharaibeh and Qumsiyeh 1995; Fenton et al. 1999; Nowak 1999; Arlettaz et al. 2000; Korine and Pinshow 2004; Daniel 2005.

myotis Myotis yumanensis, the Mexican free-tailed bat Tadarida cinereus; when Ta was low, males were torpid, while pregnant brasiliensis, and the pallid bat Antrozous pallidus (Licht and females (Cryan and Wolf 2003) or those attending pups were Leitner 1967). On a larger spatial scale, some bat species migrate active (Audet and Fenton 1988). rather than face inhospitable changes in environmental con- When Ta is high, some bat species are facultatively hyper- ditions (McNab 1974). thermic (Licht and Leitner 1967; Bronner et al. 1999). That is,

The physiological mechanisms of thermoregulation in bats as Ta increases, the animal allows its Tb to increase passively. are not different from those in other mammals of comparable This response reduces energy expenditure and lessens heat gain size (McNab 1982). In response to cold, most insectivorous by reducing the temperature gradient between the body and bats reduce their body temperature (Tb) and become torpid. the environment. For example, in T. brasiliensis, M. yumanensis, Њ Њ Torpor is well known as a means to conserve energy and water and A. pallidus, Tb was elevated by 1 –2 C above Ta when Ta (Herreid and Schmidt-Nielsen 1966) and is used by many bat was 41ЊC. In addition, studies of desert-dwelling bat species species from different environments, even bats of subtropical suggest that they generally have relatively low MRs (Licht and origin (Turbill et al. 2003). Bats do not have sweat glands; thus, Leitner 1967; Bell et al. 1986; Maloney et al. 1999; but see evaporative water loss (EWL) is respiratory and transcutaneous Speakman and Thomas 2003). (Wimsatt 1970). EWL might be especially great during ﬂight Licht and Leitner (1967) posed the question of whether the because the wings are exposed to strong convection (Carpenter above-described physiological traits occur in all desert bat spe- 1969). By entering torpor and reducing their metabolic rate cies or whether they are phylogenetically independent. We ad- (MR), bats also reduce their EWL by 50%–90% (Carpenter dressed this question by comparing insectivorous bat species 1969; Morris et al. 1994). Torpor may be short and shallow sympatric in the Negev highlands (Korine and Pinshow 2004) and last several hours, or it may be prolonged and deep, usually but with different zoogeographical distributions and different during hibernation, and last up to 80 d (Brack and Twente phyletic origins, the European free-tailed bat Tadarida teniotis 1985). and Hemprich’s long-eared bat Otonycteris hemprichii (Table Entry into and emergence from torpor depend on food avail- 1). Two-species comparisons can provide only weak evidence ability, Ta, day length, and the physiological condition and re- for adaptation (Garland and Adolph 1994), but T. teniotis and productive status of an individual bat (Altringham 1996; O. hemprichii belong to different families (Jones et al. 2002), Chruszcz and Barclay 2002). As a result, a given colony of bats and we compared six distinct physiological variables (body may contain both active and torpid individuals at the same mass, MR, EWL, Tb, Tlc, and the slope of the line relating MR time (Speakman 1988). This phenomenon has been observed to Ta below Tlc). Although these variables are not independent, in the big brown bat Eptesicus fuscus and the hoary bat Lasiurus along with the many behavioral and morphological differences 946 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow

between the species (Korine and Pinshow 2004), they should CO2, and dried using columns of ﬂuffed Dacron, soda lime sufﬁce to reduce the probability that the species might be dif- (mixed calcium oxide and sodium hydroxide), and Drierite ferent simply by chance to less than 5% (Garland and Adolph (anhydrous calcium sulfate), respectively. The vapor density

1994). (rE), fractional concentration of oxygen (FEo2), and fractional

We hypothesized that the species will differ in the relative concentration of carbon dioxide (FEco2) in the air exiting each magnitudes of their physiological responses when faced with metabolism chamber was monitored. First, air passed through the same environmental conditions and predicted that Hem- a dew point and relative humidity analyzer (Sable Systems prich’s long-eared bat, a species of Saharo-Arabian and Irano- model RH-100) to measure rE. The stream was then redried Turanian distribution, will generally have lower resting MR through a Drierite-ﬁlled column and split, a portion being (RMR) and EWL and a higher dry heat transfer coefﬁcient than diverted to a differential oxygen analyzer (Applied Electro-

T. teniotis, a species of broad central Palearctic distribution. chemistry model S104) and the remainder to an infrared CO2 analyzer (LiCor 6252). ˙˙ Material and Methods Oxygen consumption (Vo22 ) and CO2 production (Vco ) were calculated following the Warthog Systems software Study Species (http://warthog.ucr.edu) arrangement of well-known respirom- Seven Tadarida teniotis and six Otonycteris hemprichii were cap- etry equations (Hill 1972; Depocas and Hart 1975; Withers ˙ Ϫ1 Ϫ1 tured with mist nets in the vicinity of Midreshet Ben-Gurion 1977).Vo2 (mL O2 h ,ormolh ) was converted to units of in the central Negev highlands, at sites permitted by the Israel power (W), taking respiratory exchange ratio into account, that

Nature and Parks Protection Authority (permit 13414, 2001– is, the oxidation of substrate by 1 L of O2 produced between 2004). Captured bats of both sexes were initially held in in- 19 and 20 kJ (Kleiber 1961; Schmidt-Nielsen 1997) and ex- dividual cages (40 cm # 40 cm # 40 cm) until they learned to pressed as mass or surface area specific. eat and drink from dishes. They were subsequently released The CO2 analyzer was calibrated by injecting into it air con- # # into an outdoor flight cage (2.5 m 4m 7 m). Between taining different precise concentrations of CO2 (0, 250, 500, experimental measurements, bats were exposed to the local and 1,000 ppm) with a gas calibration cylinder (LiCor model ppm was determined for the 5ע climate and natural photoperiod but in the attenuating confines 6000-01). An accuracy of of the flight cage. They were provided with roosting boxes measurement range used. The O2 analyzer was adjusted against suited for the specific needs of each species and fed commer- gas of known composition (21% O2, 79% N2 by mass; Maxima, cially raised mealworms and crickets, and water was provided Mizpeh Ramon, Israel). The gas was applied to both reference ad lib., with necessary vitamins added (Barnard 1995). All ex- and sample cells of the O2 analyzer, and the differential was set periments were done at least 12 h after a bat’s last meal, assuring to 0. Calibration equations were automatically generated and that they were postabsorptive (Kleiber 1961). used by the computer software of the system to read differences

