J Comp Physiol B (2000) 170: 153±162 Ó Springer-Verlag 2000

ORIGINAL PAPER

F. Geiser á R. M. Brigham Torpor, thermal biology, and energetics in Australian long-eared bats (Nyctophilus)

Accepted: 22 November 1999

Abstract Previous studies have suggested that Austra- had torpor metabolic rates that were up to ten times lian long-eared bats (Nyctophilus) di€er from northern- higher than those measured overnight. The steady-state hemisphere bats with respect to their thermal physiology torpor metabolic rate of thermoconforming torpid bats and patterns of torpor. To determine whether this is a showed an exponential relationship with body temper- general trait of Australian bats, we characterised the ature (r2 = 0.94), suggesting that temperature e€ects are temporal organisation of torpor and quanti®ed meta- important for reduction of metabolic rate below basal bolic rates and body temperatures of normothermic and levels. However, the 75% reduction of metabolic rate torpid Australian bats (Nyctophilus geo€royi, 7 g and between basal metabolic rate and torpor metabolic rate N. gouldi, 10 g) over a range of air temperatures and in at a body temperature of 29.3 °C suggests that metabolic di€erent seasons. The basal metabolic rate of normo- inhibition also plays an important role. Torpor meta- thermic bats was 1.36 ‹ 0.17 ml g)1 h)1 (N. geo€royi) bolic rate showed little or no seasonal change. Our study and 1.22 ‹ 0.13 ml g)1 h)1 (N. gouldi), about 65% of suggests that Australian Nyctophilus bats have a low that predicted by allometric equations, and the corre- basal metabolic rate and that their patterns of torpor are sponding body temperature was about 36 °C. Below an similar to those measured in bats from the northern air temperature of about 25 °C bats usually remained hemisphere. The low basal metabolic rate and the high normothermic for only brief periods and typically en- proclivity of these bats for using torpor suggest that they tered torpor. Arousal from torpor usually occurred are constrained by limited energy availability and that shortly after the beginning of the dark phase and torpor heterothermy plays a key role in their natural biology. re-entry occurred almost always during the dark phase after normothermic periods of only 111 ‹ 48 min Key words Bats á Body temperature á Metabolic (N. geo€royi) and 115 ‹ 66 min (N. gouldi). At air rate á Nyctophilus á Torpor bouts temperatures below 10 °C, bats remained torpid for more than 1 day. Bats that were measured overnight had Abbreviations BMR basal metabolic rate á C thermal steady-state torpor metabolic rates representing only conductance á MR metabolic rate á RMR resting 2.7% (N. geo€royi) and 4.2% (N. gouldi) of the basal metabolic rate á Ta air temperature á Tb body metabolic rate, and their body temperatures fell to temperature á TMR torpor metabolic rate á TNZ _ minima of 1.4 and 2.3 °C, respectively. In contrast, bats thermoneutral zone á V O2 rate of oxygen consumption measured entirely during the day, as in previous studies,

