Temperature regulation in subtropical tree bats

Camp. Eiochem. Physiol. Vol. 104A,No. 2, pp. 321-331, 1993

0300-9629/93 166.00+ 0.00 Pergamon Press Ltd

Printedin Great Britain

TEMPERATURE

REGULATION IN SUBTROPICAL TREE BATS

MICHEL GENOUD Department of Zoology, University of Florida, Gainesville, FL 32611, USA (Received 8 May 1992; accepted 12 June 1992) Abstract-l. Rate of metabolism and temperature regulation were studied in five species of subtropical tree bats (Lusiurus seininolus, L. borealis, L. intermedius, L. cinereus and Nycticeius humeralis). 2. All species showed two states while resting below thermal nebtrality: normothermia and torpor. Below @-YC, bats in torpor maintained an intermediate body temperature. 3. Basal rate of metabolism was lower than expected on the basis of body mass (4478%) and average body temperature in the normothermic state ranged between 32.5 and 357°C. 4. Lasiurines have a high thermogenic capacity. 5. The metabolic and thermoregulatory traits studied in tree bats are generally similar to those of non-tropical bats roosting in caves and buildings.

INTRODUCI’ION

metabolic and thermoregulatory traits, which are related to their particular roosting habits (Davis, In contrast to other temperate zone bats, North 1970; McNab, 1974). Both cave bats and tree bats American tree bats of the genus Lasiurus roost singly, from temperate and subtropical regions are prone to or in small groups, in tree foliage (Barbour and enter torpor when exposed to low ambient temperaDavis, 1969). They are much more exposed to ture in the laboratory (Davis and Reite, 1967; Hock, climatic fluctuations than bats of the same regions 1951; Kulser, 1965; McNab, 1974, 1982; Reite and roosting in buildings or caves. In subtropical regions, Davis, 1966). When ambient temperature approaches tree bats may have to cope with high environmental 0°C they react by slightly increasing their rate of temperatures on hot summer days as well as with metabolism, thereby avoiding tissue freezing, but a temperatures well below freezing on cold winter further decrease in ambient temperature may elicit nights, although many of them may roost within epiphytic clumps of Spanish moss (7’iZZunrisiu various responses, including complete arousal (Davis usneoides; Constantine, 1958; Jennings, 1958). and Reite, 1967; Hock, 1951; McNab, 1974; Reite and Davis, 1966). Davis and Reite (1967) observed Moreover, they may be exposed to large thermal that in contrast to several other vespertilionids, variations even during a single day. L. borealis was able to remain in torpor when ambiIn subtropical regions of the southeastern United States (Florida), three species of Lusiurus can be ent temperature was changed, even at temperatures slightly below 0°C or as high as 15°C. Shump and found at all seasons: the Seminole bat (L. seminolus), Shump (1980) showed that the fur of L. borealis and the yellow bat (L. intermedius) and the red bat L. cinereus has a higher insulative value than that of (L. borealis). The hoary bat (L. cinereus) appears colonial species roosting in caves or buildings. Furthere only during its migrations (Zinn and Baker, thermore, McNab (1974) found that during torpor, 1979). Several other species, including the evening bat (Nycticeius humerulis), also use trees as roosts to N. humerulis has a lower rate of metabolism than several subtropical cave bats. This author suggested some extent (Barbour and Davis, 1969), but they that a low rate of metabolism (relative to that of other often select more protected sites, such as holes in the non-tropical bats) may be a common feature of tree tree trunk or crevices under bark. In such roosting bats, enabling them to save energy while in torpor at sites, thermal variations are often attenuated and bats potentially higher ambient temperatures. Indeed, benefit from the insulative value of the wood or bark wintering subtropical tree bats are often exposed to (Kurta, 1985; Kurta et al., 1987). mild temperatures (e.g. more than loOC), and may The few observations available on temperature therefore use their energy reserves at a relatively high regulation in bats of the genera Lusiurus and rate while in torpor. Nycticeius suggest that tree. bats may have distinctive In this paper, the rate of metabolism and temperature regulation of tree bats (genera Lasiurus and Present address: Institute of Zoology and Animal Ecology, Nycticeius) from subtropical North America are University of Lausanne, Lausanne, Switzerland (Tel. 4121-692-2453; Fax 021-692-25-40). investigated at ambient temperatures ranging 321

MICHEL GENOUD

322

between -10 and 40°C. The paper is intended to provide physiological data that can bc compared with data from other bat species, especially from those roosting in buildings or caves (Hock, 1951; McNab, 1974). It is not intended to describe the “thermoregulatory abilities” of tree bats in the wild, since environmental factors other than ambient temperature (e.g. the presence of protective and insulating material as wood or bark in N. humeralis, and Spanish moss or foliage in lasiurines) may extensively modify the thermal behavior of bats (Kurta et al., 1987). MATERIALS AND METHODS

Animals

This study is based on 21 L. seminolus, 3 L. borealis, 2 L. intermedius, 1 L. cinereus and 29 N. humeralis caught between April 1984 and June 1985 in Florida. L. seminolus and N. humeralis were captured at all seasons, L. borealis was caught only in summer, and L. cinereus and L. intermedius were caught only in winter. Most bats were mist-netted over creeks and ponds in northern Florida (counties of Alachua, Marion and Columbia). However, 15 evening bats were also captured by hand in attics (one on Captiva Island, Lee county and 14 in Inverness, Citrus county), and the two yellow bats were found on the soil under a tree in Gainesville (Alachua county) on two very cold mornings. All bats could be maintained in captivity for several weeks (up to 5 months) in a healthy state and most of them were released afterwards. They were kept under a natural photoperiod regime and at an ambient temperature ranging between 20 and 24°C. The Lasiurus were kept individually in large cloth bags and were permitted to fly in a room every other night. Cloth bags were chosen after two unsuccessful attempts to maintain these bats in cages. Presumably cloth bags resemble somewhat Spanish moss, in which many Lasiurus hide during the day. The Nycticeius were kept in small groups (up to 10 individuals) in a cage. The bats were watered and fed by hand three times each night on mealworms and a vitamin supplement for insectivores. Various insects were also given when available. Except for pregnant females, body mass could be maintained within a range of 1.5 g around initial mass at capture. Respirometric measurements