in O2 concentrations between the reference and sample channels, so that drift was taken into account in each measurement Metabolic Rate cycle. MR was measured by indirect calorimetry using an eight- channel open-flow gas analysis system.V˙ o was calculated from 2 Evaporative Water Loss measurements made while bats hung on a metal mesh foothold inside metabolic chambers built from PVC piping. Each cham- Total EWL (respiratory ϩ cutaneous ) was measured from the ber had a volume of 730 cm3. To keep absorptivity of the vapor density difference between incurrent and excurrent air- chamber walls as close to unity as possible and thereby avoid streams with the dew point and relative humidity analyzer, complications caused by reflected thermal radiation (Porter following Bernstein et al. (1977): 1969), they were painted matte black. The chambers were p ˙ Ϫ mounted in a controlled-temperature cabinet (Precision 850), EWL V(EEr r I), (1) and measurements were made at temperatures between 15ЊC ˙ Њ עЊ and40 0.5 C , which covers the range of daily air temper- where EWL is in grams of H2O per minute,VE is excurrent air Ϫ1 atures in our research area. flow (mL min stpd), and rI and rE are water vapor density in the incurrent and excurrent air, respectively (g mϪ3). The dew point and relative humidity analyzer was calibrated Gas Analysis System by pumping air of known vapor density through it. The air of Air was pumped from outside the building into the system at known vapor density was produced by bubbling room air a pressure 7.5 kPa higher than atmospheric (∼96.66 kPa) and through warm water (35ЊC) in a column. The water-saturated controlled by a precision gas regulator (Qubit G260). The in- air then passed through a spiral copper tube placed in a water current air was filtered for fine particulate matter, scrubbed of bath at a temperature lower than room air temperature. Some Thermoregulation in Desert-Dwelling Insectivorous Bats 947 of the vapor condensed in the tube or in a downstream water where trap, leaving the air saturated at the new, controllable temperature (which sets the desired vapor density). This vapor- arccos(c/a) r p .(3) saturated air then exited the water bath and warmed to room ͱ1 Ϫ (c 22/a ) air temperature, retaining the desired vapor density, which was read by the dew point hygrometer. A calibration curve was calculated by least squares linear regression of the readings Dry Heat Transfer Coefficient versus vapor density. The dry heat transfer coefficient (thermal conductance; Tie- leman and Williams 1999) was calculated following Dawson Body Temperature and Schmidt-Nielsen (1966):

We used interscapular temperature as a gauge of Tb and mea- Ϫ sured it continuously with subcutaneous, indwelling type-T p MR EHL C Ϫ ,(4) (Cu-constantan) thermocouples inserted over the interscapular (TbabT )A brown adipose tissue (Wojciechowski et al. 2006). Although it is not a precise measure of deep-body temperature, this site where MR is the rate of heat production (W), EHL is rate of p was chosen to represent it because (1) interscapular tempera- evaporative heat loss (W, 1 gHO2 2.49 kJ), and Ab is body ture in the big brown bat Eptesicus fuscus is within 1ЊC of heart surface area (cm2). temperature (Hayward and Lyman 1967) and (2) bats could not reach and disturb the thermocouples. At least 24 h before Experimental Protocol experiments, a thermocouple guide-cannula was implanted subcutaneously in the interscapular region of each bat (0.8 mm The multichannel system allowed us to sequentially sample air i.d. polyethylene; Portex, Hythe, England). The cannula was from six bats at a time. One of the other two channels served inserted under the skin through a small incision at back of the as a blank with an empty metabolism chamber, and the eighth neck and fixed to the skin with surgical thread and tissue glue. was connected directly to the reference channel of the oxygen -g), fitted with a ther 0.01ע) During experiments, the cannula served as a guide for a type- analyzer. Each bat was weighed T thermocouple (0.6 mm in diameter; W-TW-36 P2, Physitemp mocouple, and placed in a metabolic chamber, where it hung Instruments, Clifton, NJ) that was inserted immediately before on the galvanized mesh. The metabolic chambers were secured an experiment and fixed in place with adhesive tape. The can- inside the controlled-temperature cabinet. The dry, CO2-free nula generally fell out spontaneously within 2 wk, and the air was pumped (Gast DOA-P601-BN) through the precision wound healed completely. Tb data were recorded with a Camp- pressure regulator and was split into eight channels by a gas bell 21X datalogger. Each thermocouple was calibrated to manifold (Qubit G245) with independent mass flow control- -0.1ЊC against a thermometer with accuracy traceable to the lers/monitors each supplying a metabolic chamber. All air chanע U.S. National Bureau of Standards. nels exited the metabolic chambers to an eight-channel gas selector (Qubit G244) that allowed the choice of a specific channel for sampling, as described above. Measurement ses- Body Surface Area sions were 5 h long. Each channel was sequentially sampled for During experiments, the bats were put inside the metabolic 8 min. No data were collected for the first hour, which was chambers and could not unfurl their wings. Their body surface allowed for stabilization of the metabolic chamber temperature ˙ area (Ab) was estimated by considering them to be flattened (Ta) and the animal’s Tb andVo2 . Thus, during the 4 h of ellipsoids. Ab was computed with measurements of the three measurement, each bat was monitored five times for 8 min (40 different radii a, b, and c, wherea 1 b 1 c . The three radii were min total). Metabolic chambers were dark during experiments. substituted into an equation that closely approximates surface Reference readings were also taken at 8-min intervals, for a area of a flattened ellipsoid (Wolfram 1999): total period of 40 min. Corrections for drift were done once each cycle, that is, every 56 min, during the 5-h measurement b 22ϩ c sessions. At the end of each experiment, bats were weighed A p 2p c 23ϩ abr ϩ r again, and the mean of the two measures of m was used for b [ 3ab b the calculation of mass-specific MR. The eight flow channels were independently calibrated using aab242422424ϩ 3ac Ϫ 12ac ϩ 8bc # 22Ϫϩ c 22 r , a bubble flow calibrator (Gilian Gilibrator 2). Ten-point cali- ()]240ab Ϫ bration curves, between 0 and 2,000 mL min 1, were created (2) and used by the computer software of the system to record flow rates, which were, on average, 1,235 mL minϪ1. 948 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow

Ϫ1 Њ Ϫ1 Measurements of MR were made at 12 different Ta ranging in O. hemprichii than in T. teniotis (1.09 vs. 0.86 mW g C ; Њ Њ p ! from 15 Cto40C. Seven individuals of each species, of both F1, 57 4.82,P 0.05 ; Fig. 1). Extrapolation of the regression sexes, were examined twice, once during the daytime (r,or lines toMR p 0 gives intercepts of 39.3ЊCforT. teniotis and Њ resting, phase) and once at night, when active but resting (a, 36.2 CforO. hemprichii. Above the Tlc, the slope of the increase or active, phase). The order of the Ta was chosen randomly. in MR was similar in O. hemprichii and T. teniotis (0.27 vs. Ϫ1 Њ Ϫ1 p 1 0.19 mW g C ;F1, 36 3.24 ,P 0.05 ); however, the MR for T. teniotis was higher, giving overall higher values (F p Data Analysis 1, 36 6.70,).P ! 0.05

Data for MR are presented as mass speciﬁc. Bats were regarded There is an apparent range of thermoneutral Ta in the r Њ as torpid when their Tb was lower than 36 C, the lowest Tb we phase of both species as well. However, by contrast with their Њ Њ found in euthermic bats (36 –38 C), at the same Ta and during a phase, wherein both species were euthermic, by day both the a phase. We did not differentiate between deep and shallow species reduced MR and became torpid at low Ta (Fig. 1). This torpor. reduction in MR began at a higher Ta in O. hemprichii than in