Introduction

Communicated by I.D. Hume With the exception of rodents, bats are the only non- marine placental mammals endemic to Australia. It is F. Geiser (&) believed that the ®rst bats migrated to Australia from Zoology, School of Biological Sciences, south-eastern Asia in the mid Tertiary and evolved into University of New England, Armidale NSW 2351, Australia several distinct Australian genera (Hall 1984). Despite e-mail: [email protected] this evolutionary background and the long separation Fax: +61-2-67-733-814 from bats in other parts of the world, little detailed work has been conducted to determine whether physiological R. M. Brigham Department of Biology, University of Regina, adaptations of Australian bats di€er from those found Regina, Saskatchewan, Canada on other continents. It is possible that the aridity, 154 unpredictable rainfall, and infertile soils characteristic of undergo short torpor bouts usually lasting for less than a the Australian continent are re¯ected in energetic and day in summer, and multi-day torpor bouts (i.e. hiber- thermoregulatory features of bats as have been observed nation) in winter (Hall 1982; Ellis et al. 1991; Brigham in marsupials (Dawson 1989; Lovegrove 1996). It is and Geiser 1998; Turbill et al. 1999) similar to northern- known that Australian microbats (Microchiroptera) as hemisphere species (Audet and Fenton 1988). As torpor well as small megabats (Megachiroptera) use torpor bout duration in many heterothermic mammals appears extensively to reduce energy expenditure (Morrison to be related to Tb and MR (Twente and Twente 1965; 1959; Kulzer et al. 1970; Morris et al. 1994; Geiser et al. French 1985; Geiser and Kenagy 1988; Barnes and 1996; Hosken 1997; Bartels et al. 1998; Coburn and Ritter 1993), it is possible that bats either change the- Geiser 1998; Hosken and Withers 1999). This is most rmoenergetics with season, or that bout duration is to a likely because of their large relative surface area, high large extent controlled by external factors such as Ta and costs for thermoregulation at low air temperature (Ta), food availability. To our knowledge there is no pub- and unpredictable or temperature-dependent food sup- lished literature on whether the seasonal change in tor- ply as in bats from Europe and North America (Hock por bout duration of bats in general re¯ects seasonal 1951; Pohl 1961; Henshaw 1970; Lyman 1970; Kunz changes in MR during torpor. 1982; Fenton 1983; Thomas et al. 1990; Speakman et al. We quanti®ed patterns of thermoregulation, ener- 1991; Barclay et al. 1996). getics, and the duration of normothermic periods and Some recent studies on endemic Australian long- torpor bouts in two species of long-eared bats, N. eared bats of the genus Nyctophilus (Family Vespertili- geo€royi and N. gouldi, to determine whether they onidae) suggest that they may di€er considerably from di€er from northern-hemisphere bats. We measured northern-hemisphere species with respect to the physi- the MR of bats overnight to allow them to enter ology of torpor. Body temperature (Tb), metabolic rate torpor during the night or near sunrise as in the ®eld, (MR), and the di€erential between Tb and Ta reported in and assessed whether these values di€ered from those several species of these tree-cavity roosting bats (Ny- obtained from bats measured entirely during daytime. ctophilus geo€royi, N. gouldi and N. major) during tor- Further, we investigated interrelations between physi- por are relatively high (Morris et al. 1994; Hosken 1997; ological variables associated with torpor and deter- Hosken and Withers 1999). Whereas the Tb of many mined whether MR during torpor (TMR) di€ers torpid vespertilionids from the northern hemisphere fall among seasons. to near 0 °C, and the di€erential between Tb and Ta is often around 1 °C (Hock 1951; Henshaw 1970; Lyman 1970; Geiser and Ruf 1995), torpid Australian long- Materials and methods eared bats appear to maintain a Tb of >10 °C, and a Tb±Ta di€erential that is more than twofold, and a MR Bats were captured in mist nests in Imbota Nature Reserve (for- that is up to tenfold that of northern-hemisphere ves- merly Eastwood State Forest) near Armidale, New South Wales pertilionids. Some of these di€erences may be due to (30°35¢S, 151°44¢E), an open woodland area at approximately 1000 m altitude. Captured animals were transferred to the labo- climatic factors; however, e€ects of captivity or experi- ratory at the University of New England and measured over a mental conditions may also play an important role period of 1±3 days before being released at the site of capture. To (Hosken 1997; Hosken and Withers 1999). With respect ensure that bats held for more than 1 day maintained condition to experimental conditions it may be important that all while in captivity, they were hand fed with mealworms and water once per day while not being measured. Measurements were recently published experiments used to quantify physi- conducted on a total of 24 N. geo€royi (September 1996 to May ological variables of torpid Australian long-eared bats 1997) and 9 N. gouldi (January to July 1997). _ (Nyctophilus spp.) were conducted with individuals in The MR, measured as rate of oxygen consumption (VO2), was which torpor was induced during the day and measured monitored over periods ranging from several hours to several 2±3 h after entry into torpor around the time when free- days over a range of Tas and natural photoperiod. TMR was averaged from minimum measurements that remained constant living Nyctophilus bats typically arouse (Turbill et al. over at least 30 min. Two types of measurements were performed: 1999). In contrast, many measurements on northern- (1) overnight and the following day(s), and (2) entirely during hemisphere bats were conducted overnight to allow daytime. them to undergo their natural thermal cycle with torpor Most bats were measured beginning in the afternoon, overnight and throughout most of the following day to allow them to un- entry at night or around dawn, or bats were transferred dergo their natural daily thermal cycle with torpor entry during the to the measuring equipment while hibernating night or near sunrise (Turbill et al. 1999; Figs. 1, 2). Readings were (Hock 1951; Pohl 1961; Henshaw 1968; Riedesel and taken in the morning after torpor bouts of at least 4 h (average Williams 1976; Thomas et al. 1990; Speakman et al. 10.5 h) at the time when bats are naturally torpid (Turbill et al. 1999). Initially, we followed the protocol of Hosken (1997) and 1991). Riedesel and Williams (1976) showed that the other recent studies of Nyctophilus spp., beginning measurements in time to reach minimum MR during torpor in Myotis mid-morning, and taking readings after bats were in torpor for 2± velifer (12 g) was up to 19 h from the beginning of 3 h. However, we noted that most bats did not appear to reach measurements. steady-state minima, but instead often maintained a MR The scarcity of data on the thermal biology and throughout the day that was about ten times that reported in northern-hemisphere bats. Therefore, we discontinued this proto- energetics of Australian bats also extends to seasonal col and concentrated instead on overnight measurements as out- e€ects. Present data suggest that Australian microbats lined above. 155