All measurements were made during the first month of captivity, but none was made on the day following capture in order to let bats adjust to captivity. Furthermore, bats were never used for more than two measurements each day. The rate of metabolism was measured as oxygen consumption (Vo,) in an open air-flow respirometer. The bats were placed in a 1 litre metabolic chamber submersed in a water (or antifreeze solution) bath at a regulated temperature. Experimental temperature ranged between - 10 and 40°C in L. seminolus,

- 10 and 37.5”C in L. intermedius and -3 and 40°C in N. humeralis and between 15 and 37.5”C in L. borealis. Inside the chamber, a vertical screen and small branches placed at the top allowed the bats to roost in their natural posture: N. humeralis clung to the screen and the lasiurines rested suspended at the top branches. Ambient temperature (T,) within the metabolic chamber was measured with a thermocouple connected to a telethermometer (BAT-& Bailey, NJ, U.S.A.). Air was pushed through the chamber by a pump at a regulated flow rate of 35-320 ml/min STPD, depending on the ambient temperature within the chamber. After the chamber, CO, and water were removed from the outcoming air with soda lime and silica gel, before the flow rate (I’) was measured with a flowmeter (Sho-Rate R-2- 15 A, AA or AAA, Brooks, PA, U.S.A.) which was calibrated with a Brooks Vol-UMeter gas calibrator. After the flowmeter, part of the air was passed through an Applied Electrochemistry Oxygen Analyser (Ametek, PA, U.S.A.) in order to measure its O2 concentration (F,O,). Oxygen concentration was plotted continuously with a chart recorder. Mass-specific oxygen consumption (ml OZ/g hr) was calculated according to Depocas and Hart (1957) as: between

L. cinereus, between

v

=

02

(40, - ho,) V (1- F,o,)m

where m is body mass. FI02 is the O2 concentration of the air entering the chamber and was measured before and after the bat was placed into the chamber. Experiments were carried out between 8 a.m. and 7 p.m., and lasted between slightly more than 2 and 10 hr. Bats had no access to food or water during runs. After having been placed into the respirometric chamber, they were left for an initial period of 2 hr to adjust to their new thermal environment. Activity periods (checked visually) were never considered in the data. The bats were removed from the metabolic chamber after having been at rest for at least 30 min and only if their oxygen consumption was relatively stable for at least 15 min. This restriction was necessary since resting bats often exhibited a variable oxygen consumption (see below). Less than 30 set after removal of the bats, their body temperature (Tb) was measured by inserting a thermocouple (connected to the BAT-8 telethermometer) 2 cm deep into the rectum. Mass-specific thermal conductance (C, ml 0,/g hr “C) was calculated for corresponding values of V02’ Tb and T, according to McNab (1980) as:

C = VoJV,-

TJ

Defined in this way, C corresponds to “wet thermal conductance” (McNab, 1980), which includes heat loss by evaporation. Complementary experiments

Experiments were carried out on two successive 24 hr periods in order to assess the effect of the

323

Thermoregulation in tree bats specific dynamic action of food on the estimation of basal rate of metabolism (BMR). Four L. seminofus were used. During the first night, the bats were fed three meals of 1.2 g mealworms each, the last meal being given shortly before midnight. At 10a.m. the following morning, they were introduced into the metabolic chamber at an ambient temperature of 35’C (i.e. within thermal neutrality, see below) and their BMR was measured as soon as a stable and minimal oxygen consumption was recorded (between 12 a.m. and 1 p.m.). Individuals were assumed to be in a postabsorptive state, as they had not taken any food for more than 12 hr (control rate of metabolism). On the following night, the same individuals were given a same amount of mealworms, but the last meal was given at 8 a.m. The bats were again introduced into the metabolic chamber at 10 a.m. and values of oxygen consumption within thermal neutrality were taken every hr after a 1 hr period of equilibration. The experiments were stopped after 2 p.m. Experiments were also carried out to check for a possible influence of the duration in captivity on the rate of metabolism. Several f.. seminolus were measured on two or three occasions at the same temperature (either 25 or 35°C) during their first month of captivity. These experiments were suggested by previous observations made by Studier and Wilson (1979) showing that captivity has an effect on the rate of metabolism and thermoregulatory patterns of a tropical frugivorous bat, Artibeus

C = 1.02 m -“.5 (C in ml 0,/g hr “C and m in g; Herreid and Kessel, 1967). Intraspecific variations in BMR in L. seminolus and N. humeralis have been examined in detail with analyses of variance in a previous paper (Genoud, 1990). In both species, BMR is influenced by season, but not by sex. Similar analyses have been carried out on C and normothermic Tb, except that the effect of season on these parameters could not be tested in N. humeralis because this species usually did not maintain normothermia in autumn and winter. Since neither season (in L. seminolus) nor sex appeared to have a significant effect on the two parameters, average values of Tb and C from all individuals within each species were pooled. RESULTS

Within the metabolic chamber, bats mostly remained inactive, although short periods of moderate activity were regularly observed, during which bats were self-cleaning or changing posture and position within the chamber (e.g. Fig. 1, metabolic peaks in curves 2 and 5). However, they often exhibited a variable oxygen consumption even while resting motionless (e.g. Fig. 1, curve 1). A large number of experiments (358 on L. seminolus, 38 on L. borealis, 85 on L. intermedius, 17 on L. cinereus, and 348 on N. humeralis) was therefore necessary to detect consistent metabolic and thermoregulatory patterns as a function of ambient temperature.

jamaicensis. Data analysis

The effect of ambient temperature on the physiological parameters measured was tested using linear regression analysis. Two indications were used to estimate the lower limit of the thermoneutral zone (for details, see Bonaccorso et al., 1992). First, the V,, of bats mamtaining the normothermic state (i.e. maintaining a high body temperature at rest) was plotted against TO.Two lines may describe these data, one within the thermoneutral zone and the other below thermal neutrality. Thus the two lines that best fit the data were determined using a least squares criterion (Nickerson et al., 1989) and their intersection was calculated. Second, the temperature above which the thermal conductance of normothermic bats clearly increases above the minimum level was identified. An average between results obtained by these two methods was used. Throughout this paper, n refers to the number of measurements and N to the number of individuals. Species averages for basal rate of metabolism, minimal thermal conductance and normothermic Tb were calculated on the basis of individual averages. BMR and minimal thermal conductance were compared to the values expected on the basis of body mass, using the allometric relationships BMR = 3.42 m-o.2s (BMR in ml 0,/g hr and m in g; Kleiber, 1961) and

0

I

0

1

2

3

4

J

Hours after beginning of experiment Fig. 1. Oxygen consumption (V,) of Lusiurus seminofus during experiments at different ambient temperatures. Measurements at the end of the runs: I: T. = 25.5”C, Tb= 35.9”C, m = 11.6g, normothermia. 2: To = 34.9”C, thermal neutrality, Tb = 36.6”C, m = 11.5 g. 3: T, = 25”C, T, = 26.5”C, m = 9.3 g, torpor. 4: T, = 14.9”C, Tb = 15.2”C, m =9.1g, torpor. 5: To= -7.5”C, T,,=35.7”C, m=9.2g, normothermia. 6: T, = -4.5”C, T,, = 1l.o”C, m = 8.3 g, torpor. 7: T, = 0.5”C, Tb = 3.4”C, m = 8.5 g, torpor.

