Torpor in Three Species of Brazilian under Semi-Natural Conditions Author(s): Claus Bech, Augusto S. Abe, John Fleng Steffensen, Martin Berger and Jose Eduardo P. W. Bicudo Source: The Condor, Vol. 99, No. 3 (Aug., 1997), pp. 780-788 Published by: University of California Press on behalf of the Cooper Ornithological Society Stable URL: http://www.jstor.org/stable/1370489 . Accessed: 15/01/2014 07:54

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

University of California Press and Cooper Ornithological Society are collaborating with JSTOR to digitize, preserve and extend access to The Condor.

http://www.jstor.org

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions The Condor 99:780-788 C The Cooper Ornithological Society 1997

TORPOR IN THREE SPECIES OF BRAZILIAN HUMMINGBIRDS UNDER SEMI-NATURAL CONDITIONS1

CLAUSBECH Department of Zoology, Norwegian University of Science and Technology, N-7034 Trondheim, Norway, e-mail: [email protected] AUGUSTo S. ABE Departmento de Zoologia, Universidade Estadual Paulista, BR-13506900 Rio Claro, Sp, Brazil JOHN FLENG STEFFENSEN2 Department of Zoophysiology, University of Aarhus, DK-8000 Aarhus C., Denmark MARTIN BERGER Westfiilisches Museum fiir Naturkunde, D-48161 Miinster, Germany JOSEEDUARDO P. W. BICUDO Uhiversidade de Sao Paulo, Instituto de Biociencias, Dept. Fisiol., BR-05508900 Sao Paulo, Brazil

Abstract. We measuredbody temperaturesin three species of Brazilianhummingbirds, the VersicoloredEmerald (Amazilia versicolor; body mass 4.1 g), the Black (Me- lanotrochilusfuscus; body mass 7.7 g) and the Swallow-tailedHummingbird (Eupetomena macroura;body mass 8.6 g), duringovernight exposure to naturalconditions of photoperiod and ambienttemperatures. All three species entered torpor.In both A. versicolor and E. macroura,individuals entered torpor even if they had access to feeders up to the time of sunset. In contrast,M. fuscus was less proneto entertorpor and did so mainlyif it had been fasting for more than two hoursbefore sunset.Furthermore, M. fuscus often spentthe whole night in torpor,whereas the two other species entered torpor for a variable,often short, period of the night. We observed more than one torporbout during a single night in all three species. We suggest that multiple nocturnaltorpors result from interruptionof the normal torpor patternby some (unknown)external stimuli. Any interruptedtorpor was always followed by a new entry into torpor,supporting the view that there is a body mass thresholdbelow which the hummingbirdsmust enter torpor. Our data also indicatethat these hummingbirdspecies might use torporeven if they are not energeticallystressed. Key words: hummingbirds,Amazilia versicolor,Melanotrochilus fuscus, Eupetomena macroura, torpor, thermoregulation, body temperature.

INTRODUCTION Hiebert 1990). Most of the studies on humming- For many small homeotherms,daily torporis an suggest that daily torpor is a strategy em- essential mechanism to cope with periods of ployed in situations involving energetic stress. food deprivationand/or severe weather condi- Thus, torpor has less often been reported in ap- tions that pose a challenge to their energeticbal- parently well-fed individuals (Krtiger et al. ance. Torporis known to occur in birds from at 1982, Carpenter and Hixon 1988, Hiebert 1993a, least six differentorders (Procellariiformes,Co- 1993b). lumbiformes, Coliiformes, Caprimulgiformes, Because it is difficult to obtain physiological , and Trochiliformes;see Reinert- information from free-living hummingbirds, sen 1983, Heller 1989, French 1993). Daily tor- most studies have been of birds kept under lab- por has been extensively studied in humming- oratory conditions. Information about the use of birds, in which significantenergetic savings are torpor under natural conditions in hummingbirds correlated with the use of daily torpor (Hain- is consequently scarce (Calder and Booser 1973, sworth et al. 1977, 1981, Beuchat et al. 1979, Carpenter 1974, Calder et al. 1990), as it is for other species entering torpor (Brigham 1 1992). The only study that has provided accept- Received 27 June 1996. Accepted 13 March1997. able that need not be en- 2Current address: Marine Biological Laboratory, proof hummingbirds University of Copenhagen,DK-3000 Helsingor,Den- ergetically stressed in order to resort to daily tor- mark. por is that of Carpenter and Hixon (1988), who [780]

