Frequent arousals from winter torpor in Rafinesque's big-eared bat

12 Pages·2015·0.66 MB·English

UUnniivveerrssiittyy ooff KKeennttuucckkyy
UUKKnnoowwlleeddggee
Forestry and Natural Resources Faculty
Forestry and Natural Resources
Publications
11-21-2012
FFrreeqquueenntt aarroouussaallss ffrroomm wwiinntteerr ttoorrppoorr iinn RRaafifinneessqquuee''ss bbiigg--eeaarreedd bbaatt
((CCoorryynnoorrhhiinnuuss rraafifinneessqquuiiii))
Joseph S. Johnson
University of Kentucky, [email protected]
Michael J. Lacki
University of Kentucky, [email protected]
Steven C. Thomas
United States National Park Service
John F. Grider
University of North Carolina, Greensboro
Follow this and additional works at: https://uknowledge.uky.edu/forestry_facpub
Part of the Forest Sciences Commons
RRiigghhtt cclliicckk ttoo ooppeenn aa ffeeeeddbbaacckk ffoorrmm iinn aa nneeww ttaabb ttoo lleett uuss kknnooww hhooww tthhiiss ddooccuummeenntt bbeenneefifittss yyoouu..
RReeppoossiittoorryy CCiittaattiioonn
Johnson, Joseph S.; Lacki, Michael J.; Thomas, Steven C.; and Grider, John F., "Frequent arousals from
winter torpor in Rafinesque's big-eared bat (Corynorhinus rafinesquii)" (2012). Forestry and Natural
Resources Faculty Publications. 1.
https://uknowledge.uky.edu/forestry_facpub/1
This Article is brought to you for free and open access by the Forestry and Natural Resources at UKnowledge. It has
been accepted for inclusion in Forestry and Natural Resources Faculty Publications by an authorized administrator
of UKnowledge. For more information, please contact [email protected].
FFrreeqquueenntt aarroouussaallss ffrroomm wwiinntteerr ttoorrppoorr iinn RRaafifinneessqquuee''ss bbiigg--eeaarreedd bbaatt ((CCoorryynnoorrhhiinnuuss
rraafifinneessqquuiiii))
Digital Object Identifier (DOI)
http://dx.doi.org/10.1371/journal.pone.0049754
NNootteess//CCiittaattiioonn IInnffoorrmmaattiioonn
Published in PLoS ONE, v. 7, no. 11, e49754.
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted,
modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available
under the Creative Commons CC0 public domain dedication.
This article is available at UKnowledge: https://uknowledge.uky.edu/forestry_facpub/1
Frequent Arousals from Winter Torpor in Rafinesque’s
Big-Eared Bat (Corynorhinusrafinesquii)
Joseph S. Johnson1*, Michael J. Lacki1, Steven C. Thomas2, John F. Grider3
1DepartmentofForestry,UniversityofKentucky,Lexington,Kentucky,UnitedStatesofAmerica,2UnitedStatesNationalParkService,MammothCave,Kentucky,United
StatesofAmerica,3DepartmentofBiology,UniversityofNorthCarolina,Greensboro,NorthCarolina,UnitedStatesofAmerica
Abstract
Extensive use of torpor is a common winter survival strategy among bats; however, data comparing various torpor
behaviorsamong species are scarce. Wintertorpor behaviors arelikely to varyamong species with differentphysiologies
and species inhabiting different regional climates. Understanding these differences may be important in identifying
differing susceptibilities of species to white-nose syndrome (WNS) in North America. We fitted 24 Rafinesque’s big-eared
bats(Corynorhinusrafinesquii)withtemperature-sensitiveradio-transmitters,andmonitored128PIT-taggedbig-earedbats,
duringthewintermonthsof2010to2012.WetestedthehypothesisthatRafinesque’sbig-earedbatsusetorporlessoften
thanvaluesreportedforotherNorth Americancave-hibernators.Additionally,wetestedthehypothesisthatRafinesque’s
big-earedbatsarouseonwinternightsmoresuitablefornocturnalforaging.Radio-taggedbatsusedshort(2.4 d60.3(SE)),
shallow (13.9uC 6 0.6) torpor bouts and switched roosts every 4.1d 6 0.6. Probability of arousal from torpor increased
linearlywithambienttemperatureatsunset(P,0.0001),and83%(n=86)ofarousalsoccurredwithin1 hrofsunset.Activity
of PIT-tagged bats at an artificial maternity/hibernaculum roost between November and March was positively correlated
with ambient temperature at sunset (P,0.0001), with males more active at the roost than females. These data show
Rafinesque’sbig-earedbatisashallowhibernatorandisrelativelyactiveduringwinter.Wehypothesizethatwinteractivity
patternsprovideCorynorhinusspecieswithanecologicalandphysiologicaldefenseagainstthefunguscausingWNS,and
thatthesebatsmaybebettersuitedtowithstandfungalinfectionthanothercave-hibernatingbatspeciesineasternNorth
America.