Statistical Methods We followed the method of Pinshow et al. (1976) to objectively determine Tlc. A range of temperatures was chosen that was broad enough to clearly include the Tlc on a plot relating MR to Ta. The data points were then successively divided into two groups, an upper temperature group and a lower temperature group, and the corresponding pair of linear regression lines and their pooled mean squares (PMS) were calculated. This process was repeated, and the intersect of the pair of regression lines with the lowest PMS was considered to be the Tlc (Pinshow et al. 1976). After determining the Tlc by this objective means, the range of Ta was divided in two: temperatures less than Tlc and temperatures greater than Tlc. Data were analyzed using SAS software, and general linear model analysis was done using the mixed model procedure with the bats’ individual identiﬁcation numbers as the categorical variable for repeated measures. The interactions among different treatments (Ta and time of day) and the measured variables were examined for each species separately. Least squares linear regressions and differences in regression slopes were analyzed by ANCOVA using SYSTAT 9. SD.P ! 0.05 was chosen as the lowest 1ע Means are presented acceptable level of signiﬁcance.

Results Metabolic Rate Mean body masses of Tadarida teniotis and Otonycteris hem- g (n p 2.09 ע g (n p 7 ) and 25.43 2.31 ע prichii were31.97

6), respectively. During their a phase, the Tlc of each species was clearly defined: 32.0ЊC and 33.5ЊCforT. teniotis and O. Figure 1. Metabolic rate as a function of ambient temperature (Ta) for hemprichii, respectively. In both species, above their Tlc there Tadarida teniotis (a) and Otonycteris hemprichii (b) resting at night was a significant but slow increase in MR up to 40ЊC, the (filled circles) and resting by day (open circles). Data are presented as ע maximum Ta to which the bats were exposed. Thus, we did means SD. Solid lines describe the least squares linear regression ! not determine their upper critical temperatures. However, the equations for two ranges of Ta at night. For T. teniotis:atTalcT , MR p Ϫ0.86T ϩ 33.93 (F p 203.95 ,P ! 0.001 ,r2 p 0.86 ), and at mean mass-specific RMR between the T and 40ЊC was sig- a1,33 lc T 1 T ,MR p 0.19T ϩ 0.33 (F p 6.17 ,P ! 0.02 ,r2 p 0.23 ). For alc a 1,21 ע nificantly lower in O. hemprichii than in T. teniotis (3.57 O. hemprichii:atT ! T ,MR p Ϫ1.09T ϩ 39.33 (F p 191.95 , Ϫ1 Ϫ1 a lc a 1, 33 mW g ,;n p 5 F p ! 2 p 1 p Ϫ p 0.46 ע mW g ,n p 5 , vs. 6.00 0.21 1, 36 P 0.001,r 0.85 ), and atTalcTMR ,0.27T a5.26 (F 1,33 ! ! 2 p 6.70,P 0.05 ). Below the Tlc, the rate of increase was greater 43.79,,P 0.05 r 0.45 ). Thermoregulation in Desert-Dwelling Insectivorous Bats 949

T. teniotis,at32ЊC and 25ЊC, respectively. MR above 32.0ЊC, the Tlc, was signiﬁcantly different between the a and r phases p ! in T. teniotis (,;Fig.1F1, 48 15.88 P 0.0005 a). In contrast, no p 1 such difference was found in O. hemprichii (,F1, 40 0.26 P 0.5;Fig.1b).

Body Temperature At night, both T. teniotis and O. hemprichii maintained ! euthermic-level Tb, and atTalcT , there was no difference be- Њ עЊ tween the mean Tb of O. hemprichii (35.66 0.34 C ) and that Њ p 1 עЊ of T. teniotis (35.10 0.34 C ;F1, 57 0.88 ,P 0.05 ), while עЊ mean Tb above the Tlc was higher for T. teniotis (38.05 Њ p עЊ Њ 1.91 C37.08) than for O. hemprichii (;,1.92 C F1, 36 6.25 P ! 0.05; Fig. 2). During the day (r phase), both species reduced

Tb and entered torpor at low Ta; however, the Tb of T. teniotis was higher than that of O. hemprichii (22.5Њ–32.2ЊC and 18.4Њ– Њ p ! 32.5 C, respectively;F1, 68 4.20 ,P 0.05 ).

The Tb of T. teniotis was significantly different between its a ! Њ p ! and r phases atTa1,7425 C (F 73.05 ,P 0.0001 ). During ! Њ the night, atTa 25 C , the Tb of T. teniotis was relatively high ,(1.12ЊC (Fig. 2aע33.5Њ–36.5ЊC), with a mean value of35.11Њ) ∼ while during the day, all individuals entered torpor at Ta Њ Њ Њ 25 C, and their Tb was between 5 C and 8 C above Ta.At 1 Њ Ta 25 C, Tb remained high, and there was no difference be- p 1 tween night and day values (F1, 48 0.86 ,P 0.05 ). In com- parison, Tb in O. hemprichii was significantly different between ! Њ p ! its a and r phases atTa1,6032 C (F 217.19 ,P 0.0001 ). ! Њ During the night, atTa 32 C , Tb of O. hemprichii was relatively .1.15ЊC (Figעhigh (31.1Њ–37.2ЊC), with a mean value of35.54Њ 2b), while during the day, all individuals entered torpor at about Figure 2. Body temperature as a function of ambient temperature for Њ Њ Њ six Tadarida teniotis (a) and six Otonycteris hemprichii (b)atnight 32 C, and their Tb was between 1 C and 2 C above Ta.At T 1 35ЊC, T remained high, and there was no difference be- (filled circles) and by day (open circles). Data are presented as .SD ע a b means p 1 tween mean night and day values (F1, 40 0.001 ,P 0.05 ). thermic than for torpid O. hemprichii (Fig. 3b). EWL in T. 1 Њ Evaporative Water Loss teniotis increased sharply atTa 33.5 C , both by day and by In both species, total EWL increased with increasing T (Fig. night. EWL in O. hemprichii also increased but to a lesser extent, a Њ 3). In T. teniotis, the minimum nocturnal EWL at 15ЊC (0.18 and the increase began at 35.0 C (Fig. 3). mg minϪ1) was only 6% of its maximum rate at 40ЊC (3.02 mg minϪ1). For O. hemprichii, minimum nocturnal EWL at Dry Heat Transfer Coefficient (Thermal Conductance) 15ЊC was slightly higher, at 9% of its maximum rate (0.19 mg minϪ1 at 15ЊC vs. 2.02 mg minϪ1 at 40ЊC). EWL in both species Mean estimated surface areas of animals with folded wings for p 2 ע ! atTalcT was low by day and by night (Table 2) but increased T. teniotis and O. hemprichii were61.54 7.6 3 cm ()n 7 p 2 ע significantly with increasing Ta above their respective Tlc.At and44.09 3.53 cm (n 6 ), respectively. In both species, in Њ Њ night, at Ta between 25 C and 35 C, mean EWL was higher in both their a and r phases, the dry heat transfer coefficient (C) p ! ! T. teniotis than in O. hemprichii (F1, 57 32.03 ,P 0.005 ), but increased with increasing Ta (Fig. 4). At night, whenTalcT , 1 Њ atTa 35 C , EWL increased more rapidly in T. teniotis than in C was significantly lower in T. teniotis than in O. hemprichii Њ Ϫ1 ע Њ Ϫ1 Ϫ2 p ע p Њ O. hemprichii, and atTa 40 C , it was 1.5 times higher in T. (mW0.66 0.12 C cm ,,vs.mWn 25 0.70 0.10 C p ! Ϫ2 p p 1 1 teniotis than in O. hemprichii (,).F1, 36 5.88 P 0.05 cm ,n 25 ;F1, 575.31 ,P 0.05 ; Fig. 4a). AtT aT lc , there The EWL of torpid T. teniotis did not differ from the EWL was no significant difference in C between the species p 1 p 1 of normothermic animals (F1, 74 0.98 ,P 0.05 ; Fig. 3a). How- (F1, 36 0.30 ,P 0.05 ), and the maximum values of C were p ! Њ Ϫ1 Ϫ2 ever, EWL was higher (F1, 60 23.52 ,P 0.0005 ) for normo- 2.70 and 2.80 mW C cm for O. hemprichii and T. teniotis, 950 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow

species (Degen 1997). Otonycteris hemprichii, as expected, has

a lower mass-speciﬁc MR, enters torpor at a higher Ta, and has

lower EWL in the higher range of Ta tested. Heat production and its associated physiological functions follow a daily rhythm that, under natural conditions, may lead to considerable energy and water savings. However, a difference in EWL at tempera-

tures below the Tlc in either species was not found during day or night.

Metabolic Rate

The responses of MR to changes in Ta in O. hemprichii and T. teniotis are consistent with the responses observed in other small insectivorous bats (Speakman and Thomas 2003). However,

although MR decreases at high Ta in both species, the paths

they follow with decreasing Ta are different. The mass-specific Њ MR of O. hemprichii between its Tlc and 40 C was significantly lower than that of T. teniotis in that range, even though T. teniotis has a greater body mass. Otonycteris hemprichii, as expected from its zoogeographic origins, uses energy frugally, as reflected by its low basal metabolic rate (BMR). This phenomenon is documented for other taxa of desert mammals, large and small (Lovegrove 2000, 2003; Williams et al. 2004), and for desert birds (Dawson 1984; Maclean 1996; Tieleman et al. 2002).

Below their Tlc, euthermic bats of both species have typical mammalian patterns of thermal adjustment; that is, MR in-

creases with decreasing Ta. In both species, this results in a substantial increase in metabolic cost. To answer the question of whether physiological traits suited for survival in deserts, Figure 3. Evaporative water loss of six Tadarida teniotis (a) and six such as low MR, occur in all desert bat species or whether they Otonycteris hemprichii (b) as a function of ambient temperature at are dependent on phylogeny, we compared the slopes of the night (filled circles) and by day (open circles). Data are presented as regression lines relating BMR (Y)tomb (X) of 39 insectivorous -SD. bat species from 10 different families and of different zooge ע means ographic distributions by ANCOVA, with log body mass as respectively. In addition, by day, there were no significant dif- covariate (Fig. 5; Table 3). In contrast to the results of this ! p p ferences in C between T. teniotis and O. hemprichii at TalcT study, there was no difference (F1, 35 3.62 ,P 0.065 ) in 1 p 1 p 1 or atTaTF lc ( 1,680.64 ,P 0.05 ;F 1,42 1.87 ,P 0.05 , re- BMRs among desert and nondesert insectivorous bat species. spectively; Fig. 4b). In addition, we did a phylogenetically informed ANCOVA to In O. hemprichii, there was no significant difference in C account for nonindependence of species within a phylogeny ע ע ! between day and night whenTalcT (0.69 0.35 vs. 0.70 (Felsenstein 1985), following the techniques of Garland et al. Њ Ϫ1 Ϫ2 p 1 1 0.10 mW C cm ;F1, 600.001 ,P 0.05 ), or when T aT lc (1993) with the PDAP software package (Garland et al. 2002). p 1 (F1, 40 0.20 ,P 0.05 ). In T. teniotis, there was also no signif- We used a phyletic tree for bats based on the topology of Jones ! icant difference in C between day and night when TalcT et al. (2002), assuming equal branch lengths (Pagel 1992). Њ Ϫ1 Ϫ2 p 1 ע ע (0.65 0.48 vs. 0.66 0.13 mW C cm ;,F1, 74 2.49 P Again, we found no difference between desert and nondesert 1 p 1 p p 0.05) or whenTalc1,48TF (0.64 ,P 0.05 ). species of insectivorous bats (F1, 35 3.62 ,P 0.082 ). Previous studies of insectivorous bats indicate that low RMR is a char- Discussion acteristic common to insectivorous bats in general (McNab 1969; Lindstedt and Boyce 1985; Genoud and Bonaccorso 1986; The physiological responses of Otonycteris hemprichii to the Speakman and Thomas 2003). Our analysis supports this as- experimental conditions in this study, both by night and by sertion and suggests that desert-dwelling species are included day, are more characteristic of small desert mammals than are in the generality. those of Tadarida teniotis, which are characteristic of a mesic It is, however, possible that a difference in RMR between Thermoregulation in Desert-Dwelling Insectivorous Bats 951

Table 2: Evaporative water loss (EWL) of 17 insectivorous bat species Body EWL Species Mass (g) (mg minϪ1) Reference Pipistrellus pipistrellus 6.2 1.4 Webb et al. 1995 Myotis thysanodes 8.2 1.9 Studier et al. 1970 Myotis lucifugus 7.9 .95 Studier et al. 1970 Nyctophilus gouldi 9 .23 Morris et al. 1994 Plecotus auritus 9.1 1 Webb 1995 Myotis velifer 9.4 .55 Studier et al. 1970 Myotis daubentonii 10 1.1 Webb 1995 Tadarida brasiliensis 10.4 .52 Herreid and Schmidt-Nielsen 1966 Macrotus californicus 11.7 1.17 Bell et al. 1986 Eptesicus fuscus 16.9 .85 Herreid and Schmidt-Nielsen 1966 Chalinolobus gouldii 17.5 .67 Hosken and Withers 1997 Leptonycteris sanborni 22 1.5 Carpenter and Graham 1967 Mops condylurus 23 .42 Maloney et al. 1999 Noctilio albiventris 40 1.9 Chappell and Roverud 1990 Artibeus hirsutus 48 2.96 Carpenter and Graham 1967 Tadarida teniotis 33.4 .29 This study Otonycteris hemprichii 27.1 .23 This study desert and nondesert bats exists but could not be observed in has a better ability to conserve energy and possibly water by this analysis for three reasons. First, because of limited data for using deeper torpor, which can be employed at higher Ta. RMR available in the literature, the number of desert bat species that we used in the statistical analysis was low compared to the Body Temperature number of nondesert species. Second, it is sometimes difﬁcult to draw a sharp distinction between desert and nondesert dis- During the night, Tb of both species remained at the euthermic tributions because bats are able to ﬂy long distances between level for hours, even at relatively low Ta. Evidently, both T. dry and moist habitats. Third, it is known that plasticity of teniotis and O. hemprichii can maintain euthermy in the cold.

