Ecological, behavioral, and thermal observations of a peripheral population of Indiana bats (Myotis sodalis)

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/351360809 Ecological, behavioural, and thermal observations of a peripheral population of Indiana bats (Myotis sodalis) Conference Paper · January 1996 CITATIONS 10 1 author: Some of the authors of this publication are also working on these related projects: Master's Thesis View project Allen Kurta Eastern Michigan University 120 PUBLICATIONS2,885 CITATIONS SEE PROFILE All content following this page was uploaded by Allen Kurta on 06 May 2021. The user has requested enhancement of the downloaded file. -'F L:tr oe 3 a :4. 7A(\ 'JE OF.o >-n L o.= EEfi ^ =q :J--> :coB Are ---:" - o cs t-- I kH= ?-.Y 'F9\? (!'i " oaL € Ao"; (€ A- .:i (J-\< 5€-l" ^\o ,:l ^d H ;.n \OEad 9^TE \' {) ,.< @ avD .9 n& a)5; iaa H C'Q €cdq) EE F "ii'E kvwA = X.n * t+>Fi ' *N f \ l O-v {3? m :CgCd < b >'.9 d* + b E.F H :i t :i d.= v aca > Ecological, Behaviourill, and Thermai Observations of a Peripheral Popultrtion of Indiana Bats ( N[yotis sodalis) ALLEN KURTA, KIMBERL\' ,J . WILLIAMS' AND ROBERT MI ES ABSTRACT We studied roost-tree selection, inter-roost movements, and body temperatur-e oI z3 adult fernale ald juvenile Indiana bats over a three-year period. Bats selected their loost tree on the basis oI species and diameter, tut ttre amount of exfbliating bark was not a factor. These bats roosted exclusively in green ash (Frur,tirlrrs pennsylvanica)rvith a mean diameter of 4o.9 ! r.z cm (se; rl = z3). Unlike Indiana bats in the centre o[ therr range, those in Michigan never used silver maple (Acer saccharinum), nev;r roosted in a shaded area, and never roosted in a healthy, living tree. The bats chalged roost tlees every 2.9 days, on average, moving from r to r78 m to an alternate t;ee. Pregnant females changed roosts most often' and lactating bats, the least. Overall, roost-switching was not satisfactorily explained by hypotheses based on predator avoidance, interruption of pai-asite life cycles, tracking of food resources, or environmental conditions. The ephemeral nature of bark roosts may be a factor in frequent roost-switching, but it has not been adequately tested. The number of trees occupied by the group each year varied From five to 18, although two to fbur trees each year were used most heavily. Contrary to published information, these bats were not adversely affected by bodv temperatures o[ 35-4o'C. INTRODUCTION Guiclelines fbr rnanagement of enciangered species generally are derived from studies performed near the centre of an animal's geographic range' where populatiops are large and stable and, therefore, easier to work with than those on rhe periphery. Nevertheless, peripheral populations are worthy of attention. lf one accepts the classical view that peripheral PoPulations are wagilg il constant battle with extinction, then the declining health of an outlying grouP may be an early warlling of potential problerns l'or the nrain population. Conversely, recent allalyses suggest that peripfieral populirtions are more likely to survive than centrally located ones, and ultinrately, per.ipheral rlreas rnay become critical refugia for a sECTIoN lrt RoosflrlS Ecologlcal and Behavloural Observatlons spccics (l.ornolino anrl (llrirnncll rcql). Irr any cvcrlt, a Pruclcnt r)ranagc nrcnt plan slioulcl spccificallv adclrcss pcriphcral, as wcll as centrrrl, populations, [rccattsc vcgctativc urtcl clirnirtic corrditions oficn diff-cr across al spccics'rangc, ancl guidclincs lirrnrulatccl {br onc group n)ay not lrc applicablc to tlrc othcr (c.c., Lovis r99.1 ). -fhc Intlia'a bat ra.tcs acr.ss 'rrch tif-thc castcr.. U.itecl Statcs (-fltortison rgBz). l)istur.bancc tluring liibcrnation