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions TORPORIN BRAZILIANHUMMINGBIRDS 781 demonstratedtorpor in a free-living, well-fed, (Mettler,accuracy 0.01g) and then placed indi- Rufous Selasphorus rufus. In the vidually in smaller overnight cages (approxi- same species, but under laboratoryconditions, mately 12 X 12 X 20-cm cardboardboxes) pro- Hiebert(1993a, 1993b) reportedincreased inten- vided with a perch. Usually the birds roosted sity of torporduring periods of food restriction, quietly on the perch during the night, although but when body mass nonetheless showed an in- in some cases they apparentlyhad spent the crease. In S. rufus the use of torporwas highest night sitting on the floor of the cage. The cages in the autumn, suggesting a function of torpor (up to six used each night) were placed outdoors as a strictly energy saving mechanism, thereby duringthe night. The walls and top of each cage minimizing the time required for premigratory were equipped with holes, to ensure that the fattening. In this case torpor is not correlated hummingbirdswere exposed to natural varia- with inadequate food intake (Carpenteret al. tions in both ambienttemperature and photope- 1993, Hiebert 1993a). However, it is still an riod. open question whether hummingbirdsnormally The small size of hummingbirdsmakes the need to be energy-stressedin order to enter tor- measurementof body temperature(Tb) a difficult por under naturalconditions (Calder 1994). task. We used a copper-constantanthermocouple We examined the use of torporby three spe- (California Fine Wire Company, type 0.005) cies of Brazilianhummingbirds kept undersemi- placed subcutaneouslyand laterally on the pec- naturalconditions. We asked whetherthese spe- toral muscle for measurementsof body temper- cies would entertorpor under conditions as close ature.The thermocouplewas fixed in place with to naturalas possible. We also wanted to study small pieces of adhesive tape. During measure- the influence of short-termchanges in their en- ment, the subcutaneouslyplaced tip of the ther- ergetic status on the use of torpor.The energetic mocouplewas covered by the wing. Pilot studies status of the hummingbirdswas manipulatedby of all three species indicatedthat such measure- experimentallydepriving them of food from the ments of pectoral temperaturedid not differ by time of captureuntil sunset. more than 0.2-0.30C from simultaneous mea- surementsof rectaltemperature. A thermocouple METHODS was placed inside one of the cages to recordthe The study was carriedout at the Museu de Biol- actual ambient temperature(Ta) to which the ogia, at SantaTeresa in the state of EspiritoSan- birds were exposed. All thermocoupleswere ex- to, Brazil (19055'S, 40036'W; about700 m above tended by using large-diametercopper-constan- sea level). We studied three species of hum- tan thermocouples(Bicc Cables, U.K., 4-5 m mingbirds, the Versicolored Emerald Amazilia length) that went to a nearby house where the versicolor (- 4.1 g), the Black Jacobin Melan- data-acquisitionequipment was placed. otrochilus fuscus (- 7.7 g), and the Swallow- Body temperatureswere measured every 40 tailed Hummingbird Eupetomena macroura (- sec throughout the night. The thermocouple 8.6 g). We did not determinesexes in any of the wires were connectedto a Data Translation(DT three species, all of which are common breeding 2805) A/D converter, via a DT-757 terminal birds in the study area (Ruschi 1982). board, and processed by a computer using a We conductedthe study from 2-20 December Labtech Notebook data acquisition program. 1987. The length of the nights did not vary Each night, up to six individuals were studied much, being approximately11 hr long. Sunrise simultaneously. We consistently used at least changed from 05:53 to 05:59 and sunset from two differentspecies, as well as variablefasting 19:10 to 19:19 duringthe studyperiod. All times times, each night. After arousal of the birds, are given in Rio de Janeirosummertime. which usually occurredbetween 06:00 and 07: The hummingbirdsin the study area are ac- 00, they were removed from their cages and re- customed to feeding at artificialfeeders. Hence, leased again after removing the thermocouples individuals of all three species were easily and re-weighing. caught at the feeders duringthe afternoon.After The total time spent in torporduring a night capture they were kept individually, at normal was calculated as the time Tb was below 350C ambient temperature,in ca. 0.5 m3 cages and during that particularnight. Mean body temper- deprived of food for a variable time (range 0- ature duringtorpor was only calculatedif a sta- 240 min) until sunset. The birds were weighed ble value had been attained.