Citation:JohnsonJS,LackiMJ,ThomasSC,GriderJF(2012)FrequentArousalsfromWinterTorporinRafinesque’sBig-EaredBat(Corynorhinusrafinesquii).PLoS
ONE7(11):e49754.doi:10.1371/journal.pone.0049754
Editor:R.MarkBrigham,UniversityofRegina,Canada
ReceivedJuly4,2012;AcceptedOctober17,2012;PublishedNovember21,2012
Thisisanopen-accessarticle,freeofallcopyright,andmaybefreelyreproduced,distributed,transmitted,modified,builtupon,orotherwiseusedbyanyonefor
anylawfulpurpose.TheworkismadeavailableundertheCreativeCommonsCC0publicdomaindedication.
Funding:FundingforthisresearchwasprovidedbytheUniversityofKentucky,CollegeofAgriculture.Thefundershadnoroleinstudydesign,datacollection
andanalysis,decisiontopublish,orpreparationofthemanuscript.
CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.
*E-mail:[email protected]
Introduction causalagentofWNSisacold-tolerantfungus,Geomycesdestructans,
which is believed to be of Old World origin and only recently
Manybatspeciesinhabitingtemperateclimatesrelyheavilyon introducedtoNorthAmerica[13,14,15,16].Manybatspresumed
torpor during winter to survive extensive periods of cold and
to have died from WNS have little to no fat reserves remaining,
reduced food availability [1]. Winter torpor, or hibernation,
leading to the hypothesis that fungal infection causes more
typically consists of numerous discrete torpor bouts separated by
frequent and/or longer duration arousals during hibernation
brief periods of normothermy, or arousal, which are proposed to
[7,16,17,18]. Ultimate and proximate causes of mortality in
serve numerous physiological purposes [2,3,4,5]. Laboratory and
infected bats are likely to be more synergistic, however. For
field studies of winter torpor show that frequency of periodic
example,erosion oftheskinfromfungal invasionmaynegatively
arousals varies among species, as do minimum torpid body
affectwaterbalance,resultingindehydration,increasedfrequency
temperatures (T ) [1,6,7,8]. Together, depth and duration of
b of periodic arousals, and additional complications with thermo-
torporboutsarepartofanenergyconservationstrategyinwhich
regulation [19,20].
successful hibernators balance the costs and benefits of arousals
NeithertheclinicalsignsofWNSorthepresenceofG.destructans
with costs and benefits of torpor [9,10,11]. Although data from
have been observed in bats in the genus Corynorhinus (big-eared
fieldstudiesarescarce,itislikelythatwintertorporstrategiesvary
bats). Although expansion of the fungus into the range of
within and among species, with animals possessing higher body
Corynorhinusspeciesiscurrentlylimited,WNShasbeendocumen-
massandlowersurfaceareatovolumeratiosabletosurvivemore
ted in five caves used by the endangered Virginia big-eared bat
frequent arousals, longer periods of normothermy, or higher
(Corynorhinus townsendii virginianus), including the largest known
torpid body temperatures [9,12]. Furthermore, regional climates
hibernaculum of the species, without visibly affecting any big-
and the unique ecology and life history of each species are also
earedbat[21].Currently,experimentaldatashowingCorynorhinus
likelytoimpact seasonaltorpor strategies.