RMR occurs within different populations of the same species Above their Tlc, Tb of both T. teniotis and O. hemprichii in- p Њ inhabiting different regions (Bonaccorso and McNab 1997). creased with rising Ta fromTa 32.0 C for T. teniotis and p Њ Both T. teniotis and O. hemprichii used torpor to reduce Ta 33.5 C for O. hemprichii (Fig. 3). This increase in Tb metabolic costs at low Ta. The use of daily torpor resulted in matched the moderate increase in MR within the same range energy savings of between 17% and 75% of the euthermic heat of Ta. Increasing Tb increases skin temperature and, along with ! Њ production in T. teniotis atTa 25 C and of between 51% and it, the animal surface–to-ambient temperature gradient. This, ! Њ 93% in O. hemprichii atTa 32 C . There were differences in in turn, increases the potential for heat loss by convection, MRs between day and night in both species (Fig. 1). At night, conduction, and radiation, thereby decreasing the need for bats consistently maintained high MRs, while during the day, evaporative cooling (Calder and King 1974). Such a mechanism torpor was used, especially at low Ta. can be advantageous for mammals inhabiting arid envi- To gain a better understanding of the differences in the ther- ronments. moregulatory abilities between species, it is necessary to ex- Tb in both T. teniotis and O. hemprichii was labile. Both amine the degree to which Tb is regulated relative to Ta (Hen- maintained euthermic Tb at the lowest temperatures to which shaw and Folk 1966). Our subjective judgment of when a bat they were exposed at night but let their Tb fall almost to the enters torpor is usually based on Tb, but as bats have labile Tb, level of Ta by day, thus reducing energy expenditure. Although

Tb should also be used in conjunction with other variables to a similar pattern of physiological response to low Ta during determine the onset of torpor. Geiser and Ruf (1995) suggested daytime was observed for the two species, Tb of O. hemprichii ˙ ! Њ that changes inVo2 may be a better indicator of entry into was signiﬁcantly lower atTa 32 C . These differences in ther- torpor than Tb alone. Otonycteris hemprichii entered torpor at moregulatory response between day and night conﬁrm other ! Њ higher Ta than T. teniotis, and its MR atTa 32 C was lower observations that torpor is not an obligatory response to low than that of T. teniotis at the same Ta. This suggests that T. air temperature (Avery 1985; Audet and Fenton 1988). In ad- teniotis has a higher tolerance of low Ta, while O. hemprichii dition, torpor is apparently a mechanism used to conserve en- 952 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow

the rates of EWL of both T. teniotis and O. hemprichii were comparatively low, the rate was signiﬁcantly lower in O. hemprichii than in T. teniotis, suggesting that it has an especially frugal water economy. The lower EWLs of O. hemprichii and T. teniotis are strong indirect evidence for effective use of physiological mechanisms, such as temporal countercurrent heat exchange, to save heat and water, as described by Schmidt- Nielsen et al. (1970). The presence of the temperature gradients required for this mechanism to be effective was demonstrated in arid (Schmidt-Nielsen et al. 1970) and mesic mammals, including three vespertilionid bats (Schmid 1976), and we as- sume that it functions in O. hemprichii and T. teniotis as well. As a result of countercurrent heat exchange in the nasal passages, reduction of respiratory water loss and energy occurs

because air is exhaled below core Tb, allowing heat (the latent heat of condensation) and water to be recovered rather than lost to the environment (Hillenius 1992). Herreid and Schmidt-Nielsen (1966) regarded torpor as an energy- and water-saving mechanism. Several studies conﬁrmed that water loss is reduced during torpor in bats (see Webb et al. 1995). In our study, differences in EWL between torpid and euthermic bats were found only for O. hemprichii. This also supports the idea that O. hemprichii tolerates xeric conditions better than T. teniotis.

Dry Heat Transfer Coefﬁcient According to Scholander et al. (1950), an animal’s dry heat transfer coefﬁcient (thermal conductance, C) in homeothermic

Figure 4. Dry heat transfer coefficient of six Otonycteris hemprichii animals is expected to be minimal at the animal’s Tlc and to (open circles) and six Tadarida teniotis (filled circles) as a function of remain constant below that temperature. We found this only ambient temperature at night (a) and by day (b). The two points marked with plus signs fall outside the 95% confidence limits of the regression relationships and were therefore not used for further analysis. ergy during the r phase in these species, and, when necessary, Њ O. hemprichii may enter torpor at Ta as high as 32 C.

The main apparent difference in Tb regulation between the two species appears to be the minimum Ta down to which each maintains euthermic Tb. While O. hemprichii allows its Tb to Њ drop beginning at a Ta of 32 C, T. teniotis remains euthermic Њ until Ta reaches 25 C, suggesting again that T. teniotis is more tolerant of lower Ta in the range tested. This might explain its presence in the research area all year round, while O. hemprichii, which is better able to endure drier and hotter environments, is not found active during the cooler months (Korine and Pinshow 2004). Figure 5. Relationships between basal metabolic rate (BMR) and body

mass (mb) for nondesert (ﬁlled circles) and desert (open circles) insec- Evaporative Water Loss tivorous bat species (see Table 3 for a list of the species used here). Solid and dashed lines describe the least squares linear regression equa- Both T. teniotis and O. hemprichii have rates of EWL approx- p ϩ 2 p tions for nondesert (BMR 4.44mb 49.62 ,r 0.84 ) and desert p ϩ 2 p imately 50% lower than other bat species (Table 2). Although (BMR 2.87mb 53.71 ,r 0.86 ) species, respectively. Thermoregulation in Desert-Dwelling Insectivorous Bats 953