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions 782 CLAUS BECH ET AL.

All statisticswere carriedout using SigmaStat software(ver. 1.0, JandelScientific). Results are N given as the mean ? SD. Statisticalsignificance 00 was set at P < 0.05. +1 +1 +1 RESULTS 0 ao 6 A total of 22 Amazilia versicolor, 29 Melano- % trochilus fuscus and 27 Eupetomena macroura were tested for one night. Of these, 16, 10 and oo 12 individuals, respectively, went into torpor. There were markedvariations in the pattern,as well as the of and the likelihood, enteringtorpor z durationof torpor among individuals and spe- cies. In M.fuscus, those individualsentering tor- I I I por had a significantlylower body mass at roost- ing time comparedto those individualsthat did +1 +1 +1 E O not enter torpor (Table 1). Thus whether this "0 species entered torporduring a particularnight 0 o 00 a0 clearly depended upon the evening body mass, and hence their energy reserves. In contrast, +1 +1 +1 z evening body mass was not associatedwith tor- o 0 por in A. versicolor and E. macroura (t-tests, P - 0 = 0.09 and P = 0.80, respectively; Table 1). In all three species the use of torporresulted in a ?1 ?1 ? significant reduction in overnight mass loss o to birds not in There t- O 0 compared torpor(Table 1). 00 0 v Al was a markeddifference between M. and •C 00 fuscus -, the other two species in the durationof torpor v, periods. Long torporperiods, lasting throughout O r- the entire night, were regularlyfound in M. fus- 666 r= cus, whereas shorter sometimes ?1 ?1 +1 torporperiods, o V less than 1 hr, were observed in the other two 00; IId II 6 s o 00 V V-00 species (Fig. 1). In both A. versicolor and E. v, macroura,total torportime duringa night varied 0 +1 +1 +1 z 0 between 2 and 11 hr, whereas in M. fuscus the 6edl were V torporperiods always long (7.5-11 hr). edl In all three of we ob- species hummingbirds v served an unusual patternof torpor with more than one torporbout during a single night (Fig. This of was seen in A. 1). pattern torpormostly 05 versicolor, in which up to three distinct torpor bouts could be recorded during a single night '~A (Fig. 1). Such a multiple torporpattern was ob- served in 4 of the 16 torpidA. versicolor, in 1 of the 10 torpidM. fuscus, and 1 of the 12 torpid E. macroura. The mean nocturnalbody temperaturesof the hummingbirdswhen in a non-torpidstate were 36.8 + 1.30C (n = 22) for A. versicolor, 37.3 + 1.5?C (n = 25) for M. fuscus, and 37.1 ? 1.3?C (n = 27) for E. macroura. None of these means were different (one way ANOVA, F2,71= 0.57, P > 0.5). During torpor the body temperature

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions TORPOR IN BRAZILIAN HUMMINGBIRDS 783