speciestobelessvulnerabletoWNSarelacking,anditisunknown
Understanding differences in winter torpor behaviors among
whetheroneormoreaspectsoftheecologyorphysiologyofthese
bat species is of heightened interest due to fatal disruptions in
bats could providea defense against fungal infection.
hibernation resulting from white-nose syndrome (WNS). The
PLOSONE | www.plosone.org 1 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Rafinesque’sbig-earedbat(Corynorhinusrafinesquii)isasmall(8– ANOVAs. We tested all datasets for homogeneity of variance
14g) forest-dwelling bat found throughout the southeastern using Variance Ratio F -tests.
max
United States [22,23]. Rafinesque’s big-eared bat hibernates in A subset of bats were fitted with 0.52g temperature-sensitive
cavesandminesinmountainousandkarstportionsofthespecies radio-transmitters (model LB-2T, Holohil Systems, Ltd., Carp,
range,butthesehibernaculaarelargelyabsentinsouthernportion Ontario) immediately below the shoulder blades using surgical
of the range, where big-eared bats have been documented adhesive (Perma-Type, Plainville, CT) [33]. We placed radio-
hibernating in trees, wells, and cisterns [23,24,25]. While there transmitters belowtheshoulderblades toavoid concentrationsof
are no data on the depth or duration of winter torpor bouts in brownadiposetissuewhichmayleadtoerrorsinestimationofT
b
Rafinesque’sbig-earedbat,observationsofwintermatingandbats during periods of non-shivering thermogenesis [34]. Bats were
frequently moving among hibernacula suggest this species may placed back in their roosts after the adhesive was allowed to dry
undergo torpor bouts of shorter duration than other cave- (ca. 15min), by which time they had aroused from torpor. We
hibernating bat species in eastern North America, making it an deployed HOBO dataloggers (models U23-001 or UA-002-08,
ideal species forwinter studies[22,26,27]. Onset Computer Corporation, Bourne, MA) inside and outside
HereinweevaluatewintertorporbehaviorsinRafinesque’sbig- roosts after releasing radio-tagged bats. Dataloggers recorded air
eared bat. We hypothesized that Rafinesque’s big-eared bats use temperatures inside roosts (Tr) and outside roosts (hereafter
less torpor than other cave-hibernating bat species in eastern ambient temperature, Ta) at 15-min intervals. Dataloggers re-
North America, predicting that Rafinesque’s big-eared bats use cording ambient temperatures were placed inside solar radiation
shallower, shorter duration torpor bouts than those observed in shields.
other eastern North American species. Additionally, we hypoth- Each radio-transmitter was individually calibrated by the
esized that Rafinesque’s big-eared bats are active foragers manufacturer, providing a unique polynomial equation for use
throughout the winter, predicting that arousals would be inconvertingtransmitterpulserateintoskin-temperature(Tsk).Tsk
associated with sunset, and would be more likely to occur on of each radio-tagged bat were recorded by three datalogging
warmer nights. receivers (model R4500S, Advanced Telemetry Systems, Inc.,
Isanti, MN) placed in watertight boxes with an external power
source.Receiverswereprogrammedtoscanforradio-taggedbats
Materials and Methods
at 5-min intervals and placed outside caves and abandoned
Study Area buildings. Receivers were checked weekly or bi-weekly for
FieldworkoccurredatMammothCaveNationalPark(MCNP; maintenance and moved to new locations when necessary. We
37.2072u N, 86.1319u W) in Barren, Edmonson, and Hart attempted to locate bats in nearby caves and buildings if their
counties, Kentucky, USA. The area is predominantly forested radio-signals were not heard outside monitored hibernacula.
and is dissected by numerous small drainages, creating a topo- Chronological accounts of the roost location of each bat were
graphicallydiverselandscape.Forestcoverconsistsofoak-hickory determined based upon which receiver recorded daytime signals.