Measurements of basal metabolic rate (BMR) were carried out separately. Postabsorptive bats were exposed to Ta 26 °C in the Results morning and Ta was slowly increased to a maximum of 35 °Cin increments of about 2 °C, which lasted for at least 1 h each. The Both species of Nyctophilus bats showed a high pro- minimum MR of normothermic individuals over at least 30 min at one of these temperature steps was used to calculate the av- clivity to enter torpor in captivity. The most common erage BMR of each bat. The thermoneutral zone (TNZ) was es- daily pattern of torpor and activity observed was: entry timated from the range of Ta in which the BMR of individuals into torpor shortly after bats were placed into the res- was measured. The MR of normothermic resting individuals pirometer during the day, torpor throughout the day, (RMR) below the TNZ was also averaged over at least 30 min. Usually RMR at only one T could be measured because bats arousal near lights o€, a short normothermic period, a and re-entry into torpor followed by a torpor bout entered torpor before a second measurement at another Ta could be completed. Time of arousal from torpor and entry into torpor lasting for the rest of the night and the following day and normothermic periods were estimated from measurements of (Fig. 1a). Arousals, characterised by an enormous in- MR. Bats were assumed to have fully aroused when the MR crease in MR near the beginning of the dark phase, overshoot during arousal episodes reached approximately twice began 18 ‹ 31 min after lights o€ (N. geo€royi) and RMR at the same Ta (Geiser 1986). Food and water were not available during measurements of MR which was performed in 37 ‹ 55 min after lights o€ (N. gouldi). The total 500-ml respirometry chambers ®tted with a wide plastic mesh to normothermic periods lasted for 111 ‹ 48 min (N. ge- allow for roosting and placed in a temperature-controlled o€royi) and 115 ‹ 66 min (N. gouldi). Normothermic (‹0.5 °C) cabinet. The ¯ow rate (100±200 ml min)1) was con- trolled with rotameters and measured with mass ¯owmeters periods usually consisted of a period of intense activity (Omega FMA-5606). and only brief resting phases. When bats were measured The percentage oxygen was measured with Ametek Applied overnight, torpor entry almost always occurred during Eletrochemistry S-3 A oxygen analysers. Most measurements were the dark phase. Only on one occasion did a bat conducted with a dual-channel S-3AII. Occasionally, a single- channel S-3AI, ®tted with a high-resolution output board (N. gouldi) measured overnight enter torpor in the light (80335SE), was used. In the dual-channel set-up, the MR of two phase 40 min after lights on. After re-entry into torpor, bats was measured simultaneously every 3 min; solenoid valves bouts lasting from early in the dark phase, throughout switched to a reference channel (outside air) for 3 min once every the day to the beginning of the next dark phase, were 15 min. The outputs from the oxygen analyser, the ¯owmeters, and frequently observed in both species. thermocouples measuring Ta inside the respirometry chamber to the nearest 0.1 °C, were transferred via a logger (Datataker DT100) A second arousal after midnight or late during the to a personal computer. In the single-channel set-up, four channels, dark phase followed by a short normothermic period which included three animal channels and one reference channel was observed less frequently (Fig. 1b). Torpor bouts (outside air) were scanned in sequence with solenoid valves. Each between two normothermic periods at night lasted for of the four channels was read for 3 min (i.e. the MR of each bat and the reference channel were measured once every 12 min). Both 368 ‹ 149 min (N. geo€royi) and 220 ‹ 194 min the output from the oxygen analysers and the ¯owmeter were (N. gouldi). transferred with an A/D 14-bit converter card to a personal com- Occasionally, bats entered torpor for more than puter. The Ta inside the respirometry chamber was read to the 1 day (Fig. 2). The longest torpor bouts, observed at nearest 0.1 °C with an Omega digital thermocouple thermometer and recorded on the personal computer via the A/D converter card. Ta<10 °C, were 41:30 h (N. geo€royi) and 33:36 h (N. The MR was calculated using standardised gas volumes and Eq. 3a gouldi), but all of these were interrupted by measure- of Withers (1977). Animals were weighed before and after the ex- ments of Tb and it is likely that bats would have re- periments and a linear decrease of body mass throughout the ex- mained torpid for longer if left undisturbed. Partial periment was assumed for calculation of mass-speci®c MR. Mass arousals, during which maximum MR was less than loss ranged between 0.1 g day)1 and 0.4 g day)1. Thermal conductance (C) was calculated using the equation: 20% of that during full arousals, were observed in some C ˆ MR(Tb±Ta). The Q10 was calculated using the equation: cases (Fig. 2). 10= T b1 ±T 2 † Both Nyctophilus species rarely maintained a high T Q10 ˆ MR1=MR2† b . Tb was measured to the nearest b 0.1 °C with an Omega digital thermometer by inserting a ®ne cal- at rest during daytime cold exposure. At high Ta, ibrated thermocouple probe 1 cm into the rectum. Since the rate of normothermic thermoregulation was more common. rewarming is slow at the beginning of the rewarming process BMR of N. geo€royi was 1.36 ‹ 0.17 ml g)1 h)1 (Geiser and Baudinette 1990), rectal Tb, which was measured within 15 s of removal from the respirometers, was within 0.2 °Cof (Tb ˆ 35.7 ‹ 0.7 °C; body mass ˆ 7.1 ‹ 0.8 g; n ˆ 6) the Tb before the disturbance. Animals were considered torpid and was measured at Ta 29.1±33.2 °C (Fig. 3a, b). The )1 )1 when Tb fell below 30 °C. When no Tb measurements were avail- BMR of N. gouldi was 1.22 ‹ 0.13 ml g h able, animals were considered to be torpid when their MR fell to (T ˆ 36.0 ‹ 1.4 °C; body mass ˆ 10.0 ‹ 1.1 g; n ˆ 8) that or below that of individuals in which both variables were b and was measured at T 30.7±32.5 °C (Fig. 4a, b). measured and which had a Tb of less than 30 °C at same Ta. All a equipment was calibrated before measurements. Data-acquisition During cold exposure the RMR of both species in- software was written by G. KoÈ rtner, B. Lovegrove and T. Ruf. creased linearly with decreasing Ta (Figs. 3b, 4b), but Results are expressed as means ‹ standard deviation (SD) for the T of N. geo€royi (33.6 ‹ 1.4 °C) was somewhat the number of individuals (n) that were measured if not speci®ed b otherwise. Di€erences among or between means were tested using a lower than in the TNZ (Fig 3a). one-way ANOVA or a Student's t-test as appropriate (Zar 1984). When torpid, bats substantially reduced MR and Tb Regression equations were ®tted using the method of least squares (Figs. 1±4). Over a wide range of Ta, bats were the- (Zar 1984). Interrelations between physiological variables, and rmoconforming and MR and T fell with T . The TMR potential seasonal physiological changes were only tested in b a of both species showed an exponential decline with Ta N. geo€royi, because it was the only species for which our sample )1 )1 size allowed for meaningful analysis. above 1 °C(N. geo€royi: TMR (ml g h ) ˆ 0:026 156

Fig. 1a, b Daily ¯uctuations of metabolic rate in Nyctophilus geo€royi. One bat (a) measured at an air temperature (Ta)of 10.7 °C aroused after lights o€, indicated by the substantial increase in metabolic rate (MR), and remained in torpor from early in the dark phase well into the next day when body temperature (Tb)was measured. The other bat (b) measured at Ta 10.5 °C aroused twice, early in the dark phase and at about 0300 hours. The black bars indicate the dark phase

Fig. 3 Tbs(a)andMRs(b) in normothermic (open symbols)and torpid ( ®lled symbols) N. geo€royi as a function of Ta. Each point represents one individual measurement. MR in resting normothermic )1 )1 bats (RMR) increased with decreasing Ta:RMR(mlg h )= 2 11.69±0.35Ta (°C); r ˆ 0.91