324

MICHEL GENOUD (BMR,

Table 1. Lower critical temperature (r,<, “C), Q,, (mean 5 SD, n). and basal rate of metabolism ml 0,/g br; mean + SD, N) in tree bats of the genera Lasiurus and Nycticeius Species

r,

Q,,

L. seminolus

32

2.74 * 0.09. 76

L. borealis L. cinereus L. intermedius N. humeralis

32 31 34 32

2.40 + 0.08, I12

lit*

%t

SeasonS

1.38 + 0.12,9 1.15*0.13, 10

9.8 f 1.3 9.0 & 1.0

71 58

s,a w

1.43 + 0.25, 2 0.83, I 0.81 f 0.21, 2 1.19 * 0.15, 18 0.82i:O.l3,6

12.2 f 1.2 19.9 22.6 f 3.9 9.0 * I. I 11.1i1.0

78 51 52 60 44

s w w s a,w

BMR

*Body mass corresponding to the BMR estimate. tBMR expressed as a percentage of the value expected on the basis of body mass (Kleiber, 1961). fThe seasons to which the BMR value is referring are indicated: s = spring and summer, a = autumn, w = winter.

When exposed to different ambient temperatures, bats of all species responded as heterotherms showing the ability to be endothermic homoiotherms or to enter torpor. Below the thermoneutral zone, two physiological states could be distinguished, in which resting bats exhibited a relatively stable oxygen consumption for periods of 15 min or more. These states, which are referred to as the normothermic state and torpor are illustrated in Fig. 1. In many cases, periods of variable oxygen consumption in resting bats clearly corresponded to transitions between these two states (Fig. 1, curve 1). In the normothermic state, resting bats maintained a high r,, usually above 30°C by an increased rate of metabolism (Fig. 1, curves 1 and 5). During torpor at ambient temperatures above a threshold value lying between 0 and 5°C Tb dropped near the ambient temperature, oxygen consumption stabilized at a low level and a highly characteristic pattern of respiratory cycles was observed, the length of which increased with decreasing TO(Fig. 1, curves 3 and 4). At T, values below &SC, all bats appeared able to enter torpor. Torpid bats regulated Tb at some intermediate level by an increased oxygen consumption, and did not show respiratory cycles (Fig. 1, curves 6 and 7). During two experiments at -9.2 and - 10°C Seminole bats completely stopped breathing. When the oxygen concentration curve was flat and equal to the base line, these experiments were stopped. The bats were found on the floor of the metabolic chamber and were thought to be dead. Tb could not be measured since the body was hard and the temperature probe could therefore not be inserted into the rectum. However, after a few minutes at room temperature (23 and 24”C), the bats slowly resumed breathing and in the evening they flew actively. These bats were probably in a state of “supercooling”, as was described by Davis and Reite (1967) and Davis (1970).

and N. humeralis, lower BMR values (relative to body mass) were observed in winter, or autumn and winter, respectively (Table 1; see also Genoud, 1990). Because red bats were caught only in summer whereas yellow and hoary bats were caught only in winter, differences in BMR between these species do not necessarily represent interspecific differences. They may also be due to seasonal changes within species. As a consequence of specific dynamic action of the food, time elapsed since the last meal significantly affected the rate of metabolism of L. seminolus within thermal neutrality (Anova, P < 0.05; Fig. 2). Three hours after the last meal, the increase due to specific dynamic action amounted to about 40% of the BMR (control rate). Thereafter, this amount declined progressively, being negligible 3 hr later. Accordingly, care was taken not to feed the bats after midnight prior to standard experiments made at ambient temperatures within or close to the thermoneutral zone. No evidence could be found that time in captivity has an effect on BMR in L. seminolus. Among the 10 individuals measured two or three times at 35°C during their first month of captivity, BMR differed significantly between individuals (P < 0.05), but did not depend on time in captivity (Ancova, with individual number treated as a factor and time in captivity as a covariate; interaction term between the factor and the covariate was not significant). The same result was obtained with the 5 individuals measured two or three times below thermal neutrality (25°C).

Thermal neutrality, BMR and spectjic dynamic action

of food The thermoneutral zone extended from between 3 1 and 34°C to about 36°C depending on the species (Table 1). All species had a low average BMR (Table l), ranging between 44 and 78% of the value expected on the basis of body mass. In L. seminolus

0-7

control

0123456

Hours fitter meal Fig. 2. Oxygen consumption (Vo,) of resting Lusiurus seminolus (N = 4) within thermal neutrality (35°C) as a

function of time elapsed since last meal. Control correspond to the same individuals in a postabsorptive

values state.

Thermoregulation

in tree bats

325

Normothermia The propensity to maintain a high T, varied according to To, season and species. Lasiurines occasionally maintained the normothermic state at all To’s to which they were exposed, although they did so more frequently at a moderate T, (20-30”(Z) and below 5°C (Figs 3, 5 and 6). In N. humeraks normothermia was also more often observed at moderate temperatures but occurred only down to a T, of 5°C (Fig. 4). In addition, the normothermic state was more frequently observed in lasiurines than in N. humeralis and at least at a moderate T,, it occurred more often in spring and summer than during the winter. Such variations are difficult to interpret, since the propensity to maintain a high Tb is likely to be influenced by the experimental conditions. Normothermic bats usually maintained their Th above 30°C (Table 2) although a large variability was observed at low T,. For example, L. seminolus maintained Th between 31.5 and 38.3”C while resting at a T, close to 0°C (Fig. 3). In L. seminolus, L. borealis and L. cinereus, Tb did not depend on T, and averaged between 34.8 and 35.7”C (Table 2). In contrast, T,, varied significantly with T, in both L. intermedius and N. humeralis, and at a moderate