(Quercus2Caryaspp.)andwesternmixedmesophyticforests[28]. Only two roosts (buildings) were located close enough to each
other that bats in either roost could be recorded by a single
Duringsummer,Rafinesque’sbig-earedbatroostsinhollowtrees,
receiver.Recordedsignal strengthdiffered notablybetweenthese
sandstoneoutcrops,cavesandabandonedman-madestructures(J
roosting sites, however, allowing for a clear determination of
Johnson, unpublished data). Hundreds of caves occur within the
roostinglocation.Batswhichcouldnotbelocatedforseveraldays
21,380 ha Park, including six known hibernacula of Rafinesque’s
(always #10d) were considered to be in the same roost for the
big-eared bat. The Park has one of the largest known winter
entire period. We compared roost-switching frequencies (i.e.,
concentrations of Rafinesque’s big-eared bats, with .1000 big-
numberofdaysabatinhabitedaroostbeforeswitchingtoanew
eared bats hibernating withinthePark [23].
roost) between sexes and among winter periods using a two-way
ANOVA.
Radio-telemetry Data Collection and Analysis
We applied Willis’ [35] equation for an energy-based temper-
All methods were approved by the University of Kentucky
ature threshold for torpor onset (T ), using the conservative
Institutional Animal Care and Use Committee (IACUC No. onset
equationbaseduponmodelparametersminus1SE.Thisequation
A3336-01)andtheNationalParkService(NPSIACUCNo.2011-
requires a simultaneous measure of T, andwewere only ableto
30).WecapturedRafinesque’sbig-earedbatshibernatingincaves r
calculate T when bats occupied roosts with HOBO datalog-
and abandoned buildings during three consecutive winters and onset
gers. Calculated values for T varied marginally (between
radio-tagged bats during March 2010, January 2011, and onset
31.5232.3uC),however,andweappliedaT valueof32uCto
November2011–January2012.Wecombineddataacrosswinters onset
all bats. Thus, we considered bats to be torpid when T was
anddividedthedatasetintobatsradio-trackedduringearly-(mid- sk
,32uC, and considered torpor bouts over when T .32uC or if
sk
November–early December), mid- (mid-December–mid Febru-
radio-signals were lost following a rapid rise in T ; presumably
sk
ary), and late-winter (mid-March–early April) to account for
signifying the bats left the roost before normothermic T was
sk
variability in torpor bout duration associated with progression of
recorded.
the hibernation season [29,30,31]. No radio-tagged bat was
AlthoughWillis’[35]equationwasdesignedforusewithT ,not
tracked during .1 winterperiod. b
T ,itstillholdsadvantagesoveranarbitrarydefinitionoftorpor
sk
We recorded age, sex, reproductive condition, right forearm
because T in bats is a good indicator of T [36]. Because T is
sk b sk
length,andbodymassofcapturedbats.Wedeterminedthebody typically a few degrees below T , especially at low T , use of the
b a
conditionofallbut4bats(forearmlengthswerenotmeasuredfor T equation likely results in a T slightly below that which
onset onset
4 individuals) by dividing body mass by forearm length [32]. We would be obtained using T measurements. However, error in
b
compared body condition between sexes and among winter determiningwhenbatsinourstudyenteredtorporresultingfrom
periodsusingatwo-wayanalysisofvariance(ANOVA).Weused useofT islikelyonthescaleofseveralminutesbecausewedid
sk
a0.05significancelevelfordifferenceandcomparedleastsquares not observe T lingering between 32uC and other typical torpor
sk
means using Tukey’s adjustment when significant, for all cutoffs,suchas25uC[37].Instead,T quicklydroppedtovalues
sk
PLOSONE | www.plosone.org 2 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
,20uC while bats roosted in hibernacula. Error in determining We performed a linear regression (PROC REG, SAS) to
arousal from torpor may be greater because of the concentration evaluatetheroleofambienttemperaturesonwinteractivity,and
of brown adipose tissue located beneath the skin, potentially test our prediction that bats would be more active on nights
leading to erroneous measures of T . While our data are suitable for nocturnal foraging. We used number of marked bats
sk
insufficient todeterminetheextent ofthiserror,wecontend that recorded each day during winter as the response variable, T at
a
it is likely small, because initiation of rapid heat production sunsetastheindependentvariable,andasignificancelevelof0.05.