Amaziliaversicolor Melanotrochilusfuscus Eupetomenamacroura 40 40

30 30

20 -J20

30 30

S20 20 3 30 30

>g,4020 4020 -- -

0 30

20 20 20 22 24 2 4 6 20 22 24 2 4 6 20 22 24 2 4 6 Hour FIGURE 1. Representativeexamples of body temperaturemeasured in individualsof Amazilia versicolor, Melanotrochilusfuscus, and Eupetomenamacroura entering torpor. Note the occurrenceof multipletorpor bouts in several of the individuals.Line based on mean values calculatedfor every seven measurements(every 280 sec). was regulated at a level close to the ambient temperature differences between the three spe- = temperature, with mean values of 23.8 ? 1.4?C cies approached significance (F2,33 2.98, P = = ? 1.70C = = (n 13), 24.7 (n 12), and 23.2 ? 0.06, and F2,33 3.05, P = 0.06, respectively), 1.4?C (n = 11) for A. versicolor, M. fuscus and suggesting that M. fuscus kept a slightly higher E. macroura, respectively. These values corre- body temperature during torpor. The overall spond to body-to-ambient temperature differ- mean body-to-ambient temperature difference ences of 2.4 3.2 ? and 2.4 ?- + 0.80C, 0.80C, was 2.6 + 1.00C (n = 36, Fig. 2). The relative 1.2?C, respectively. The mean body tempera- constancyof this differenceis best illustratedby tures during torpor and the body-to-ambient those individualsexperiencing sudden tempera- ture changes during the recording. During one 28 1 particularnight therewas a heavy rainstormthat i 0 A. versicolor lowered the ambienttemperature by about 0 M. fuscus 2?C. 26 A E.macruora o10 This caused the body temperaturesof the torpid birds to change in parallel (Fig. 3). 24 As an index of overnightenergy expenditure, *o we calculated the mean overnight body-mass E 22 o loss (BML). The overnightBML (expressed as a percentage of the initial body mass lost per 0 20 hour duringthe night) decreasedin all threespe- cies with the use of torpor(Fig. 4 and Table 1). 18 In A. versicolorthere also was a decreasein noc- 18 19 20 21 22 23 24 turnal energy expenditurewith increasing fast- Ambient (oC) temperature ing time (FT) before sunset. This relationshipis FIGURE 2. Body temperatureof three species of described by the equation: BML = 0.779 - hummingbirdsduring torpor, as a functionof ambient 0.003FT (r = 0.70, n = 15, P < 0.05; Fig. 4). Solid line shows = the brokenline temperature. Tb Ta; Interestingly,even those individualsthat did not shows the linearregression line obtainedfor all the Tb- values: Tb = 1.05 + 1.08Ta (r = 0.77, P < 0.05, n = enter torpor but remained normothermic 36). throughoutthe night, showed a decrease in noc-

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions 784 CLAUS BECH ET AL.

1.2 36 Melanotrochilusfuscus Amaziliaversicolor oo 0.8 0o 32

0.4 28 E r- a o 0.0 24 ,o h 1.2 Melanotrochilusfuscus 0 0D o a 20 I 0.8 0 0 0 20 22 24 2 4 6 E 0 Hour

FIGURE3. Ambient and body temperaturesof two O0 Melanotrochilus a The fuscus throughout single night. 0.0 suddendrop in ambienttemperature at 01:40 was due to heavy rain. 0 Eupetomenamacroura 0.8 o o

turnal energy expenditure with increasing fast- 0.4 ing time. This relationship was statistically sig- o?9go nificant for M. fuscus (BML = 0.934 - 0.003FT, r = 0.78, n = 19, P < 0.001) and E. macroura 0.0 (BML = 0.675 - 0.001FT, r = 0.57, n = 15, P 0 60 120 180 240 Fastingtime (min) < 0.05; Fig. 4). A similar relationship could not be demonstrated for A. versicolor, probably be- FIGURE 4. Overnightbody mass loss (% decrease cause too few individuals failed to enter into tor- in body mass-hr-') in three species of hummingbirds as a function of time before sunset. cir- 4). Because of this be- fasting Open por (Fig. relationship cles indicate birds that were normothermic above tween BML and time in E. (Tb fasting non-torpid 350C) throughoutthe night, and filled circles signify macroura and M. fuscus, there may be a rela- hummingbirdsthat enteredtorpor for a periodof vari- tionship between non-torpid body temperatures able duration.Lines signify significant relationships and BML. This was in fact the case for E. ma- (for equations:see text). For two individuals(one A. versicolor and one we did not croura, in which there was a M. fuscus) record the significant positive accuratetime of food deprivationprior to sunset, and relationship between nocturnal non-torpid body the numberof individualsconsequently do not corre- temperature and overnight body-mass loss spond to the numberof individualsfrom which body (BML = 0.627 + 0.0327Tb, r = 0.46, one-tailed mass recordingswere obtained(Fig. 5). t-test, t14 = 1.86, P < 0.05). In M. fuscus there no such was, however, relationship. As BML is an indirect measure of energy ex- In the smallest species (A. versicolor), torpor penditure, the energy saved by entry into torpor occurred even in one individual that was al- can be calculated using the regression lines re- lowed to feed up to the time of sunset (Fig. 4). lating overnight mass loss to torpor duration It seems that A. versicolor is less tolerant of fast- (Fig. 5). Using these regression lines, we cal- stress than the other ing two larger hummingbird culated the decrease in overnight body-mass loss which species, could remain normothermic (assumed to correspond to the change in energy the even after for throughout night fasting al- expenditure) associated with a torpor period of most three hours (Fig. 4). However, for all three 10 hours out of a total night of 11 hours. The the of species, saving energy achieved by entry results indicated a mean nightly energy savings into torpor is reflected in lower overnight of 49% for A. versicolor, 61% for M. fuscus, and body-mass loss compared to those which did not 60% for E. macroura. enter torpor. This also is manifested in Figure 5, which shows that the overnight mass loss was DISCUSSION correlated, negatively and significantly, with the The present results indicate that, at the prevail- total time spent in torpor during the night. ing ambient temperatures, a nonregulated torpor