17). The minimum TMR of N. gouldi was 0.052 ‹ )1 )1 0.01 ml g h (Tb ˆ 10.1 ‹ 2.1 °C; Ta ˆ 8.9 ‹ 1.4 °C; n ˆ 6). The steady-state TMR of thermoconforming N. ge- o€royi was related to Tb (Fig. 5). This relationship was best described by an exponential equation (r2 ˆ 0.94; )1 )1 P<0.0001). TMR was around 0.05 ml g h at Tb below 10 °C and increased about sixfold to values )1 )1 Fig. 2 MR over 2 days in N. geo€royi. The bat was measured at Ta around 0.3 ml g h at Tb between 25 °Cand30°C. 9.3 °C and remained in torpor from late afternoon on the 1st day, The Q10 was 3.4 between BMR and the minimum TMR throughout the night and the following day. A partial arousal and 3.0 for TMR of thermoconforming bats. occurred on the 2nd day shortly after lights o€. The black bars indicate the dark phase Whereas the TMR and Tb of torpid Nyctophilus bats declined with Ta over a wide range of Ta, the Tb±Ta di€erential in thermoconforming torpid bats was not a Ta(°C) 2 )1 )1 1:102 , r ˆ 0.92; N. gouldi: TMR (ml g h )= function of Ta (Figs. 3a, 4a). In N. geo€royi, the Tb±Ta 0.022 ´ 1.124Ta(°C), r2 ˆ 0.59 (Figs. 3b, 4b). The mini- di€erential of all torpid thermoconforming bats was mum TMR of N. geo€royi was 0.037 ‹ 0.014 ml described by the equation: Tb±Ta (°C) ˆ 1.08+0.02Ta )1 )1 2 g h (Tb ˆ 6.3 ‹ 3.2 °C; Ta ˆ 5.9 ‹ 2.7 °C; n= (°C) (r ˆ 0.035; P>0.3) and the mean Tb±Ta was 157

Fig. 5 MRs as a function of Tb in torpid thermoconforming N. geo€royi. Each point represents one individual measurement. MR in torpid bats (TMR) decreased exponentially with Tb:TMR )1 )1  2 (ml g h ) ˆ 0:0249  1:09Tb C†; r ˆ 0.94

measured for the species was 1.4 °C (Fig. 3a, b). One N. gouldi at Ta 1.0 °C regulated Tb at 3.7 °C and had a )1 )1 TMR of 0.42 ml g h ; the lowest Tb measured in this species was 2.3 °CatTa 0.7 °C (Fig. 4a, b). TMR di€ered between bats that were measured overnight versus bats in which torpor was induced in the morning and TMR was measured 2±3 h later (Fig. 6). At Ta 4±10 °C, the TMR of N. geo€royi was 0.04 ‹ 0.02 (n ˆ 18) when measured overnight, which was only about 10% of the value obtained from bats measured entirely during the daytime (TMR ˆ 0.41 ‹ Fig. 4 Tbs(a)andMRs(b) in normothermic (open symbols)and torpid (®lled symbols) N. gouldi as a function of Ta. Each point represents one individual measurement. MR in resting normothermic )1 )1 bats (RMR) increased with decreasing Ta:RMR(mlg h ) ˆ 9.91± 2 0.26Ta (°C); r ˆ 0.83