throughnon-shiveringthermogenesislikelysignifiesthatthebatis Winter was defined as 01 November–01 March. We did not
either undergoing spontaneous arousal or maintenance of a T include data from March 2012 in this analysis due to unusually
b
notably higher than T. In the former scenario, error in high temperatures and high activity of bats at the roost (see
r
determining the length of torpor bouts is likely to be a small Results).Aseparateanalysiswasconductedforadultmales,adult
periodoftime.Inthelatterscenario,torpidT shouldnotexceed females, male young-of-the-year, andfemale young-of-the-year.
sk
T .
onset
Bats were considered torpid for the entire time they were not Results
recorded by any datalogging receiver (always #10d). While this
Radio-telemetry
might overestimate torpor bout duration while un-located (bats
mayhavearousedandnotleftthehibernaculumduringthistime), We captured and measured 33 bats during periodic roost
it provides a conservative measure for testing the prediction that searches. Body condition differed among bats (F3, 29=60.9,
P,0.0001),withdifferencesdetectedamongbatscapturedduring
Rafinesque’s big-eared bats use shorter duration torpor bouts
different winter periods (F=90.9, P,0.0001), but not between
compared to other species. We determined the duration (days),
sexes (F=0.39, P=0.54). Mean body condition was greatest
minT ,andaverageT ofeachtorporbout,andaveragedeach
sk sk
during early-winter (0.2860.01) compared to mid- (0.216.001,
measurewithinbatsforstatisticalanalysis.Wealsodeterminedthe
P,0.0001) and late-winter (0.1960.001, P,0.0001), and mean
average duration (hrs) of normothermic periods when entry and
body condition was greater during mid-winter than late-winter
arousalfromtorporweresuccessfullyrecordedfortwoconsecutive
(P=0.006). Body conditions of males and females both averaged
torpor bouts. Each variable was compared between sexes and
0.2360.01. Body mass of females averaged 10.060.46, versus
among winterperiods usinga two-way ANOVA.
10.060.52 for males. Range in body mass declined from 11.4–
We determined the time difference (hrs) between arousal and
14.0 g(bodyconditions:0.27–0.32)inmid-Novemberto7.1–8.3 g
sunsetondayswherewerecordedtorpidT datapriortoarousal.
sk (body conditions:0.16–0.20)in early March.
We used generalized estimating equations (PROC GENMOD,
Weradio-tagged14femaleand10maleRafinesque’sbig-eared
SAS) to assess the role of varying ambient temperatures on the
batsbetween2010and2012.Allbatsre-enteredtorporfollowing
probability of arousal from torpor [38]. Generalized estimating
radio-taggingbutarousedfromtorporwithin1 hrofsunset.These
equationsareanextensionofthegenerallinearmodel,butdonot
initial bouts were not included in analyses or summaries. Four
provide correlation coefficients. The strength of generalized
femaleswereneverlocatedfollowingemergencefromhibernacula
estimating equations is that they allow for analysis of clustered
onthefirstnightandwerenotincludedinanytorporanalysesor
(i.e., many days of observation from an individual bat which are
summaries. Bats switched roosts every 4.1d 60.6, with no
notindependentsamples)binary(i.e.,torpidoractive)datawhich
difference detected between sexes or among winter periods (F
cannot befitted usingtypical linear models. 3,
=0.88, P=0.47). Bats traveled 25356437m (range=549–
We calculated the Heterothermy Index (HI) for each bat to 16
5964)betweenconsecutiveroosts.WewereabletodetermineHI
quantifyvariabilityinT ofeachindividual[39].Wefollowedthe
sk values for 8 males and 8 females. Optimal skin temperatures
recommendationofBoylesetal.[39]tousethemoderepresenting
ranged from 28–34uC (mean=32.860.4) and HI ranged from
the greatest T as T . We used all of the recorded T for
b b-opt sk 14.4–20.7(mean=17.760.5).