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions TORPORIN BRAZILIANHUMMINGBIRDS 785

1.2 , elicit a of Tb 2). The en- Amaziliaversicolor regulation (Fig. nightly ergy savings (49-61% of potential nightly en- 0.8 - ergy expenditure)resulting from using torporat these ambienttemperatures falls within the same as for other 0.4 0 range previouslyreported humming- birds (Krtigeret al. 1982, Wang 1989). Entry into torpor clearly depended upon the 0.0L hummingbirds'energetic state (the durationof 1.2 Melanotrochilusfuscus fasting time before sunset), and supports the 0 general view that torpor in hummingbirdsis a mechanism evoked as a to S0.8- normally response energy stress (Hainsworthet al. 1977). This also - accords with in other of S0.4 findings many species * I birds in which a nightly hypothermiahas been CY) .-E- 0.0 described as a response to energy stress (Bie- bach 1977, Reinertsenand Haftorn 1984, Graf 0 Eupetomenamacroura 0.8 et al. 1989). Interestingly,the present study has revealed that, in M. fuscus and E. macroura, those individuals that did not resume a torpid 0.4 state during the night, but suffered from a long fasting period before dusk, could decrease their overnightenergy expenditure(Fig. 4). In E. ma- 0.0 croura this was paralleledby a decrease in Tb 0 2 4 6 8 10 12 as well. even before reach Timein torpor(hours) Thus, hummingbirds the body mass threshold for torpor initiation FIGURE5. Overnightbody mass loss (% decrease they may enter a low metabolic/low Tb state, in bodymass-hr-1) in threespecies of hummingbirdswhich be similarto the as a functionof thetime in The may qualitatively nightly spent torpor. overnight seen in other of birds. The body mass losses (BML)expressed as a functionof hypothermia groups torportime (TT) are described by the followingequa- body temperatures (36.8-37.3'C) observed tions: BML = 0.790 - 0.039TT (r = 0.78, P < 0.05, when hummingbirdswere resting at night in a n = 22) for Amazilia versicolor; BML = 0.671 - nontorpidstate are within the range of nightly = P n = for Melanotro- 0.041TT (r 0.71, < 0.05, 29) body temperatures in hum- chilusfuscus;and BML = 0.604 - 0.036TT (r = 0.85, previously reported P < 0.05, n = 27) for Eupetomenamacroura. mingbirds(35.3-39.5'C; Prinzingeret al. 1991). In additionto the use of torpor as a defense mechanism against energy stress, some studies occurredin all three hummingbirdspecies. Thus have indicated that hummingbirdsalso may en- Tb was not regulated at the lowest level. Our ter into torpor during presumablynormal non- unpublisheddata, based on metabolic studies at stressed periods (Krtigeret al. 1982). Carpenter ambient temperaturesexperimentally lowered and Hixon (1988) showed that the Rufous Hum- below that of the outside conditions, indicate mingbirdmay become torpid duringthe premi- that M. fuscus regulates its Tb at 19-23TC, and gratory period, apparentlyin order to enable a E. macrouraregulates its Tb at 13-16TCduring more rapidbuild up of fat reserves.This recently torpor.We have no equivalent data for A. ver- has been supportedby furtherstudies (Carpenter sicolor, but for five other species of humming- et al. 1993, Hiebert 1993a, 1993b). Carpenter birds (Lophornis magnifica, Calliphlox amethys- (1974) reported torpor in the Andean Hillstar tina, Clytolaema rubricauda, Leucochloris al- HummingbirdOreotrochilus estella during nat- bicollis, and Phaethornis pretrei) from the same ural roosting conditions, and found a clear sea- area in Brazil, the regulatedbody temperatures sonal difference in the use of torpor;both the recorded during torpor range between 12 and number of incidences and the duration of the 180C (unpubl.data). Thus, at ambient tempera- torpor periods were greater during the winter. tures of between 19 and 23'C, as encounteredin Carpenter(1974) concludedthat neitherthe am- the present study, the hummingbirdsapparently bient temperature nor low food availability were not exposed to temperatureslow enough to could explain the use of torporby this species.