-10

0

10

30

40

T, (Z;” Fig. 3. Body temperature (Tb), oxygen consumption (V,,), and thermal conductance (C) of Lasiurus seminolus as a function of ambient temperature (TJ. Open circles: normothermia, and values within and above thermal neutrality. Closed circles: torpor at low T, where minimum and stable values were reached. Crosses: torpor at low To where V,, varied irregularly. Lines for normothermic bats: 1 = mean Tb, 2 = regression line for Vo, as a function of To, 3 = mean C. Curves for bats in torpor at T, values above 5°C (individual data points not represented): 4 = average T,, 5 = average V,,, 6 = average C. Dashed line: line of equality between Tb and T,,

-10

6

10

30

40

T* (OC;” Fig. 4. Body temperature (Tb), oxygen consumption (V,,), and thermal conductance (C) of Nycticeius humeralis as a function of ambient temperature (T,). Open circles: normothermia, and values within and above thermal neutrality. Closed circles: torpor at low T, where minimum and stable values were reached. Crosses: torpor at low T, where V, varied irregularly. Lines for normothermic bats: I = regression line for Tb as a function of T, , 2 = regression line for V,, as a function of T,, 3 = mean C. Curves for bats in torpor at T, values above 5°C (individual data points not represented): 4 = average Tb, 5 = average Vo, , 6 = average C. Dashed line: line of equality between Tb and T,.

ambient temperature, it was on average lower than in the three other species (32.5 and 33.6”C at a T, of 20-30°C Table 2). In all species, Vo, of normothermic bats increased with decreasing T, (Table 2), and T, explained most of the variance in Vo, (0.79 < r* < 0.98). However, residual variability, independent of T,, was appreciable especially at low T, (Fig. 3) and was mainly related to variations in the T, that was maintained. This is demonstrated by the significant correlation which was obtained in all species (P < 0.05) except L. intermedius (P = 0.15) between residuals of the regression analyses of Vs and Tb on T,. In L. seminolus, Vo, exceeded 20 ml 0*/g hr during some experiments below -5°C (Fig. 3). Although no attempt was made to estimate summit metabolism in any of the studied species, some observations suggested that summit metabolism was reached in L. seminolus. Between -5 and - 10°C normothermic Seminole bats shivered violently and often descended on the floor of the metabolic chamber where they curled in a ball shaped posture, whereby the uropatagium covered the belly and reached the chin. Two of them became supercooled. Furthermore, another bat exposed to -8°C exhibited a maximum

MICHEL GENOUD

326

c e

40 1 30 Ix+@+ +, 20 g:+ .

c”

,O

.

0

/

lb

,.,,.’

. :*s :A’” St ,/ ,’ 1-p-1-

OJ , -10

A

meP9Al ,/

0

10

A

20

30

rJ

40

-io

Ta (“‘9

b

Ib

3b

4’0

T, &”

Fig. 5. Body temperature (Tb), oxygen consumption (V,,), and thermal conductance (C) of Lusiurus intermedius as a function of ambient temperature (T,). Open circles: normothermia, and values within and above thermal neutrality.

Fig. 6. Body temperature (Tb) oxygen consumption (V,,), and thermal conductance (C) of Lasiurus borealis (circles and points) and L. cinereus (triangles) as a function of ambient temperature (T,). Open symbols: normothermia,

Closed circles: torpor at low T, where minimum and stable values were reached. Crosses: torpor at low T, where Vo, varied irregularly. Points: torpor at moderate temperatures. Lines for nonnothermic bats: 1 = regression line for T, as a

and values within and above thermal neutrality. Closed

function of T,, 2 = regression line for VP1 as a function of T,, 3 = regression line for C as a function of 7’,. Dashed line: line of equality between T,, and T,.

Vo2 of 22.2 ml 0,/g hr for l-2 hr, and subsequently decreased Vs in a rather irregular way, suggesting that the bat was entering irreversible hypothermia. This experiment was stopped to avoid death of the bat.

Below thermal neutrality, thermal conductance (C) in the normothermic state was significantly related to T. only in L. intermedius (Table 2). In this species the minimal thermal conductance has been calculated on the basis of data available below 2O”C, whereas in the

L. seminolus

VO2

0.29 + 0.11 ml 0*/g hr “C was measured

L. cinereus

in the only

hr “C) of normothermic

L. intermedius

N. humeralis

n, P,

12 Mean T,

:, 0.10,35.1 k 0.6(15)

;:0.25, 35.7 f 0.8 (2)

gnSO.38, 34.8 (1)

30.8 + 0.10 T, 24, <0.001,0.66 32.5 + 0.1 (2)

30.8 + 0. I I To 36, <0.001,0.25 33.6+ 1.2(18)

f(TJ

14.07 - 0.41 T, 92, <0.001,0.90

9.50 - 0.25 T, IO, <0.001,0.79

7.93 - 0.23 T, 10,
6.21 - 0.16 7’. 34,
12.79 - 0.37 T, 53, <0.001,0.83

:;, 0.32, 0.39 f 0.05 (17)

;:0.,2,0.30 f 0.04 (1)

gnso.35, 0.23 + 0.02 (1)

0.20 + 0.003 To 24,0.001,0.32 0.21 * 0.05 (2)

;1,0.13,0.47 f 0.09 (13)

KT,) n, P, 12 Mean minimal

For

L. borealis

f(TcJ

n, P, r* C

other species, a general average has been calculated. intermedius, L. cinereus and L. borealis had a thermal conductance close to the value expected on the basis of their body mass (97, 101 and 105% of expected, respectively). L. seminolus (116%), and especially N. humeralis (137%) had a higher thermal conductance. However, it must be stressed that most values for N. humeralis have been obtained during spring and summer. A thermal conductance of L.

(Tb. “C), oxygen consumption (Vo,, ml 0,/g hr) and thermal conductance (C, ml 0,/g tree bats of the genera Lariurus and Nycriceius Mow thermal neutrality

Table 2. Body temperature

Th

triangles and points: torpor. Lines for normothermic L. borealis: la = mean Tb, 2a = regression line for V,, as a function of To, 3a = mean C. Lines for normothermic I,. cinereus: lb = mean Tbr 2b = regression line for V,, as a function of T,, 3b = mean C. Dashed line: line of equality between Tb and T,.