Table 3: Basal metabolic rates (BMR) of 39 insectivorous bat species from 10 different families Desert/ Body BMR Species Nondesert Family Mass (g) (mW) Reference Peropteryx macrotis ND Em 5.1 65.84 Genoud et al. 1990 Saccopteryx bilineata ND Em 8.2 85.36 Genoud and Bonaccorso 1986 Saccopteryx leptura ND Em 4.2 53.00 McNab 1982 Hipposideros galeritus ND Rh 8.5 52.45 McNab 1989 Rhinolophus ferrumequinum ND Rh 28 259.56 Speakman et al. 2003 Noctilio albiventris ND No 40 220.8 Chappell and Roverud 1990 Noctilio leporinus ND No 61 262.30 McNab 1969 Mormoops blainvilli ND Mo 8.9 48.51 Rodrı´guez-Durań 1995 Pteronotus davii ND Mo 9.4 85.35 Bonaccorso et al. 1992 Pteronotus parnellii ND Mo 19.2 171.26 Bonaccorso et al. 1992 Pteronotus personata ND Mo 14 128.38 Bonaccorso et al. 1992 Pteronotus quadridens ND Mo 4.8 33.46 Rodrı´guez-Durań 1995 Macrotus californicus D Ph 13 125.58 Bell et al. 1986 Phyllostomus discolor ND Ph 33.5 192.63 McNab 1969 Phyllostomus elongatus ND Ph 35.6 216.45 McNab 1969 Phyllostomus hastatus ND Ph 84.2 394.90 McNab 1969 Antrozous pallidus D An 22 118.36 Licht and Leitner 1967 Eumops perotis D Mol 56 222.32 Leitner 1966 Molossus molossus ND Mol 16.5 112.20 McNab 1969 Tadarida brasiliensis ND Mol 11 85.36 Licht and Leitner 1967 Tadarida teniotis ND Mol 33.4 200.4 This study Chalinolobus gouldii ND Ve 17.5 140.53 Hosken and Withers 1997 Eptesicus serotinus ND Ve 27 240.57 Speakman et al. 2003 Histiotus velatus ND Ve 11.2 55.78 McNab 1969 Lasiurus cinereus ND Ve 30 270 Cryan and Wolf 2003 Miniopterus schreibersii ND Ve 10.5 133.88 Brown 1999 Myotis austroriparius ND Ve 7.5 60.83 McNab 1969 Myotis lucifugus ND Ve 7.7 104.34 Kurta and Kunz 1988 Myotis myotis ND Ve 25 139.5 Hanus 1959 Myotis thysanodes ND Ve 8.1 97.12 O’Farrell and Studier 1970 Myotis yumanensis D Ve 5.5 68.64 O’Farrell and Studier 1970 Nyctophilus geoffroyi D Ve 8 63.60 Hosken and Withers 1999 Nyctophilus major ND Ve 13.6 113.83 Hosken 1997 Otonycteris hemprichii D Ve 27.1 101.63 This study Pipistrellus pipistrellus ND Ve 6.6 67.52 Speakman 1993 Plecotus auritus ND Ve 11.1 89.24 McLean and Speakman 2000 Vespadelus vulturnus ND Ve 4 22.4 Willis et al. 2005 Natalus tumidirostris ND Na 5.4 46.33 Genoud et al. 1990 Rhinonycteris aurantius ND Hi 8.3 90.47 Baudinette et al. 2000 Note.D p desert species, ND p nondesert species. Em p Emballonuridae , Rh p Rhinolophidae , No p Noctilionidae , Mo p Mormoopidae , Ph p Phyllostomidae, An p Antrozoidae , Mol p Molossidae , Ve p Vespertilionidae , Na p Natalidae , and Hi p Hipposideridae . for O. hemprichii. In contrast, T. teniotis reached a minimal Heath 1970). Since bats in our experiments were in opaque p Њ value of C at a Ta lower than its Tlc (Ta 29 C ). Bats can metabolic chambers, we could not observe their behavior and change their dry heat transfer coefficient by behavioral and/or thus could not differentiate between behavioral and physio- physiological means. They can spread and flap their wings logical changes in C. (Reeder and Cowles 1951; Bartholomew et al. 1964), and they In this study, no significant differences between day and night can modulate peripheral blood flow by vasoconstriction and values of dry heat transfer coefficient were found for either vasodilation, especially in their membranous wings (Kluger and species. By contrast, previous studies reported differences be- 954 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow tween day and night values of whole-animal “dry” conductance Avery M.I. 1985. Winter activity of pipistrelle bats. J Anim Ecol in other species of microchiropteran bats (Chalinolobus gouldii, 54:721–738. Hosken and Withers 1997; Nyctophilus geoffroyi, Hosken and Barnard S.M. 1995. Bats in Captivity. Wildlife Conservation So- Withers 1999; Mops condylurus, Maloney et al. 1999). We can- ciety, Morrow, GA. http://www.basicallybats.org/onlinebook. not offer a substantiated explanation for all the differences in Bartholomew G.A., P. Leitner, and J.E. Nelson. 1964. Body conductance, but in the case of C. gouldii and N. geoffroyi,the temperature, oxygen consumption and heart rate in three explanation may be that the differences were found at tem- species of Australian flying foxes. Physiol Zool 37:179–198. peratures below the range covered in our study and in torpid Baudinette R.V., S.K. Churchill, K.A. Christian, J.E. Nelson, and and hibernating bats. P.G. Hudson. 2000. Energy, water balance and the roost mi-

During the night, the difference between Tb and Ta increased, croenviroment in three Australian cave-dwelling bats (Mi- and the rate of metabolic heat production increased as well, in crochiroptera). J Comp Physiol B 170:439–446. order to maintain euthermic Tb. In addition, during the day, Bell G.P., G.A. Bartholomew, and K.A. Nagy. 1986. The roles the difference between Tb and Ta decreased, and the rate of of energetics, water economy, foraging behaviour and geo- metabolic heat production also decreased. As a result, the MR thermal refugia in the distribution of the bat, Macrotus cal- Ϫ toTbaT ratio was maintained. Evaporative heat loss, which ifornicus. J Comp Physiol B 156:441–450. in principle could also affect C, was found to be very low and Bernstein M.H., D.M. Hudson, J.M. Stearns, and R.W. Hoyt. could not result in signiﬁcant changes between its day and night 1977. Measurement of evaporative water loss in small ani- values. mals by dew-point hygrometry. J Appl Physiol 43:382–385. In conclusion, we found that the physiological responses of Bonaccorso F.J., A. Arends, M. Genoud, D. Cantoni, and T. Mor-