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions 786 CLAUS BECH ET AL.

In the Poorwill (Phalaenoptilus nuttalli), daily in hummingbirds,the birds in the present study torporalso may occur in free-rangingindividu- were exposed to naturalphotoperiod, tempera- als independentof their energeticstate (Brigham ture and sounds. Thus, it is likely that the mul- 1992). The results of the present study comple- tiple torporbouts could be related to the exper- ment the above observationswith the Versicol- imental conditions. Any external sound stimuli ored Emerald Hummingbird A. versicolor, from the tropical night could have initiated which may undergo nightly torporduring peri- arousal. However, we never detected any obvi- ods of apparentlyhigh food availability,as was ous external stimuli that occurred during the indicated by the use of torpor in an individual nights on which we observed multiple bouts of caught at normalroosting time (Fig. 4). For the torpor.In addition, often only one of the indi- other two species studied, a fasting period viduals studied during a single night would ex- seemed to be necessary for the inductionof tor- hibit multiple torporbouts, while the othershad por. The shortest time of food deprivationnec- either one long period, or did not enter torporat essary to induce torporin M. fuscus and E. ma- all. This obviously could stem from differences croura was 100 and 20 min, respectively (Fig. in energetic state and differentlevels of suscep- 4). Thus, our data suggest that E. macrouraalso tibility to externalstimuli. Regardless of the rea- has the ability to entertorpor under normal, non- son for multiple nightly torpor bouts, further stressed circumstances.M. fuscus, on the other studies employinghummingbirds in theirnatural hand, seems to be less prone to enter torpor, habitat are needed to establish how widespread since torporwas not recordedeven after fasting this pattern is. Most earlier studies were con- periods of up to 100 min durationbefore dusk ducted on hummingbirdswith lower body tem- (Fig. 4). Heavy rain or cold weather probably peraturesduring torpor. Assuming a gradualloss could hinderhummingbirds in feeding for a pe- of response with body temperature,the relative- riod before dusk, and it is conceivable that pe- ly high Tb during torpor in the tropical hum- riods of food-deprivationof up to 100 minutes mingbird species of the present study might could indeed occur under normalcircumstances cause a higher degree of susceptibilityto exter- in the study area. There are no apparentinter- nal stimuli, in contrast to hummingbirdsfrom specific differences in the biology of the three temperateand montaneareas, which often have species which could explain the observed dif- lower levels of torpor body temperatures(5- ferences in torporpattern. 10'C; Calder and Booser 1973, Calder 1974, The three species of hummingbirdsused in Carpenter1974). the presentstudy were able to enter torpormore When hummingbirdsexperienced these inter- than once duringa single night (Fig. 1; see also ruptedtorpor periods, they would invariablyen- Bech et al. 1994). This implies that some of the ter torpor again after having reached the nor- single torporperiods were shortduration, lasting mothermic nightly Tb level (Fig. 1). This still for only a few hours. Hainsworthet al. (1977) lends credit to the theory that a threshold-value also reported torpor periods of only 2.5 hr in of body mass is operating (Hainsworthet al. Rivoli's Hummingbird Eugenes fulgens and 3.5 1977, Hiebert 1992). The observationof multi- hr in the Black-chinnedHummingbird Archilo- ple torporbouts, on the other hand,raises a fun- chus alexandri, whereas Hiebert(1990) showed damentalquestion about the energeticsof torpor, that torporbouts of 2.5-3.0 hr could occur late namely whetherthe individualswill still have an in the night in Rufous Hummingbirds.Thus, energetic advantagefrom such very shorttorpor most.hummingbirdsmay have the ability to en- bouts. Ourresults indicatethat some individuals ter torporfor short periods at a time. However, may not even utilize the full time requiredto the present description of multiple periods of entertorpor, but may actuallyarouse from torpor torpor in hummingbirdswould seem to be the before their body temperaturehas reached its first report of such cases. We cannot offer any lowest level. Thus, Tb only was lowered to a explanation as to the cause of such a pattern, value between the normothermicand normal tor- which seems to conflict with the assumptionthat pid values (A. versicolor, Fig. 2). However,our there is a minimumbody mass (set-point)below data do not allow us to test whether such very which the hummingbirdis obliged to enter tor- short periods of torporare of thermoregulatory por (Hainsworthet al. 1977). However, in con- significance to the birds. The hummingbirds trastto most other (laboratory)studies on torpor would benefit from these torporbouts if the cost