1.36 ‹ 0.85 °C. In N. gouldi, the Tb±Ta di€erential of all torpid thermoconforming bats was described by the 2 equation: Tb±Ta (°C) ˆ 1.02+0.006Ta (°C) (r ˆ 0.003; P>0.9) and the mean Tb±Ta was 1.08 ‹ 0.46 °C. It is likely that the constant Tb±Ta di€erential despite the rise of TMR at high Ta was possible because C increased from values around 0.05 ml g)1 h)1 °C)1 in thermo- conforming torpid bats at Ta below 10 °C to values of )1 )1 )1 about 0.1±0.3 ml g h °C at Ta above 15 °C (Song et al. 1997). At very low Ta, torpid bats increased their TMR to maintain Tb above a threshold (Figs. 3, 4). Most indi- vidual N. geo€royi increased TMR at Ta 0.4 ‹ 0.3 °C )1 )1 by about ®vefold to 0.209 ‹ 0.57 ml g h and Tb Fig. 6 MRs in torpid N. geo€royi measured overnight (hatched was 2.7 ‹ 0.9 °C(n ˆ 5). These bats had a raised C columns) and entirely during the daytime (open columns)atTasof4± 10 °C and 10±15 °C. Means ‹ SE are presented here because of the similar to that in torpid thermoconforming individuals large variation; means ‹ SD are provided in the text. TMR at both at T >15 °C. One N. geo€royi showed no evidence a Tas was signi®cantly greater during the daytime measurements than of thermoregulation at Ta 0.4 °C, and the lowest Tb during the overnight measurements in both Ta ranges 158 Physiological variables of normothermic bats mea- sured here were very similar to those reported in pre- vious studies. Regressions of RMR in N. geo€royi as a function of Ta, and the Tb below thermoneutrality were close to reported results, as were BMR and Tb in the TNZ (Hosken and Withers 1999). However, for both N. geo€royi and N. gouldi, BMR was only about 65% of that predicted from allometric equations for bats of equivalent size (Hayssen and Lacy 1985). This suggests that maintenance metabolism in Australian bats is rel- atively low, as in many other Australian monotreme, marsupial, and placental mammals (Hulbert and Daw- son 1974; Dawson 1989; Geiser and Coburn 1999; Hume 1999). This frugal use of energy for maintenance me- tabolism during normothermia in Australian mammals is likely a re¯ection of unpredictable climate and food supply and infertile soils of the continent (Hulbert and Dawson 1974; Lovegrove 1996). The minimum Tb of bats reported here are similar to Fig. 7 Metabolic rates (means ‹ SD) in torpid thermoconforming those of hibernating endotherms of equivalent body N. geo€royi measured at similar Tasinspring(Ta 10.3 ‹ 1.6 °C), mass (Geiser and Ruf 1995). Similarly, the Tb±Ta dif- summer (Ta 9.8 ‹ 0.8 °C), and autumn (Ta 10.2 ‹ 0.6 °C). TMR did not di€er among seasons (ANOVA; P > 0.08) although TMR in ferential of about 1 °C in torpid thermoconforming autumn appeared somewhat lower than those in the other seasons. Nyctophilus is close to that reported for hibernating bats Body mass was indistinguishable among seasons from the northern hemisphere (Hock 1951; Henshaw 1968, 1970). Moreover, the minimum TMRs reported here for Australian long-eared bats fall near the mean 0.46; n ˆ 4; t-test; P < 0.001). However, the variance in for hibernating endotherms (Geiser and Ruf 1995), and the daytime only sample was very large and the mini- are 88% (N. geo€royi) and 126% (N. gouldi) of the mum TMR in one bat was close to that measured in TMR predicted from the allometric equation for hiber- overnight experiments. At Ta 10±15 °C, TMR measured nators of the same body mass at a Tb of 5 °C (Geiser entirely during the day was about double the value of 1988). They are also similar to those of northern-hemi- those measured overnight (Fig. 6; t-test; P < 0.05). sphere microbats and the Australian broad-nosed bat The TMR of N. geo€royi did not appear to di€er (Scotorepens balstoni, 7 g) at similar Tb (Table 1). strongly with season (Fig. 7). When measured in a However, physiological variables of torpid Nyctophilus similar Ta range (ANOVA; P > 0.7), TMR in spring bats reported in our study are well below values previ- )1 )1, (0.065 ‹ 0.019 ml g h Ta 10.3 ‹ 1.6 °C; body ously reported for the same species (Morris et al. 1994; mass 7.4 ‹ 0.3 g; n ˆ 4) and summer (0.068 ‹ Hosken and Withers 1999), which has profound impli- )1 )1 0.01 ml g h ; Ta 9.8 ‹ 0.8 °C; body mass cations for energy expenditure and whether they are able 7.1 ‹ 0.7 g; n ˆ 7) appeared somewhat greater than to undergo multi-day torpor bouts. Some of the di€er- TMR measured in autumn (0.047 ‹ 0.009 ml g)1 h)1; ences between the studies may be explained by regional Ta 10.2 ‹ 0.6 °C; body mass 7.6 ‹ 0.4 g; n ˆ 4); di€erences, as long-eared bats studied previously came however, the means did not di€er signi®cantly (ANO- from areas that are warmer than the Armidale region. VA; P ˆ 0.083). Body mass and total TMR (ml h)1) However, the minimum TMRs reported (Hosken 1997; also did not di€er among seasons (ANOVA; P > 0.1). Hosken and Withers 1999) were similar to values re- ported here, and it appears therefore that the minimum TMR rather than the mean values reported in previous Discussion studies represent steady-state torpor in Nyctophilus.This interpretation is supported by our data on bats mea- Our study provides the ®rst evidence that Australian sured entirely during the daytime, following the protocol long-eared bats exhibit similar thermal characteristics of previous studies, when TMRs were about ten times and patterns of torpor to their relatives from the the mean value of bats measured overnight (Fig. 6). The northern hemisphere. The Tb of both Nyctophilus species large di€erences between bats measured entirely during fell to about 1±3 °C and torpor bouts could last for the day and those measured overnight are likely to be more than 1 day. TMR of the bats we measured was as related to the natural thermal cycle of Nyctophilus bats low as 0.5% of that in normothermic bats at the same Ta in the wild. Free-ranging Nyctophilus geo€royi invari- and about 3% of the BMR. As torpor was used fre- ably enter torpor in the night or early morning and often quently and reduced energy expenditure substantially it arouse during the middle of the day when Ta rises appears that it plays a central role in the biology of (Turbill et al. 1999). If torpor is induced around mid- Australian microbats. morning it coincides with the time when free-ranging 159

Table 1 Physiological variables of torpid northern-hemisphere and Australian microbats. Data for species measured overnight or hibernating (Tb body temperature, TMR torpid metabolic rate) Species Body Min. Min. TMR Torpor bout Source mass (g) Tb (°C) (ml O2 length (days) g)1 h)1)