O. hemprichii to a range of Ta, both by night and by day, are ton. 1992. Thermal ecology of moustached and ghost-faced characteristic of a mammal adapted to arid environments, while bats (Mormoopidae) in Venezuela. J Mammal 73:365–378. those of T. teniotis are characteristic of a mesic species. Both Bonaccorso F.J. and B.K. McNab. 1997. Plasticity of energetics species use torpor equally well to save energy and water, and in blossom bats (Pteropodidae): impact on distribution. J both are extremely frugal compared to other small insectivorous Mammal 78:1073–1088. bats that have been studied. On a larger scale, we did not find Brack V. and J.W. Twente. 1985. The duration of the period of that the desert-inhabiting species among the insectivorous bats hibernation of three species of vespertilionid bats: field stud- that we compared have lower MRs than the mesic species. ies. Can J Zool 65:1240–1242. Bronner G.N., S.K. Maloney, and R. Buffenstein. 1999. Survival tactics within thermally-challenging roosts: heat tolerance Acknowledgments and cold sensitivity in the Angolan free-tailed bat, Mops condylurus. S Afr J Zool 34:1–10. We thank Amram Zabari for his exemplary technical help in Brown C.R. 1999. Metabolism and thermoregulation of indi- this research. We also thank Anton Fennec, Ran Marom, and vidual and clustered long-fingered bats, Miniopterus schrei- the students in the Physiological Ecology Laboratory, in par- dersii, and the implications for roosting. S Afr J Zool 34: ticular Shai Daniel, for their assistance in capturing and main- 166–172. taining the bats, Dr. Eli Groner for his help with the statistical Calder W.A. and J.R. King. 1974. Thermal and caloric rela- analysis, and Dr. Ted Garland for his comments on a draft of tionships of birds. Pages 259–413 in D.S. Farner and J.R. this manuscript. This research was funded by Israel Science King, eds. Avian Biology. Academic Press, New York. Foundation grant 653/01-1 to B.P. and C.K. and a student Carpenter R.E. 1969. Structure and function of the kidney and support grant from the Mitrani Department of Desert Ecology water balance of desert bats. Physiol Zool 42:288–302. (MDDE). This is paper 523 of the MDDE. Carpenter R.E. and J.B. Graham. 1967. Physiological responses to temperature in the long-nosed bat, Leptonycteris sanborni. Literature Cited Comp Biochem Physiol 22:709–722. Chappell M.A. and R.C. Roverud. 1990. Temperature effects Altringham J.D. 1996. Bats: Biology and Behaviour. Oxford on metabolism ventilation and oxygen extraction in a Neo- University Press, London. tropical bat. Respir Physiol 81:401–412. Arlettaz R., C. Ruchet, J. Aeschimann, E. Brun, M. Genoud, Chruszcz R. and R.M.R. Barclay. 2002. Thermoregulatory ecol- and P. Vogel. 2000. Physiological traits affecting distribution ogy of a solitary bat, Myotis evotis, roosting in rock crevices. and wintering strategy of the bat Tadarida teniotis. Ecology Funct Ecol 16:18–26. 81:1004–1014. Cryan P.M. and B.O. Wolf. 2003. Sex differences in the ther- Audet D. and M.B. Fenton. 1988. Heterothermy and the use moregulation and evaporative water loss of a heterothermic of torpor by the bat Eptesicus fuscus (Chiroptera: Vesperti- bat, Lasiurus cinereus, during its spring migration. J Exp Biol lionidae): a field study. Physiol Zool 61:197–204. 206:3381–3390. Thermoregulation in Desert-Dwelling Insectivorous Bats 955

Daniel S. 2005. Body Temperature Regulation and Foraging Henshaw R.E. and G.E. Folk. 1966. Relation of thermoregulation Activity in Reproductive Female Hemprich’s Long-Eared to seasonally changing microclimate in two species of bats Bats (Otonycteris hemprichii) in the Negev Desert. PhD thesis. (Myotis lucifugus and M. sodalis). Physiol Zool 39:223–236. Ben-Gurion University of the Negev, Beer Sheva, Israel. Herreid C.F. and K. Schmidt-Nielsen. 1966. Oxygen consump- Dawson T. and K. Schmidt-Nielsen. 1966. Effect of thermal tion, temperature and water loss in bats from different en- conductance on water economy in the antelope jack rabbit. vironments. Am J Physiol 211:1108–1112. J Cell Physiol 67:463–472. Hill R.W. 1972. Determination of oxygen consumption by use Dawson W.R. 1984. Physiological studies of desert birds: pre- of the paramagnetic oxygen analyzer. J Appl Physiol 33:261– sent and future considerations. J Arid Environ 7:133–155. 263. Degen A.A. 1997. Ecophysiology of Small Desert Mammals. Hillenius W. 1992. The evolution of nasal turbinates and mam- Springer, Berlin. malian endothermy. Paleobiology 18:17–29. Depocas F. and J.S. Hart. 1975. Use of the Pauling oxygen Hosken D.J. 1997. Thermal biology and metabolism of the analyzer for measurement of oxygen consumption of animals greater long-eared bat, Nyctophilus major (Chiroptera: Ves- in open circuit and in short-lag, closed-circuit apparatus. J pertilionidae). Aust J Zool 45:145–156. Appl Physiol 10:388–392. Hosken D.J. and P.C. Withers. 1997. Temperature regulation Felsenstein J. 1985. Phylogenies and the comparative method. and metabolism of an Australian bat, Chalinolobus gouldii Am Nat 125:1–15. (Chiroptera: Vespertilionidae) when euthermic and torpid. Fenton M.B., B. Shalmon, and D. Makin. 1999. Roost switching, J Comp Physiol B 167:71–80. foraging behavior and diet of the vespertilionid bat, Oto- ———. 1999. Metabolic physiology of euthermic and torpid nycteris hemprichii. Isr J Zool 41:501–506. lesser long-eared bats Nyctophilus geoffroyi (Chiroptera: Ves- Garland T., Jr., and S.C. Adolph. 1994. Why not to do two- pertilionidae). J Mammal 80:42–52. species comparative studies: limitations on inferring adap- Jones K.E., A. Purvis, A. MacLarnon, O.R.P. Bininda-Emond, tation. Physiol Zool 67:797–828. and N.B. Simmons. 2002. A phylogenetic supertree of the Garland T., Jr., A.W. Dickerman, C.M. Janis, and J.A. Jones. bats (Mammalia: Chiroptera). Biol Rev 77:223–259. 1993. Phylogenetic analysis of covariance by computer sim- Kleiber M. 1961. The Fire of Life: An Introduction to Animal ulation. Syst Biol 42:265–292. Energetics. University of California, Davis. Garland T., Jr., P.E. Midford, J.A. Jones, A.W. Dickerman, and Kluger M.J. and J.A. Heath. 1970. Vasomotion in the bat wing R. Diaz-Uriarte. 2002. PDAP: Phenotypic Diversity Analysis and thermoregulatory response to internal heating. Comp Programs. Version 6.0. University of Wisconsin, Madison. Biochem Physiol 32:219–226. Geiser F. and T. Ruf. 1995. Hibernation versus daily torpor in Korine C. and B. Pinshow. 2004. Guild structure, foraging space mammals and birds: physiological variation and classification use, and distribution in a community of insectivorous bats of torpor patterns. Physiol Zool 68:935–966. in the Negev Desert. J Zool (Lond) 262:187–196. Genoud M. and F.J. Bonaccorso. 1986. Temperature regulation, Kurta A. and T.H. Kunz. 1988. Roosting metabolic rate and rate of metabolism and roost temperature in the greater body temperature of male little brown bats, Myotis lucifugus, white-lined bat Saccopteryx bilineata (Emballonuridae). in summer. J Mammal 69:645–651. Physiol Zool 59:49–54. Genoud M., F.J. Bonaccorso, and A. Arends. 1990. Rate of Leitner P. 1966. Body temperature, oxygen consumption, heart metabolism and temperature regulation in two small tropical rate and shivering in the California mastiff bat, Eumops per- insectivorous bats (Pteropteryx macrotis and Natalus tumi- otis. Comp Biochem Physiol 19:431–443. dirostris). Comp Biochem Physiol 97:229–234. Licht P. and P. Leitner. 1967. Physiological responses to high Gharaibeh B.M. and M.B. Qumsiyeh. 1995. Otonycteris hem- environmental temperatures in tree species of microchirop- prichii. Mamm Species 514:1–4. teran bats. Comp Biochem Physiol 22:371–387. Hanus K. 1959. Body temperatures and metabolism of bats at Lindstedt S.L. and M.S. Boyce. 1985. Seasonality, fasting en- different environmental temperatures. Physiol Bohemoslov durance, and body size in mammals. Am Nat 125:873–878. 8:250–259. Lovegrove B.G. 2000. The zoogeography of mammalian basal Hartman F.A. 1963. Some flight mechanisms of bats. Ohio J metabolic rate. Am Nat 156:201–219. Sci 63:59–65. ———. 2003. The influence of climate on the basal metabolic Hayward J.S. and Lyman C.P. 1967. Nonshivering heat pro- rate of small mammals: a slow-fast metabolic continuum. J duction during arousal from hibernation and evidence for Comp Physiol B 173:87–112. the contribution of brown fat. Pages 346–355 in K.C. Fisher, Maclean G.L. 1996. The Ecophysiology of Desert Birds. A.R. Dawe, C.P. Lyman, E. Scho¨nbaum, and F.E. South, eds. Springer, Berlin. Mammalian Hibernation III, Proceedings. Oliver & Boyd, Maloney S.K., G.N. Bronner, and R. Buffenstein. 1999. Ther- Edinburgh. moregulation in the Angolan free-tailed bat Mops condylurus: 956 S. Marom, C. Korine, M. S. Wojciechowski, C. R. Tracy, and B. Pinshow