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions TORPOR IN BRAZILIAN HUMMINGBIRDS 787 of arousal would not counteract the, even avian energetics.Publ. Nuttall Ornithol.Club 15: 86-144. short-term, reduction in body temperature and CALDER,W. A. 1994. Whendo hummingbirdsuse tor- metabolic rate. Recent data obtained from both por in nature?Physiol. Zool. 67:1051-1076. hummingbirds (Hiebert 1990) and mammals CALDER,W. A., ANDJ. BOOSER.1973. Hypothermia (Ruf and Heldmaier 1992) seem to indicate that of Broad-tailedHummingbirds during incubation homeotherms will in fact benefit energetically in nature with ecological correlations. Science 180:751-753. from bout regardless of its length. any torpor CALDER,W. A., L. L. CALDER,AND T. D. FRAZIER. In we have shown that there are summary, 1990. The hummingbird'srestraint: a natural large interspecific differences in the use of tor- model for weight control. Experientia46:999- por between the three Brazilian hummingbird 1002. species studied. Whereas A. versicolor and E. CARPENTER,F L. 1974. Torporin an Andeanhum- macroura seem to enter mingbird:its ecological significance.Science 183: torpor very readily 545-547. without M. any previous energy stress, fuscus CARPENTER,E L., AND M. A. HIXON. 1988. A new apparently only enter torpor when energetically functionfor torpor:fat conservationin a wild mi- stressed (low body mass). The reason for these granthummingbird. Condor 90:373-378. differences is unknown. All three species will at CARPENTER,F L., M. A. HIXON,C. A. BEUCHAT,R. W. AND D. C. PATON. 1993. mass times have multiple torpor bouts during the RUSSELL, Biphasic We that this is caused inter- gain in migranthummingbirds: body composition night. suggest by changes, torpor,and ecological significance.Ecol- ruption of the normal torpor period by external ogy 74:1173-1182. stimuli. FRENCH,A. R. 1993. Hibernationin birds:comparison with mammals,p. 43-53. In C. Carey,G. L. Flor- ACKNOWLEDGMENTS ant, B. A. Wunder,and B. Horwitz [eds.], Life in the cold: ecological, physiological and molecular The study was initiatedby the late Kjell Johansen.Fi- mechanisms.Westview Press, Boulder,CO. nancial supportwas generouslyprovided by the Erna GRAF,R., S. KRISHNA,AND H. C. HELLER.1989. Reg- and Victor HasselbladsStiftelse (to KJ and JFS), Ib ulated nocturnalhypothermia induced in pigeons HenriksensFond (to KJ and JFS), the Danish Natural by food deprivation.Am. J. Physiol. 259:R733- Science Research Council (to KJ and JFS), and the R738. Deutsche Forschungsgemeinschaft(grant Be 536 to HAINSWORTH,F R., B. G. COLLINS,AND L. L. WOLF. MB). We also wish to thank the Brazilian National 1977. The function of torpor in hummingbirds. Research Council (CNPq) for its support (grants to Physiol. Zool. 50:215-222. to ASA, JFS and JEPWB) and for permission carry HAINSWORTH,F R., M. E TARDIFF, AND L. L. WOLF. out this study in Brazil (permitno. EX-16/87). Thanks 1981. control for also are due to the staff at the Museu de Prof. Proportional daily energy reg- Biologia ulation in hummingbirds.Physiol. Zool. 54:452- Mello Leitao in SantaTeresa