Northern hemisphere Myotis myotis 25 4 0.04 41 Pohl 1961; Kulzer 1965; Harmata 1987 Myotis lucifugus 5.2 1.3 0.02 31 Hock 1951; French 1985; Thomas et al. 1990 Myotis velifer 12 0.07 Riedesel and Williams 1976 Myotis natteri 8 9 0.031 Kulzer 1965; Speakman et al. 1991 Myotis daubentoni 9 0.07 Speakman et al. 1991 Barbastella barbastellus 7 0.04 Pohl 1961 Pipistrellus pipistrellus 7.4 3 0.07 Kulzer 1965; Speakman et al. 1991 Eptesicus fuscus 22 5 0.03 25 Kulzer 1965; French 1985; Szewczak and Jackson 1992 Rhinolophus hipposideros 6 18 Harmata 1987 Australian Nyctophilus geo€royi 7 2.7 0.037 >1 Ellis et al. 1991; present study Nyctophilus gouldi 10 3 0.052 >1 Present study Scotorepens balstoni 7 3.2 0.044 Geiser and Brigham, unpublished data Miniopterus schreibersii 15 12 Hall 1982 bats often show passive rewarming caused by the increased in Ta and arousal. Torpid bats at that time appear to maintain Tb and MR above minimum steady- state levels perhaps because their normal daily thermal cycle has been interrupted. It is also likely that distur- bance by handling during the day results in a substantial increase in MR, as has been shown for several bat spe- cies (Speakman et al. 1991; Thomas 1995). Thus, the temporal organisation of an animal's daily cycle must be considered if reliable data on steady-state torpor, suit- able for comparison with other species and for investi- gation of mechanisms of MR reduction, are to be obtained. Our results support the argument that the reduction of MR during torpor in species capable of undergoing prolonged torpor bouts is due to three major factors (Geiser 1988; Guppy and Withers 1999). First, early in entry into torpor, normothermic thermoregulation ceases (Heller et al. 1977), and the MR can fall to ap- proximately BMR (i.e. by about 85±90% at low Ta in Fig. 8 MRs as a function of Tb torpid (TMR) and normothermic our bats) from RMR without lowering Tb (Geiser 1988; (BMR; open circles) N. geo€royi. Each point represents one individual )1 )1 Withers 1992; Song et al. 1996). measurement. The linear regression log10 TMR (ml g h ) ˆ )1.77+0.047T (°C); r2 ˆ 0.91, ®tted to the logarithm of MR in Second, as the Q10 of 8.6 (i.e. a 75% reduction of b thermoconforming torpid bats ( ®lled circles) intersected Tb of MR) between BMR and TMR at Tb of 29.3 °C, is far normothermic bats with BMR at 59.5% BMR. TMR of thermoreg- above that typical for biochemical reactions of 2±3 ulating torpid bats are shown by ®lled triangles (Schmidt-Nielsen 1997) and for many, especially larger, hibernating species (Geiser 1988; Nicol et al. 1992), it appears that physiological inhibition is involved in in comparison to active animals (Storey and Storey lowering the MR (Geiser 1988; Storey and Storey 1990; 1990; Martin et al. 1999), whereas the TMR is as low as Malan 1993; Song et al. 1997; Guppy and Withers 1999; 3% of the BMR (i.e. a di€erence of more than one order Hosken and Withers 1999; Martin et al. 1999). This in- of magnitude!). terpretation is further supported by the observation that Thus, the third in¯uence clearly must be the e€ect of the regression of the logarithm of TMR versus Tb in- lowering Tb by about 35 °C. As not enough heat is tercepts with Tb at BMR 40.5% below the BMR (Fig. 8) produced for normothermic thermoregulation, because suggesting that physiological inhibition contributes to the threshold for Tb is lowered at torpor entry and lowering TMR in small species capable of hibernation. thermogenesis is inactivated, Tb begins to decline and However, in vitro evidence suggests that enzyme activity this fall of Tb in turn should a€ect TMR. That TMR during hibernation is usually reduced by less than 50% and Tb are related is emphasised by the 94% prediction 160 of TMR by Tb in thermoconforming torpid bats same habitats (Geiser 1987; Withers et al. 1990; Geiser (Fig. 5), and by the exponential relationship that is and Ruf 1995). characteristic of e€ects of temperature on rates (Withers Seasonal changes in TMR of N. geo€royi were not 1992; Schmidt-Nielsen 1997; Guppy and Withers 1999) pronounced. Spring and summer TMR were almost (Fig 5). Overall, the reduction of MR from RMR to the identical, but there was some indication of a slight de- minimum TMR in N. geo€royi is 99.6%. Most of this crease in autumn. It is therefore possible that TMR in reduction (85.9%) can be explained by the fall from winter are even lower and that TMR shows some sea- RMR to BMR, 5.7% is explained by metabolic inhibi- sonal acclimatisation. Unfortunately, we were unable to tion (from BMR to the predicted TMR at Tb 35.7 °C; ®nd N. geo€royi in the winter of our study, most likely Fig. 8), and 8.0% is explained by the fall in Tb assuming because they were hibernating. However, as our summer that the Q10 of about 3 for TMR (Fig. 8) is explained data were similar to those obtained in northern hemi- entirely by temperature e€ects. If only values below the sphere bats during hibernation in winter, it is likely that BMR are considered as in many studies on this subject TMR show only small seasonal changes and that the (summarised in Geiser 1988), the 97.3% reduction in known seasonal changes in torpor patterns and bout MR can be explained by metabolic inhibition (40.5%) duration in bats are to a large extent determined by and temperature e€ects (56.8%) applying the assump- environmental conditions. Clearly, more detailed ®eld tions outlined in the previous sentence. data on torpor patterns and environmental variables are Although Nyctophilus in steady-state torpor show no needed to put our results on thermal biology and ener- evidence for physiological thermoregulation over a wide getics in perspective. range of Ta, regulation of Tb commenced in most indi- Most Nyctophilus investigated here displayed short viduals when Ta fell below about 1 °C. Evidence for bouts of torpor and aroused daily after sunset. The thermoregulation comes from the increase in the Tb±Ta abrupt signal of lights o€ in the laboratory at the be- di€erential and a substantial increase in TMR (Figs. 3, ginning of the dark phase appears to be a strong trigger 4, 8), as in other heterothermic endotherms (Heldmaier for arousals even at low Ta. However, it is clear from the and Ruf 1992). Thermoregulatory thresholds during present study and other laboratory and ®eld investiga- torpor appear to be common to most (all?) heterother- tions that Nyctophilus spp. are capable of prolonged mic endotherms (Heller and Hammel 1972; Heller et al. torpor and most likely hibernate for part of the winter 1977). It appears that extra energy is spent to prevent Tb (Ellis et al. 1991; Brigham and Geiser 1998; Turbill et al. of torpid animals from reaching temperatures that 1999). Prolonged bouts of torpor also conform with the would cause physiological dysfunction. In hibernating TMR of the bats measured here. TMR in Nyctophilus species capable of prolonged torpor, most minimum Tb was similar to that of hibernating species and only about lie within the range 0±10 °C (Geiser and Ruf 1995). A 10% of that in species which enter daily torpor exclu- reduction much below 0 °C is obviously not possible sively and are unable to undergo prolonged torpor because of the implications of the freezing of body ¯uids (Geiser and Ruf 1995). It thus appears that although (Barnes 1989). Higher minimum Tb appear to be a re- Nyctophilus bats may frequently arouse on a daily basis, ¯ection of the thermal environment experienced during their short bouts of torpor do not resemble daily torpor, hibernation (Kulzer 1965). Increased thermoregulatory but are, physiologically speaking, short bouts of hiber- costs and in turn more frequent arousals from torpor nation. It would be interesting to investigate whether and the associated increase in energy expenditure at Ta this is also the case for other bats from temperate cli- below the Tb threshold (Geiser and Kenagy 1988; mates and whether short torpor bouts in microbats from Ransome 1990) should represent a strong selective the tropics constitute daily torpor or short hibernation. pressure to maintain minimum Tb below or near the Ta experienced for most of the hibernation season, espe- Acknowledgements We thank Becky Francis, Steve Debus, Robin cially in species relying to a large extent on stored energy Gutsell, Sandy Hamdorf, Gerhard KoÈ rtner, and Richard Wiacek for their help with netting the bats. This work was supported by a during winter. While this argument may appear plausi- grant from the Australian Research Council to FG. ble from an energetic point of view, why should Aus- tralian bats living on a warm continent let Tb fall to such low temperatures just a few degrees above the freezing References point of water? Although most of Australia is warm in summer, more than 80% of the continent experiences Audet D, Fenton MB (1988) Heterothermy and the use of torpor regular frost in winter, and the average minimum Ta in by the bat Eptesicus fuscus (Chiroptera: Vespertilionidae): winter for most of the southern half of Australia is be- a ®eld study. Physiol Zool 61: 197±204 Barclay RMR, Kalcounis MC, Crampton LH, Stefan C, Vonhof low 6 °C (Colls and Whitaker 1993). As tree roosts MJ, Wilkinson L, Brigham RM (1996) Can external radio- provide limited thermal bu€ering, and tree-dwelling bats transmitters be used to assess body temperatures and torpor in usually are torpid when Ta is low, it is likely that Aus- bats? J Mammal 77: 1102±1106 Barnes BM (1989) Freeze avoidance in a mammal: body temper- tralian bats in many regions regularly experience Tb between 1 °C and 5 °C, as do pygmy possums (Cercar- atures below 0 °C in an arctic hibernator. Science 244: 1593± 1595 tetus spp.), honey possums (Tarsipes rostratus) and Barnes BM, Ritter D (1993) Patterns of body temperature change feathertail gliders (Acrobates pygmaeus) living in the in hibernating arctic ground squirrels. In: Carey C, Florant GL, 161