individual bats whendetermining each HIvalue.
We documented 7.461.2 (range=1–16, n=147) torpor bouts
perbat.Batsarousedfromtorporevery2.4d60.3(Fig.1A,2A),
PIT-tagging
withnodifferencedetectedbetweensexesoramongwinterperiods
We used a harp-trap (Faunatech, Bairnsdale, Victoria, Aus- (F =0.94, P=0.44). Duration of normothermic periods
3, 16
tralia)tocaptureRafinesque’sbig-earedbatsexitingaman-made between torpor bouts differed among bats (F =5.72,
3, 12
roost during April and August 2011. We measured each bat as P=0.011), with differences detected among winter periods
described above, and subcutaneously implanted a 12.5mm PIT (F=8.27, P=0.006, Fig. 2B), but not between sexes (F=0.01,
tag (model TX1411SST, Biomark, Inc., Boise, ID), sealing the P=0.91, Fig. 1B). Duration of normothermy was shorter during
injectionsitewithsurgicaladhesive(Perma-Type,Plainville,CT), mid-winterthanearly(P=0.039)andlatewinter(P=0.008).Only
as part of a long-term study overseen by MCNP’s Science and one torpor bout was successfully monitored for four bats radio-
Resources Management Division. Roost entrances were sur- tagged during January 2011, and duration of normothermic
rounded by custom-made antennas designed to read and record periods were not determined for these bats. Skin temperatures
PIT tags of marked bats entering or exiting the roost. We infrequentlyfellbelow10uCduringtorpor.AveragetorpidT was
sk
summarized PIT tag readings into the number and identity of 13.9uC 60.6 (Fig. 1C, 2C), with no difference detected between
taggedbatspassingthroughthesensorfieldeachdaybetween03 sexesoramongwinterperiods(F =1.74,P=0.20).Minimum
3,16
May2011and01April2012.Becausebatswereoftenrecordedat torpidT averaged12.1uC60.8(Fig.1D,2D),withnodifference
sk
the roost more than an hour after sunrise and an hour prior to detected between sexes or among winter periods (F =1.87,
3, 16
sunset,weconsideredadaytobethe24hrperiodbetween1200 P=0.18).
and1200thefollowingday.Thisallowedustoefficientlyprocess Increasingambienttemperatureatsunsetsignificantlyincreased
large quantities of data while ensuring we captured all nightly the probability of arousal (b=0.1360.03, P,0.0001), and the
activity in our summaries. We report the percentage of adult timing of arousals was centered on sunset, with 50% (n=51)
males, adult females, male young-of-the-year, and female young- occurring in the 30min following sunset, and 83% (n=86)
of-the-year recorded each day. occurringwithin61 hrofsunset(Fig.3).Themajorityofarousals
PLOSONE | www.plosone.org 3 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Figure 1. Winter torpor behaviors of male and female Rafinesque’s big-eared bats determined through radio-telemetry at
MammothCaveNationalPark,Kentucky,USA.Numberofradio-taggedbatsisincludedinparentheses.