C

each variable, the regression line r&tins ~ the varial _____ __.. _ ._._ _.._ble to ambient temperature is given [f(TJ, unless the regression is not significant (ns). The number of mea surements involved in the regression analysis (n), the significance level (P) and the goodness of fit (r2) are indicated. For TA and C, the mean + SD of the individual means is also given, and the number of individuals involved is indicated in brackets. _

Thermoregulation

three evening bats which maintained occasionally normothermia in autumn. This is 96% of the expected value. Also, two evening bats caught in summer had their fur interrupted by spots of naked skin; their thermal conductance corresponded to 135 and 179% of the expected value, respectively. Thus thermal conductance may vary seasonally in the evening bat, although more data are needed to demonstrate this statistically. Torpor

At a T, above 0-5”C, the average Vo, of bats in torpor had to be estimated by integrating the oxygen consumption curves over periods of l-4 hr, because of the regular respiratory cycles exhibited in this state (Fig. 1, curves 3 and 4). This was done only for L. seminolus and N. humeralis, for which enough data were available. Average V,, in torpor was related to T,, and this relationship was quasi exponential. A linear regression model fitted to the logarithmically transformed values of Vo, as a function of T, (Fig. 7) explained a high proportion of the total variance (0.86 < r2 < 0.92). Adding a quadratic term to the model would increase r2 by only 0.01-0.014. A comparison of the average V,, of L. seminolus and N. humeralis in torpor was done by covariance analysis, where species was a factor and T, a covariate. The interaction term species x T0 was significant (P = 0.006) which demonstrates that these two species exhibit a different thermal dependence of the rate of metabolism in torpor. They have a similar rate at a low T,, but at high temperature, N. humeralis

0.5,

Laslurvs seminolus o.o-

,

-1.5 0.5

Nyctlcdus humeralis 0.0 1 ry

2

-0.5

f -1.0

1

-1.5 '

I 5

10

15

&”

25

30

T,

and Nycticeius humeralis in torpor as a function of ambient temperature (To). Regression lines are: log V,,, = - 1.482 + 0.047 To for L. seminolus, and log Vo, = - 1.423 + 0.040 To for N. humeralis.

327

exhibits a lower rate than L. seminolus. The relationship between logarithmically transformed V,, values and Tb was also quasi linear, as was expected since the rate of metabolism in torpor should follow a Q,, function. A best fit line was obtained by linear regression, and Q,, was estimated as the slope of this line. It was slightly lower in N. humeralis than in L. seminolus (Table 1), which parallels the above results on the relationship between V,, in torpor and T, in these species. The Tb measured within 30 set of removal of bats in torpor was always very close to ambient temperature (Figs 3-6). However, arousal is known to be very fast in small bats (e.g. see Heldmaier, 1969; Kulzer, 1965) and therefore, a slight, but systematic over estimation of T,, (possibly by a few tenths of a degree) is probable. This should not affect significantly the relationship between Vo, and Tb and the estimation of Q,,, but may lead to a strong underestimation of thermal conductance in torpor. The low conductance values which have been calculated for individuals in torpor at low T, (Figs 3-6), may have resulted from the partial arousal of the bats when their Tb was measured. At ambient temperatures below 0-5°C bats in torpor increased their rate of metabolism (Figs 3-6). At such low ambient temperatures, V,, often varied irregularly for several hours. These irregular fluctuations were probably related to a variable T6, since they occurred in completely inactive bats. Experiments presenting such irregular fluctuations were stopped after 4-7 hr, regardless of the particular form of the oxygen consumption curve and the final values were recorded (Figs 3-5, crosses). In many other experiments at low ambient temperature, Vo2 reached a minimum and relatively stable level that was maintained for several hours (Fig. 1, curves 6 and 7; Figs 3-6, closed circles and triangles). In L. seminolus, L. intermedius and L. cinereus, for which enough data were obtained at low T,, these minimum Vo, values increased as ambient temperature decreased below 5°C (Figs 3, 5 and 6). The precise form of the curve describing the relationship between Vo, and T, cannot be specified with the available data, but it must be stressed that at least in L. seminolus, Tb also increased as T, decreased (Fig. 3). Thermal conductance tended to be lower than in normothermic bats (Figs 3-6), but again it may have been under estimated due to rapid arousal of the bats when removed from the metabolic chamber. Reactions

I

Fig. 7. Oxygen consumption (Vo, in ml 0, /g hr) of Lusiurus seminolus

in tree bats

to high ambient temperature

attempt was made to assess the maximum ambient temperature which can be sustained for a certain time by the different species. The highest ambient temperature to which the bats were exposed was 40°C in L. seminolus and N. humeralis and 37.5”C in the other species. Exposure to these high temperatures was tolerated for the duration of the runs No

328

MICHEL GENOUD

(2.5-3 hr). Bats mostly remained inactive and became hyperthermic, but generally maintained their T, less than 1S”C above T, (Figs 3-6). L. seminolus and N. humeralis maintained similar differences between body and air temperatures. For example, at a TO of 37-38°C this difference amounted to 1.0 + 0.3”C (N = 12) in Seminole bats and 1.2 f 0.4”C (N = 16) in evening bats. At 39-40°C these values dropped to 0.8 f 0.4”C (N = 6) and 0.6 f 0.4”C (N = 9), respectively. These temperature differentials do not differ significantly (t-test, P > 0.05). Not enough data was obtained on the three other species to permit a statistical comparison. Thermoregulation at high ambient temperature was achieved by postural changes and, presumably, by an intense evaporative water loss. Above 37°C the bats usually spread their legs and uropatagium and held their wings slightly apart from the body. Occasionally, the posterior part of their abdomen was found wet when they were removed from the metabolic chamber. DISCUSSION