Wunder BA, Horwitz B (eds). Life in the cold. Westview, Henshaw RE (1968) Thermoregulation during hibernation: appli- Boulder, pp 119±130 cation of Newton's law of cooling. J Theor Biol 20: 79±90 Bartels W, Law BS, Geiser F (1998) Daily torpor and energetics in Henshaw RE (1970) Thermoregulation in bats. In Slaugher BH, a tropical mammal, the northern blossom-bat Macroglossus Walton DW (eds). About bats. Southern Methodist University minimus (Megachiroptera). J Comp Physiol B 168: 233±239 Press, Dallas, pp 188±232 Brigham RM, Geiser F (1998) Seasonal activity patterns of Ny- Hock RJ (1951) The metabolic rates and body temperatures of ctophilus bats based on mist-net captures. Aust Mammal 20: bats. Biol Bull 101: 289±299 349±352 Hosken DJ (1997) Thermal biology and metabolism of the greater Coburn DK, Geiser F (1998) Seasonal changes in energetics and long-eared bat, Nyctophilus major (Chiroptera: Vespertilion- torpor patterns in the subtropical blossom-bat Syconycteris idae). Aust J Zool 45: 145±156 australis (Megachiroptera). Oecologia 113: 467±473 Hosken DJ, Withers PC (1999) Metabolic physiology of euthermic Colls K, Whitaker R (1993) The Australian weather book. Na- and torpid lesser long-eared bats, Nyctophilus geo€royi (Chir- tional Book Distributers, Brookvale optera: Vespertilionidae). J Mammal 80: 42±52 Dawson TJ (1989) Responses to cold of monotremes and marsu- Hulbert AJ, Dawson TJ (1974) Standard metabolism and body pials. In: Wang LCH (ed) Advances in comparative and envi- temperature of perameloid marsupials from di€erent environ- ronmental physiology. Springer, Berlin Heidelberg New York, ments. Comp Biochem Physiol 47A: 583±590 pp 255±288 Hume ID (1999) Marsupial nutrition. Cambridge University Press, Ellis WA, Marples TG, Phillips WR (1991) The e€ect of a tem- Cambridge perature-determined food supply on the annual activity cycle of Kulzer E (1965) Temperturregulation bei FledermaÈ usen aus the lesser long-eared bat, Nyctophilus geo€royi Leach, 1921 verschiedenen Klimazonen. Z Vergl Physiol 50: 1±34 (Microchiroptera: Vespertilionidae). Aust J Zool 39: 263±271 Kulzer E, Nelson JE, McKean J, MoÈ hres FP (1970) Un- Fenton MB (1983) Just bats. University of Toronto Press, Toronto tersuchungen uÈ ber die Temperaturregulation australischer French AR (1985) Allometries of the duration of torpid and eu- FledermaÈ use (Microchiroptera). Z Vergl Physiol 69: 426±451 thermic intervals during mammalian hibernation: a test of the Kunz TH (ed) (1982) Ecology of bats. Plenum Press, New York theory of metabolic control of the timing of changes in body Lovegrove BG (1996) The low basal metabolic rates of marsupials: temperature. J Comp Physiol B 156: 13±19 the in¯uence of torpor and zoogeography. In: Geiser F, Hulbert Geiser F (1986) Thermoregulation and torpor in the kultarr, An- AJ, Nicol SC (eds) Adaptations to the cold: tenth international techinomys laniger (Marsupialia: Dasyuridae). J Comp Physiol hibernation symposium. University of New England Press, B 156: 751±757 Armidale, Australia, pp 141±151 Geiser F (1987) Hibernation and daily torpor in two pygmy pos- Lyman CP (1970) Thermoregulation and metabolism in bats. In: sums (Cercartetus spp., Marsupialia). Physiol Zool 60: 93±102 Wimsatt WA (ed) Biology of bats. Academic Press, New York, Geiser F (1988) Reduction of metabolism during hibernation and pp 301±330 daily torpor in mammals and birds: temperature e€ects or Malan A (1993) Temperature regulation, enzyme kinetics, and physiological inhibition? J Comp Physiol B 158: 25±37 metabolic depression in mammalian hibernation. In: Carey C, Geiser F, Baudinette RV (1990) The relationship between body Florant GL, Wunder BA, Horwitz B (eds) Life in the cold. mass and rate of rewarming from hibernation and daily torpor Westview, Boulder, pp 241±252 in mammals. J Exp Biol 151: 349±359 Martin SL, Maniero GD, Carey C, Hand SC (1999) Reversible Geiser F, Coburn DK (1999) Field metabolic rates and water up- depression of oxygen consumption in isolated liver mitochon- take in the blossom bat Syconycteris australis (Megachirop- dria during hibernation. Physiol Biochem Zool 72: 255±264 tera). J Comp Physiol B 169: 133±138 Morrison P (1959) Body temperatures in some Australian mam- Geiser F, Kenagy GJ (1988) Torpor duration in relation to tem- mals. I. Chiroptera. Biol Bull 116: 484±497 perature and metabolism in hibernating ground squirrels. Morris S, Curtin AL, Thompson MB (1994) Heterothermy, torpor, Physiol Zool 61: 442±449 respiratory gas exchange, water balance and the e€ect of feed- Geiser F, Ruf T (1995) Hibernation versus daily torpor in mam- ing in Gould's long-eared bat Nyctophilus gouldi. J Exp Biol mals and birds: physiological variables and classi®cation of 197: 309±335 torpor patterns. Physiol Zool 68: 935±966 Nicol S, Andersen NA, Mesch U (1992) Metabolic rate and ven- Geiser F, Coburn DK, KoÈ rtner G, Law BS (1996) Thermoregu- tilatory pattern in the echidna during hibernation and arousal. lation, energy metabolism, and torpor in blossom bats, In: Augee ML (ed) Platypus and echidnas. Royal Zoological Syconycteris australis (Megachiroptera). J Zool London 239: Society NSW, Sydney, pp 150±159 583±590 Pohl H (1961) Temperarurregulation und Tagesperiodik des Guppy M, Withers PC (1999) Metabolic depression in animals: Sto€wechsels bei WinterschlaÈ fern. Z Vergl Physiol 45: 109±153 physiological perspectives and biochemical generalisations. Biol Ransome RD (1990) The natural history of hibernating bats. Helm, Rev 74: 1±40 London Hall LS (1982) The e€ect of cave microclimate on winter roosting Riedesel ML, Williams BA (1976) Continuous 24-hour oxygen behaviour in the bat, Miniopterus schreibersii blepotis. Aust J consumption studies of Myotis velifer. Comp Biochem Physiol Ecol 7: 129±136 54A: 95±99 Hall LS (1984) And then there were bats. In: Archer M, Clayton G Schmidt-Nielsen K (1997) Animal physiology. Cambridge Uni- (eds) Vertebrate zoogeography and evolution in Australasia. versity Press, Cambridge Heperian Press, Carlisle, pp 837±852 Song X, KoÈ rtner G, Geiser F (1996) Interrelations between meta- Harmata W (1987) The frequency of winter sleep interruptions in bolic rate and body temperature during entry into daily torpor in two species of bats hibernating in limestone tunnels. Acta Sminthopsis macroura. In: Geiser F, Hulbert AJ, Nicol SC (eds). Theriol 32: 331±332 Adaptations to the cold: tenth international hibernation sym- Hayssen V, Lacy RC (1985) Basal metabolic rates in mammals: posium. University of New England Press, Armidale, pp 63±69 taxonomic di€erences in the allometry of BMR and body mass. Song X, KoÈ rtner G, Geiser F (1997) Thermal relations of metabolic Comp Biochem Physiol 81A: 741±754 rate reduction in a hibernating marsupial. Am J Physiol 273: Heldmaier G, Ruf T (1992) Body temperature and metabolic rate R2097±R2104 during natural hypothermia in endotherms. J Comp Physiol B Speakman JR, Webb PI, Racey PA (1991) E€ects of disturbance on 162: 696±706 the energy expenditure of hibernating bats. J Appl Ecol 28: Heller HC, Hammel HT (1972) CNS control of body temperature 1087±1104 during hibernation. Comp Biochem Physiol 41: 349±359 Storey KB, Storey JM (1990) Metabolic rate depression and bio- Heller HC, Colliver GW, Beard J (1977) Thermoregulation during chemical adaptations in anaerobiosis, hibernation and estiva- entrance into hibernation. P¯uÈ gers Arch 369: 55±59 tion. Q Rev Biol 65: 145±174 162