for their invaluablesup- 462. port duringour stay. We are gratefulto Philip A. Tal- H. C. 1989. and tor- lantireand H. for the HELLER, Sleep, hypometabolism, Stephen Clayborough correcting por in birds, p. 231-245. In C. Bech and R. E. languageand two refereesfor valuablesuggestions on the Reinertsen [eds.], Physiology of cold adaptation manuscript. in birds. PlenumPress, New York. HIEBERT,S. M. 1990. Energy costs and temporal LITERATURE CITED organization of torporin the Rufous Humming- A. S. J. E M. AND bird (Selasphorus rufus). Physiol. Zool. 63: BECH,C., ABE, STEFFENSEN, BERGER, 1082-1097. J. E. P. W. BICUDO.1994. Multiplenightly torpor bouts in 323-328. In K. Plesch- HIEBERT,S. M. 1992. Time-dependentthresholds for hummingbirds,p. initiationin the Rufous ka and R. Gerstberger[eds.], Integrativeand cel- torpor Hummingbird(Se- J. 162:249-255. lular aspects of autonomicfunction: temperature lasphorusrufus). Comp. Physiol. and osmoregulation.John Libbey Eurotext,Paris. HIEBERT,S. M. 1993a. Seasonalchanges in body mass BEUCHAT,C. A., S. B. CHAPLIN,AND M. L. MORTON. and use of torpor in a migratoryhummingbird. 1979. Ambient temperatureand the daily ener- Auk 110:787-797. getics of two species of hummingbirds,Calypte HIEBERT, S. M. 1993b. Seasonalityof daily torporin anna and Selasphorus rufus. Physiol. Zool. 52: a migratoryhummingbird, p. 25-32. In C. Carey, 280-295. G. L. Florant, B. A. Wunder,and B. Horwitz BIEBACH,H. 1977. Reduktiondes Energiestoffwech- [eds.], Life in the cold: ecological, physiological sels und der KOrpertemperaturhungernder Am- and molecular mechanisms. Westview Press, seln (Turdusmerula). J. Ornithol.118:294-300. Boulder,CO. BRIGHAM, R. M. 1992. Daily torporin a free-ranging KRUGER,K., R. PRINZINGER,AND K.-L. SCHUCHMANN. goatsucker,the Common Poorwill (Phalaenopti- 1982. Torporand metabolismin hummingbirds. lus nuttallii).Physiol. Zool. 65:457-472. Comp. Biochem. Physiol. 73A:679-689. CALDER,W. A. 1974. Consequencesof body size for PRINZINGER,R., A. PRESSMAR,AND E. SCHLEUCHER.

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions 788 CLAUS BECH ET AL.

1991. Body temperaturein birds. Comp. Bio- RUF,T., ANDG. HELDMAIER.1992. The impact of daily chem. Physiol. 99B:499-506. torporon energy requirementson the Djungarian REINERTSEN,R. E. 1983. Nocturnalhypothermia and hamster, Phodopus sungorus. Physiol. Zool. 65: 994-1010. its energeticsignificance for small birds living in the arctic and subarctic A review. Polar RusCHI,A. 1982. Hummingbirdsof state of Espirito regions. Santo. Editora Sao Brazil. Res. 1:269-284. Rios, Paulo, WANG,L. C. H. 1989. Ecological, physiological,and AND S. 1984. The effect REINERTSEN,R. E., HAFTORN. biochemical aspects of torpor in mammals and of short-timefasting on metabolismand nocturnal birds, p. 361-401. In L. C. H. Wang [ed.], Ad- hypothermiain the Willow Tit Parus montanus.J. vances in comparativeand environmentalphysi- Comp. Physiol. 154:23-28. ology, Vol. 4. Springer-Verlag,Berlin.

This content downloaded from 186.217.234.34 on Wed, 15 Jan 2014 07:54:58 AM All use subject to JSTOR Terms and Conditions