Metabolic

rate and thermoregulation

in Lasiurus

and

Nycticeius The present results only partly corroborate previous observations which have been published on the rate of metabolism and temperature regulation of North-American tree bats (Davis and Reite, 1967; McNab, 1974; Reite and Davis, 1966). The rates of metabolism reported by McNab (1974) for N. humer alis in torpor are much lower than those measured in the present study. For example, this author obtained a rate of metabolism lower than 0.05 ml 0,/g hr at 20°C and of about 0.2 ml 0,/g hr at 30°C. In the present study, the corresponding rates are 0.24 ml 0*/g hr and 0.6 ml 0,/g hr, respectively. Such a large difference is difficult to understand. At least at a moderate or high ambient temperature (e.g. 20-3O”C), it cannot be explained by the fact that the present values correspond to an average obtained by integrating the oxygen consumption curve, for at such temperatures the regular fluctuations in oxygen consumption represented only minor variations around this average (Fig. 1, curve 3). Part of the difference may reflect the different nutritional conditions of bats. The increase in rate of metabolism exhibited by torpid lasiurines (L. seminolus, L. intermedius and L. cinereus) when environmental temperature was lowered below 0°C confirms observations previously made by Reite and Davis (1966) and Davis and Reite (1967) on L. borealis. These authors did not measure oxygen consumption, but they showed that red bats exposed to temperatures between 0 and -5°C increase their heart rate and respiratory rate, and increase the difference between rectal and ambient temperature. However, their results do differ quantitatively from the present ones. For example, at an ambient temperature of -5°C red bats maintained

their Tb only l-4°C above ambient (Davis and Reite, 1967; Reite and Davis, 1966), whereas in the present experiments lasiurine bats maintained their Th at a higher level, usually between 5 and 15°C depending on ambient temperature. This discrepancy may reflect interspecific differences but it may also be due to differences in the experimental procedure. Reite and Davis (1966) and Davis and Reite (1967) restrained their bats by tapping the wings to wooden blocks during the experiments, a procedure which may influence the thermoregulatory abilities of the bats. Further observations on the metabolic and thermal reactions of unrestrained L. borealis to subfreezing temperatures are needed. In the present study, supercooling appeared as exceptional (only two observations). It is probable that this phenomenon is exhibited only by bats that have exhausted their energy reserves. Davis and Reite (1967) also observed supercooling when they maintained their restrained bats for several hours at subfreezing temperatures. Active rewarming from this stage is apparently not possible. In the present study, the two L. seminolus which became “supercooled” rewarmed passively after the experiments, when they were left at room temperature. However, Davis and Reite (1967) showed that supercooled bats are also able to resume breathing when ambient temperature is increased to 5°C. As these authors already concluded, it is clear that supercooling can only be a transient phenomenon, whereby the survival of the bat depends on the imminence of the next increase in ambient temperature. Lasiurine bats have a high thermogenic capacity. The highest mass-specific rates of metabolism measured in L. seminolus (slightly above 20mlO,/g hr, Fig. 3) are similar to the summit metabolism of soricine shrews of a similar size (e.g. 20.3 ml 0,/g hr for Sorex coronatus and 17.5 ml OZg hr for Neomys ,fodiens; Sparti, 1992). However, to the contrary of soricine shrews, lasiurine bats do not have a high basal rate of metabolism. Several authors (e.g. Packard, 1968; Wunder, 1985; Dawson and Olson, 1987) have suggested that one function of having a high BMR for a small mammal may be to permit higher rates of thermogenesis, and that having a high BMR may be a prerequisite to having a high summit metabolism. The present data on lasiurine bats clearly do not support this hypothesis, since these mammals associate a high thermogenic capacity with a low BMR. In their comparative studies, neither Koteja (1987) nor Sparti (1992) could find a significant correlation between BMR and maximum rate of metabolism among mammals and shrews, respectively, and therefore, the generality of this hypothesis appears at best doubtful. Thermal

ecology

of subtropical

tree bats

Care must be taken when extrapolating the present laboratory results to the thermoregulatory abilities of tree bats in the field. In Georgia, Constantine (1958)

Therrnoregulation discovered many lasiurines hanging in or on pendant clumps of Spanish moss and in Florida, Jennings (1958) purchased many from Spanish moss gatherers. As is the case for wood or bark (Kurta, 1985; Kurta et al., 1987), Spanish moss probably retards heat loss from the bats. In addition, roosting sites are selected according to several criteria, including thermal ones, such as radiation and air movements (Constantine 1958, 1966). Constantine (1958) also observed three L. seminolus being warmed by the sun and exhibiting greatly accelerated respiratory rates. Exposure to sun and/or wind, as well as the insulative value of Spanish moss, wood or bark may clearly modify the thermoregulatory patterns of tree bats in the wild. Nevertheless, the present results suggest that lasiurines loose energy at a high rate during episodes where ambient temperature is below 0°C despite the probable insulative value of Spanish moss. For example, the oxygen consumption measured at - 10°C in a resting Seminole bat of 9.5 g represents a rate of energy loss of about 1.5 kJ/hr, and a rate of fat use of approximately O.O4g/hr (Kleiber, 1961). This is a significant loss since several cold nights without feeding may follow in a row. For this reason, hibernation appears incompatible with the roosting behaviour of lasiurines. Wintering L. seminolus forage at rather low ambient temperatures. In this study, they have been netted at ambient temperatures as low as 10°C at dusk. This contrasts with previous observations suggesting that Seminole bats do not fly below 18°C (Jennings, 1958) or 21°C (Constantine, 1958) and with the suggestion that a high arousal temperature “protects” lasiurine bats from arousing too frequently (Davis, 1970). Arousal probably also depends on the amount of energy reserves that are still available to the bat. The overwintering strategy of N. humeralis is still poorly known. This species accumulates large amounts of fat in August both in northern Georgia (Baker et al., 1968) and in north-central Florida (Genoud, 1990, this study). This fat accumulation, together with occasional recoveries of banded bats at large distances south of the place they were tagged, led several authors to suggest that the species undergoes long migratory flights from northern parts of its range (Baker et al., 1968; Barbour and Davis, 1969; Watkins, 1972). In Florida, evening bats are present throughout the winter (Jennings, 1958; Bain and Humphrey, 1986) although they usually cannot be seen flying. These bats probably hibernate in sites which protect them from the most extreme temperatures, despite the relatively mild average ambient temperatures that predominate during winter (mean January temperature ranges between about 13 and 16°C in northern Florida; see McNab, 1974). During the present study, four female evening bats hibernated successfully in the laboratory. These bats were kept for three months, between 26 December 1984 and 26 March 1985, in a 1 m3 tank, in complete darkness and constantly humid air, and at a regulated