a small mammal that uses hot roosts. Physiol Biochem Zool 1950. Heat regulation in some arctic and tropical mammals 72:385–396. and birds. Biol Bull 99:237–258. McLean J.A. and J.R. Speakman. 2000. Effects of body mass and Speakman J.R. 1988. Position of pinnae and thermoregulatory reproduction on the basal metabolic rate of brown long-eared status in brown long-eared bats. J Therm Biol 13:25–30. bats (Plecotus auritus). Physiol Biochem Zool 73:112–121. ———. 1993. The evolution of echolocation for predation. McNab B.K. 1969. The economics of temperature regulation Symp Zool Soc Lond 65:39–63. in Neotropical bats. Comp Biochem Physiol 31:227–268. Speakman J.R., W.C. Lancaster, W. Ward, G. Jones, and K.C. ———. 1974. The behavior of temperate cave bats in a sub- Cole. 2003. Energy costs of echolocation in stationary in- tropical environment. Ecology 55:943–958. sectivorous bats. Pp. 361–365 in J. Thomas, C.F. Moss, and ———. 1982. Evolutionary alternatives in the physiological M. Vater, eds. Echolocation in Bats and Dolphins. University ecology of bats. Pages 151–200 in T.H. Kunz, ed. Ecology of of Chicago Press, Chicago. Bats. Plenum, New York. Speakman J.R. and D.W. Thomas. 2003. Physiological ecology ———. 1989. Temperature regulation and rate of metabolism and energetics of bats. Pages 430–490 in T.H. Kunz and M.B. in three Bornean bats. J Mammal 70:153–161. Fenton, eds. Bat Ecology. University of Chicago Press, Morris S., A.L. Curtin, and M.B. Thompson. 1994. Hetero- Chicago. thermy, torpor, respiratory gas exchange, water balance and Studier E.H., J.W. Proctor, and D.J. Howell. 1970. Diurnal body the effect of feeding in Gould’s long-eared bat Nyctophilus weight loss and tolerance of weight loss in five species of gouldi. J Exp Biol 197:309–335. Myotis. J Mammal 51:302–309. Nowak R.M. 1999. Walker’s Mammals of the World. 6th ed. Tieleman B.I. and J.B. Williams. 1999. The role of hyperthermia Johns Hopkins University Press, Baltimore. in the water economy of desert birds. Physiol Biochem Zool O’Farrell M. and E.H. Studier. 1970. Fall metabolism in relation 72:87–100. to ambient temperature in three species of Myotis. Comp Tieleman B.I., J.B. Williams, and M.E. Buschur. 2002. Physi- Biochem Physiol 35:697–703. ological adjustments to arid and mesic environments in larks Pagel M.D. 1992. A method for the analysis of comparative (Alaudidae). Physiol Biochem Zool 75:305–313. Turbill C., B.S. Law, and F. Geisera. 2003. Summer torpor in data. J Theor Biol 156:431–442. a free-ranging bat from subtropical Australia. J Therm Biol Pinshow B., M.A. Fedak, D.R. Battles, and K. Schmidt-Nielsen. 28:223–226. 1976. Energy expenditure for thermoregulation and loco- Webb P.I. 1995. The comparative ecophysiology of water balance motion in emperor penguins. Am J Physiol 231:903–912. in microchiropteran bats. Symp Zool Soc Lond 67:203–218. Poole E.L. 1936. Relative wing ratios of bats and birds. J Mam- Webb P.I., J.R. Speakman, and P.A. Racey. 1995. Evaporative mal 17:412–413. water loss in two sympatric species of vespertilionid bat, Porter W.P. 1969. Thermal radiation in metabolic chambers. Plecotus auritus and Myotis daubentonii: relation to foraging Science 166:115–117. mode and implications for roost site selection. J Zool (Lond) Reeder W.G. and R.B. Cowles. 1951. Aspects of thermoregu- 235:269–278. lation in bats. J Mammal 32:389–402. Williams J.B., A. Munõz-Garcia, S. Ostrowski, and B.I. Tiele- Riskin D.K. and M.J. Pybus. 1998. The use of exposed diurnal man. 2004. A phylogenetic analysis of basal metabolism, total roosts in Alberta by the little brown bat, Myotis lucifugus. evaporative water loss and life-history among foxes from Can J Zool 76:767–770. desert and mesic regions. J Comp Physiol B 174:29–39. Rodrı´guez-Durań A. 1995. Metabolic rates and thermal con- Willis C.K.R., C. Turbill, and F. Geiser. 2005. Torpor and thermal ductance in four species of Neotropical bats roosting in hot energetics in a tiny Australian vespertilionid, the little forest caves. Comp Biochem Physiol 110:347–357. bat (Vespadelus vulturnus). J Comp Physiol B 175:479–486. Schmid W.D. 1976. Temperature gradiance in the nasal passage Wimsatt W.A. 1970. Biology of Bats. Academic Press, New York. of some small mammals. Comp Biochem Physiol 54:305–308. ˙˙ Withers P.C. 1977. Measurement ofVo22 ,Vco and evaporative Schmidt-Nielsen K. 1964. Desert Animals. Oxford University water loss with a flow-through mask. J Appl Physiol 42:120– Press, Oxford. 123. ———. 1997. Animal Physiology. 5th ed. Cambridge Univer- Wojciechowski M., M. Jefimow, and E. Te˛gowski. 2006. En- sity Press, Cambridge. vironmental conditions rather than season determine torpor Schmidt-Nielsen K., F.R. Hainsworth, and D.E. Murrish. 1970. use and temperature selection in large mouse-eared bats Counter-current heat exchange in respiratory passages: effect (Myotis myotis). Comp Biochem Physiol A (forthcoming). on water and heat balance. Respir Physiol 9:263–276. Wolfram S. 1999. The Mathematica Book. 4th ed. Cambridge Scholander P.F., R. Hock, V. Walters, F. Johnson, and L. Irving. University Press, Cambridge.