doi:10.1371/journal.pone.0049754.g001
were documented while bats were hibernating in caves (n=84, detected on 40% of days (n=25). Daily activity of adult females
82%), where T’s were between 5u and 11uC throughout the begantoresemblefallactivitypatternsbyearlyMarch.Numberof
r
winter.Manybatshibernatingincavesexhibitedperiodsofrapid adult females detected per day during winter increased linearly
thermogenesis occurring within 1 hr of sunset that were not with T at sunset (r2=0.23, F =35.1, P,0.0001). Activity of
a 1,119
considered arousals, because T failed to reach 20uC before female young-of-the year was similar to that of adult females
sk
declining (Fig. 4). Temperatures inside buildings, where 18% (Fig. 5A, Fig. 6A), and also increased linearly with T at sunset
a
(n=19) ofarousals weredocumented, ranged from28to21uC. (r2=0.27,F =44.7,P,0.0001).
1,119
Daily activity of adult males declined steadily throughout the
PIT-tagging summer,butremainedrelativelyhighduringthewintercompared
WePITtagged128Rafinesque’sbig-earedbats(38adultmales, to females (Fig. 5B). More than 20% of PIT-tagged males were
71adultfemales,10juvenilemales,and9juvenilefemales).PIT- detectedon13%ofdaysbetween01Decemberand01February
tagsfrom13bats(11adultfemalesand2adultmales)werefound (n=8),whilefewerthan10%ofmalesweredetectedon71%days
atthebaseoftheroostbeforetheendofthestudy,eitherasaresult (n=44), and no male was detected on 32% of days (n=20).
ofmortalityorbeingshed,andthesebatswereremovedfromour Number of adult males detected per day during winter increased
analysis.DailyactivityofadultfemalesdeclinedinearlyOctober, linearly with Ta at sunset (r2=0.26, F1,119=41.8, P,0.0001).
and#20%ofalladultfemaleswererecordedonalldaysbetween Activity of male young-of-the year was similar to that of adult
04Decemberand01February(Fig.5A).Fewerthan10%ofPIT- males (Fig. 5B, Fig. 6B) and also increased linearly with Ta at
tagged adult females were detected on 89% of days between 01 sunset (r2=0.19,F1,119=28.1,P,0.0001).
December and 01 February (n=55), and no adult female was
PLOSONE | www.plosone.org 4 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Figure 2. Winter torpor behaviors of Rafinesque’s big-eared bats during early (mid-November–mid-December), mid- (mid-
December–mid-February), and late (March) winter determined through radio-telemetry at Mammoth Cave National Park,
Kentucky,USA.Numberofradio-taggedbatsisincludedinparentheses.
doi:10.1371/journal.pone.0049754.g002
Discussion mid-winter, short torpor bouts were also common during this
period,resultinginsubstantialvariationandlackofdifferencesin
Rafinesque’s big-eared bats hibernating in caves and aban-
bout duration among winterperiods.
doned buildings used short, shallow torpor bouts, and bats were
Similar patterns in torpor use and body conditions between
frequently active during winter. Torpor patterns did not differ
males and females throughout the winter are not consistent with
between male and female big-eared bats, but consistently varied
thethriftyfemalehypothesis[10,37].Thethriftyfemalehypothesis
with the progression of winter. Normothermic periods were
predicts that females are more conservative with fat reserves
significantly shorter during mid-winter, and while torpor bout
during winter because it is critically important to carry some of
durationandT didnotdiffersignificantlyamongwinterperiods,
sk these reserves into spring, when gestation begins. To accomplish
werecordedthelongesttorporboutsandlowestT ’sduringmid-
sk this, females use longer torpor bouts at lower temperatures than
winter. These findings are consistent with other mammalian
males. Males are not as energetically constrained in spring, and,
hibernators, where torpor bouts are longest, and normothermic
therefore, opt to avoid some of the ecological and physiological
periods shortest, during the middle of the hibernation season
costsoftorporthroughuseofshallowertorpor.Further,wefound
[29,30,31]. Probability of arousals increased with increasing
noevidencethatheavierbats,orbatswithahigherbodycondition
aboveground T at sunset, and it is likely that the numerous
a index,exhibitedhigherminimumoraverageT ’s,whichhasbeen
warm daysoccurring duringthemid-winter of 2011–2012,when sk
found in studies of little brown myotis (Myotis lucifugus) [9].