in tree bats

329

temperature of 15 f 1“C. Food was not provided, but water was constantly available within the tank. This observation shows that N. humeralis is at least able to hibernate at ambient temperatures as high as the average winter temperatures that prevail in northern Florida. Tree bats versus cave or house dwelling bats North-American tree bats do not tend to have a lower rate of metabolism than other non-tropical bats. Their BMR (44 to 78% of the expected value) is similar to that of attic- or cave bats, including Phyllostomidae (Macrotus californicus, 68%; Bell et al., 1986) Molossidae (Eumops perotis, 57% and Tadarida brasiliensis, 64%; Leitner, 1966; Licht and Leitner, 1967) and Vespertilionidae (Myotis yumanensis, 74%, Antrozous pallidus, 54% and Myotis chiloensis, 80%; Bozinovic et al., 1985; Licht and Leitner, 1967). In addition, the four studied lasiurines tended to have a slightly higher BMR than N. humeralis, although they exhibited the most extreme tree roosting behaviour. The rate of metabolism of tree bats in torpor is also roughly similar to that of temperate cave- or attic bats in torpor (Myotis lucifugus, Hock, 1951; Tadarida brasiliensis, Herreid, 1963; Myotis austroriparius and Pipistrellus subflavus, McNab, 1974). However, previous studies on cave bats have often reported steeper relationships between rate of metabolism and ambient temperature, suggesting high Q,, values. For example, the ratio of the rates measured at 30 and 10°C is equal to 21.6 in Myotis lucifugus (Hock, 1951) and 11.6 in Tadurida brasiliensis (Herreid, 1963). Similar values can be obtained from Fig. 4 in McNab (1974) for Myotis austroriparius and Pipistrellus subjlavus. this ratio is only 6.3 in N. humeralis and 8.7 in L. seminolus. It would be unwise to interpret this difference as a physiological difference between tree bats and cave bats. As ambient temperature decreases, the rate of metabolism of bats in torpor drops to such low levels, that measurement errors and/or methodological aspects (e.g. whether minimal or average oxygen consumption values are retained) become increasingly important. Also, the Q,, estimates obtained in this study for N. humeralis and L. seminolus (2.4 and 2.74) are similar to that reported for Myotis lucifugus (2.1) by Henshaw (1968). In their review of metabolic rate in animals, Robinson et al. (1983) found an average Q,, of 2.4 for homeotherms in general. No clear difference can be found either between tree bats and cave bats in the body temperature that is maintained in a normothermic state. Tree bats maintain an average Tb of 32.5-35.7”C. Myotis chiloensis tends to maintain a slightly higher normothermic body temperature (36.6 + 2.2”C; Bozinovic et al., 1985), but at “room temperature” (23-25”C), Tadarida brasiliensis, Antrozous pallidus and Myotis yumanensis exhibit a Tb of 32-38°C 32.5-36°C and 30-35.5”C, respectively (Licht and

330

MICHELGENOUD

Leitner, 1967). Also, between 10 and 30°C normothermic Plecotus auritus maintain an average Tb of 34%37.8”C, depending on ambient temperature (Speakman, 1988). In their comparative study of pelt insulation in North American bats, Shump and Shump (1980) have shown that in summer L. borealis and L. cinereus have a better insulated fur than several sympatric cave bats (Myotis lucifugus, M. keenii and Eptesicus fuscus). Although these results stress an important thermal characteristic of tree bats, they cannot easily be compared to data on thermal conductance obtained on live bats, since properties other than pelt insulation influence thermal conductance (e.g. patterns of peripheral blood circulation). The same is true for estimates based on cooling curves of dead bats. For example, using this technique, Kurta (1985) obtained a value of 0.0327 W/C for Myotis lucifugus (average body mass 8.9 g). Assuming an energetic equivalent for oxygen of 20.1 J/ml 0,) this corresponds to 0.66 ml 0,/g hr “C, or 193% of the expected value. Present estimates of the minimal thermal conductance of tree bats were equal to, or slightly higher than, the predicted values (97--137%). The highest values were found in N. humeralis, but L. seminolus also tended to have a thermal conductance above the expected value. Unfortunately, Myotis ch~~oen.~is is the only other non-tropical insectivorous bat for which an estimate of the minimal thermal conductance is available (0.41 ml OZ/g hr “C or 97% of the predicted value; Bozinovic et al., 1985), and this estimate has been obtained from the slope of the hne relating oxygen consumption and ambient temperature, a method which may give wrong values of minimal thermal conductance (McNab, 1980). As Davis and Reite (1967) and Davis (1970) pointed out, lasiurine bats are very efficient in coping with low environmental temperatures. They are able to remain torpid when exposed to temperatures well below 0°C and react by maintaining their body temperature at some intermediate level. Thus, they avoid both tissue freezing and the large expenditure associated with arousals and normothermia. They can also sustain temporary exposure to very low ambient ~peratures (- 1OC) through a high rate of thermogenesis. N. humeralis was equally able to remain in torpor when ambient temperature was decreased to -3°C. The few data available on cave bats suggest a variety of responses to subfreezing temperatures. When ambient temperature drops to slightly betow 0°C Myotis lucr~ugus and M. soda& are usually able to remain in torpor, and react by slightly increasing their rate of metabolism (Davis and Reite, 1967; Henshaw and Folk, 1966; Hock, 1951; Reite and Davis, 1966). In contrast, Eptesicus fuscus invariably arouses, whereas Pi~istre~l~s sub&us appears unable to arouse and dies within a few hours (Davis, 1970; Davis and Reite, 1967). These differences appear related to the overwintering strategy of the bats (Davis, 1970).

Acknowledgements-1

am grateful to Brian K. McNab, of the University of Florida, who provided room, equipment and invaluable help at all stages of this work. Frank J.

Bonaccorso, Willard W. Hennemann and Alexis Arends provided help for mist netting the bats, and the manuscript was improved by the suggestions of Frank J. Bonaccorso and Brian K. McNab. The research was supported by a grant of the “Fonds national suisse de la recherche scientifique”. REFERENCES Bain J. R. and Humphrey S. R. (1986) Social org~i~tion and biased sex ratio of the evening bat, Nycticeius humeraiis. Florida Sci. 49, 22-3

1.

Baker W. W., Marshall S. G. and Baker V. B. (1968) Autumn fat deposition in the evening bat Nycriceius humeralis. J. Mammal. 49, 314-317.