Szewczak JM, Jackson DC (1992) Ventilatory response to hypoxia Twente JW, Twente JA (1965) Regulation of hibernating periods and hypercapnia in the torpid bat, Eptesicus fuscus. Respir by temperature. Proc Natl Acad Sci USA 54: 1058±1061 _ _ Physiol 88: 217±232 Withers PC (1977) Measurements of VO2, VCO2, and evapora- Thomas DW (1995) The physiological ecology of hibernation in tive water loss with a ¯ow through mask. J Appl Physiol 42: vespertilionid bats. Symp Zool Soc London 67: 233±244 120±123 Thomas DW, Cloutier D, Gagne D (1990) Arrhythmic breathing, Withers PC (1992) Comparative animal physiology. Saunders, Fort apnea and non-steady-state oxygen uptake in hibernating little Worth brown bats (Myotis lucifugus). J Exp Biol 149: 395±406 Withers PC, Richardson KC, Wooller RD (1990) Metabolic Turbill C, KoÈ rtner G, Geiser F (1999) Thermal biology and roost physiology of euthermic and torpid honey possums. Aust J selection of long-eared bats, Nyctophilus geo€royi. Proceedings Zool 37: 685±693 of the 1999 Scienti®c Meeting of the Australian Mammal So- Zar JH (1984) Biostatistical analysis. Prentice Hall, Englewood ciety, Abstracts, p 60 Cli€s