50%ofdays(n=29)haddailyhightemperaturesexceeding10uC,
Comparisonofourresultstostudiesoflittlebrownmyotisshould
were partly responsible for the large variance in torpor duration
be interpreted with caution, however, as our sample sizes were
amongbats.Thus,whilethelongesttorporboutsoccurredduring
PLOSONE | www.plosone.org 5 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Figure3.Timingofperiodicarousalsfromhibernationinrelationtosunsetdeterminedthroughradio-telemetryatMammothCave
NationalPark,Kentucky,USA.
doi:10.1371/journal.pone.0049754.g003
Figure 4. Skin temperatures (red circles) of a female Rafinesque’s big-eared collected through radio-telemetry and concurrent
ambienttemperaturesoutsidethecave(blackline)recordedover12daysatMammothCaveNationalPark,Kentucky,USA.Shaded
areasincludehoursbetweensunsetandsunrise.
doi:10.1371/journal.pone.0049754.g004
PLOSONE | www.plosone.org 6 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Figure 5. Daily activity of PIT-tagged adult female (A) and male (B) Rafinesque’s big-eared bats at a man-made structure in
MammothCaveNationalPark,Kentucky,USA.SolidlineindicatesasecondPIT-taggingeffortinmid-August,increasingthenumberofPIT-
taggedbats.
doi:10.1371/journal.pone.0049754.g005
small, and because we did not concurrently monitor little brown ula, may play important roles in the breeding biology of this
myotis inour study area. species [26,27]. We did not document winter copulation, but
Although we present the first published data on winter torpor a large amount of stored sperm at the base of the tail was
patternsofRafinesque’sbig-earedbat,frequentmovementsduring immediatelyevidentinallcapturedmales,indicatingthatmating
winter by this species have been reported for decades [22]. likely occurs throughout the hibernation season. We also found
Rafinesque’s big-eared bat copulates during winter, and periodic evidence that suggests big-eared bats encounter more potential
arousals from hibernation, as well as switching among hibernac- matesduringthewinterthantheyencounterduringothertimesof
PLOSONE | www.plosone.org 7 November2012 | Volume 7 | Issue 11 | e49754
WinterArousalsofRafinesque’sBig-earedBat
Figure 6. Daily activity of PIT-tagged female (A) and male (B) Rafinesque’s big-eared bat young-of-the-year at a man-made
structureinMammothCaveNationalPark,Kentucky,USA.
doi:10.1371/journal.pone.0049754.g006
theyear.Bothmalesandfemalesswitchedhibernaculafrequently, distances reported for this species during summer research in
and traveled up to 5964m between consecutive roosts during other parts of the range [24,40]. On several occasions bats
winter months, compared to a maximum distance of 3395m travelled.4kmbetweenconsecutivehibernacula,onlytoreturn
observed during three years of summer research at Mammoth totheprevious roostafter several days.
CaveNationalPark(J.Johnson,unpublisheddata).Longdistance Frequent movement among distant hibernacula may serve to
movements between hibernacula were not uncommon, as we maximize gene flow among populations rarely interacting during
tracked 8 bats (40%; n=4 males, 4 females) .2 km between the summer. Similar movements among hibernacula, accompa-
consecutive hibernacula. By comparison, only 3 of 64 (5%) bats nied by mating, may also occur during the fall, but there are no
radio-tracked during the summer traveled .2km between published data of these behaviors in Rafinesque’s big-eared bat.
consecutiveroosts(J.Johnson,unpubl.data).Wetracked3female Our finding that both sexes make these movements does not
and 2 male bats .4 km between hibernacula, longer than support the hypothesis of Clark [27] that Rafinesque’s big-eared
PLOSONE | www.plosone.org 8 November2012 | Volume 7 | Issue 11 | e49754