Barbour R. W. and Davis W. H. (1969) Bats of America. University Kentucky Press, Lexington, KY. Bell G. P., Bartholomew G. A. and Nagy K. A. (1986) The roles of energetics, water economy, foraging behavior, and geothermal refugia in the distribution of the bat, Macrotus californicus. J. camp. Physiol. 156B, 44-450. Bonaccorso F. J., Arends A., Genoud M., Cantoni D. and Morton T. (1992) Thermal ecology of moustached and ghost-faced bats (Mormoopidae) in Venezuela. J. Mammal. 73, 3655378. Bozinovic F., Contreras L. C., Rosenmann M. and TorresMura J. C. (1985) Bioenergetica de Myotis chiloensis (3Q&;,ptera: Vespertilionidae). Rev. Chilena Hist. Nat. 58, Constantine D. G. ($958) Ecological observations on lasiurine bats in Georgia. f. mammal. 39, 6470. Constantine D. G. (1966) Ecological observations on lasiurine bats in Iowa. J. Mammal. 41, 34-41. Davis W. H. (1970) Hibernation: ecology and physiological ecology. In Biology of Buts (Edited by Wimsatt W.-A.), Vol. 1. on. 265-330. Academic Press. New York. Davis W. H. and Reite 0. 8. (1967) Responses of bats from temperate regions to changes in ambient tem~rature. Biol. Bull. 132, 320-328. Dawson T. J. and Olson J. M. (1987) The summit metabolism of the short-tailed shrew Blarina hreuicauda: a high summit is further elevated by cold acclimation. Physiol. .

.

2001. 60, 631-639.

Depocas F. and Hart J. S. (1957) Use of the Pauling oxygen analyser for measurement of oxygen consumption of animals in open-circuit systems and in a short-lag, closedcircuit aunaratus. J. aoul. Phvsiol. 10. 388-392. Genoud M’(1990) Seasonal variations in the basal rate of metabolism of subtropical insectivorous bats (Nycficeius humeralis and Lasiurns seminolus): a comparison with other mammals. Revue suisse Zool. 97, 77-90. Heldmaier G. (1969) Die Thermogenese der Mausohr~edermaus (Myotis myofis Borkh.) beim Erwachen aus dem Winterschlaf. Z. vergl. Physiologie 63, 59-84. Henshaw R. E. (1968) Thermoregulation during hibernation: application of Newton’s law of cooling. J. theor. Biol. 20, 79-90.

Henshaw R. E. and Folk G. E. Jr. (1966) Relation of thermoregulation to seasonally changing mi~r~limate in two species of bats (iWyotis tucifugus and M. sodalis). Physiol. Zool. 39, 223-236.

Herrkid C. F. (1963) Metabolism of the mexican free-tailed bat. J. Cell camp. Phvsiol. 61, 201-207. Herreid C. F. and -Kessel B. (1967) Therms1 conductance in birds andmammals. Camp. Biochem. Physiol. 21,405-414. Hock R. J. (1951) The metabolic rates and body temperatures of bats. Biot. Bull. 101, 289-299. Jennings W. L. (1958) The Ecological Distribution of Bats in Florida Ph.D., University of Florida, Gainesville, FL. Kleiber M. (1961) The Fire of Life: An Introduction to Animal Energetics. John Wiley, New York.

Thermoregulation Koteja P. (1987) On the relation between basal and maximum metabolic rate in mammals. Camp. Eiochem. Physiol. 87A, 205-208. Kulzer E. (1965) Temperaturregulation bei Fledermausen (Chiroptera) aus verschiedenen Klimazonen. Z. vergl. Physiol. 50, l-34.

Kurta A. (1985) External insulation available to a nonnesting mammal, the little brown bat (Myotis lucifugus). Comp. Biochem.

Physiol. 82A, 413-420.

Kurta A., Johnson K. A. and Kunz T. H. (1987) Oxygen consumption and body temperature of female little brown bats (Myotis lucifugus) under simulated roost conditions. Physiol.

Zool. 60, 386-397.

Leitner P. (1966) Body temperature, oxygen consumption, heart rate and shivering in the California mastiff bat Eumops perotis. Camp. Biochem. Physiol. 19, 431-443.

Licht P. and Leitner P. (1967) Physiological responses to high environmental temperatures in three species of microchiropteran bats. Camp. Biochem. Physiol. 22, 371-387.

McNab B. K. (1974) the behavior of temperate cave bats in a subtropical environment. Ecology 55, 943-958. McNab B. K. (1980) On estimating thermal conductance in endotherms.. Phyiiol. Zool. 53, i45-156. McNab B. K. (1982) Evolutionarv alternatives in the physiological ecology of bats. In Eiology of Buts (Edited by Kunz T. H.), pp. 151-200. Plenum Publishing Corporation, New York. Nickerson D. M., Facey D. E. and Grossman G. D. (1989) Estimating physiological thresholds with two-phase regression. Physiol. Zool. 62, 866-887.

in tree bats

331

Packard G. C. (1968) Oxygen consumption of Microrus montunus in relation to ambient temperature. J. Mammal. 49, 215-220.

Reite 0. B. and Davis W. H. (1966) Thermoregulation in bats exposed to low ambient temperatures. Proc. Sot. exp. Biol. Med. 121, 1212-1215. Robinson W. R., Peters R. H. and Zimmerman J. (1983) The effects of body size and temperature on metabolic rate of organisms. Can. J. Zool. 61, 281-288. Shump K. A. and Shump A. U. (1980) Comparative insulation in vespertilionid bats. Camp. Biochem. Physiol. 66A, 351-354. Sparti A. (1992) Thermogenic capacity of shrews (Mammalia, Soricidae) and its relationship with basal rate of metabolism. Physiol. Zool. 65, 77-96. Speakman J. R. (1988) Position of the pinnae and thermoregulatory status in brown long-eared bats (Plecotus nuritus). J. therm. Biol. 13, 25-29.

Studier E. H. and Wilson D. E. (1979) Effects of captivity on thermoregulation and metabolism in Artibeus (Chiroptera: Phyllostomatidae). Camp. jamaicensis Biochem. Physiol. 62A, 347-350.

Watkins L. C. (1972) Nycticeius humeralis. Mammal. spec. 23, l-4.

Wunder B. A. (1985) Energetics and thermoregulation. In Biolopv of New World Microtus (Edited by Tamarin R. H), pp. 812-844. Special publication No. 8. The American Society of Mammalogists. Zinn T. L. and Baker W. W. (1979) Seasonal migration of the hoary bat, Lasiurus cinereus, through Florida. J. Mammal. 60, 634-635.