und intcntional nrocjification of.rnajor hibcrrac'la lccl t<l scrious cleclincs in p.pulation sizc during thc l96os and car.ly r97os, ancl conscclucrrtlv, thc species was declarcci cndanscrccl in r97j (llichtcr ct al. r993). Ilibernatirrg indiana birrs .tre now rvcll p|otcctccl, [rLrt thc spccics contirrrrcs to cicclinc, suggesting problcnrs in surnrner as r.vcll (Olalvson r9.S7). During thc warrn-weathe r rnorrths, thcsc bats ro<tst filintarily under. cxfbliating bark, in upland woocilots ancl liparian fi>rcsts.'lb clate, sumnlcr roosting habits l'rave been rnost extcnsively studicd in Illinois ((iarclrrcr.et al. r99r) and Missouri (Calahan r993), in tirc hcart of thc spccics' distribution. ln the prcscnt lcPort, rvc dcscribc a nrulti ycar sttrcly of a srrmme r Po1'rulation of Incliana bats, roosting only r5 knr fl.orn thc northcrn cdge of their range. OLrr goals wcrc to provicle bascline, ecological, and behirvioural data, applicablc ro thc lnanilscmenr of this specics in northern areas, zrncl to nrirkc conrPaIisons rvith previous stuciics per[ormed in the core of the spccies'rangc. wc \verc p.lrtrcula|ly rrrtcrcstecl in factors affect inc roost selection and roost lidelity, brlt rve also cxarnined the question of body tenrpcratr-rrc. Ilasccl on lalrolatorv cxpcrirnents, Hensharv and Folk (1966) suggested that body tentperatures of only 34-35oC were fhtal to this bat. If true, such a low, lethal body temperature rvould have a profound ccological inrpact, sl,erely lirniting the types of roost sites that this treedlvclling anirtral cor-rlcl usc. METHODS The study took place near Vermontville, Eaton County, Michigan, frorn 3o June r99t to r5 August r99j. Indiana bats were captured in mist nets rvhile flying over streams or in flight paths leading from their roosts. After capture, sonte dorsal fur rvas trirnrned rvith scissors, and a o.7-g radiotransnritter (Holohil Sysrenrs Inc., woodlawn, ontario) was attached using a surgical adhesive (Skin-llond). The roost tree was located on subsequent clays, Lrsing a thrcc-clenrcnt Yrgi antcltna and recciver (Wildlife Instrunrclrts, t'RX-20oos).'l)'rnsrnittcrs rcnrtrined attachecl and functional for u;r to u d.lys. I)opulation sizc *'as rlctcrrrinccl by counting bats as they left their roost, betrveen sunset rrnd 5o nrinutcs alicr sunset. we consistently observed trces that sheltcred an Incliana birt rvith a functioning transmitter. in addition, rvc rnadc counts at otl)cr roost trccs (originally dctected b1'radiotracking), r.vhen multiplc obscrvers rvere available, or when rve did not have a tr-ansrriitter on anv bat. 'lb cllarilctcrize tlre haltit,rt, rvc cstablishecl o.r-ha, circular plots surror,rncling eirch roost, and iclentrfied every tree > 10 cnl DIIH within the plots. I1-plots lbr adjaccnt r'oosts o',,erlappecl, rve clnly used data from the ro3 Thermal Observatlons Statlstlcs Maternlty Roost Parameters lirst loost tltirt rvits cliscovcrcrl to nr,rirrtain inclcpcrttlcncc o1-thc nicasrrrccl clianrctcr at [rrcirst hcight (ttnu ) anrl clctcrrrrinccl thc cach roost trcc arrcl thc prcl'crrccl cxit 1.loint rrsing;r clinonrctcr. clirssificcl cach trcc irccorrlirrg to tlrc anroLrrrt ol'loosc [r.rrk that trsirrs a systcnr rccorrrrncrrtlccl [ry Girldncr ct al. (rq9r). A rating rvas givcrt to trccs rvith > z5'Zr krosc lrirrk, nrcc'liunr ) r<l()6 but < low was ) o(Xr lrut < ro(kr. ciata. Wc hcight of Wc also wirs prescltI of high 25(Z), and In this rcport, wc includc irrlirrrnation obtainctl fl.orn fivc o{-cight roost tr.ecs prcviously clcscr.i[rctl in Kurta ct al. (199].1 ).Wc cliscarclcd clata on tltlcc tlccs (rclosts Ir, C, and ll) nrcntionccl in that paper bccause thosc roosts wcrc discttvcrctl only bv acciclcntirlly olrscrving bats leaving tfre trees .rt clusl<. 'l hc lcccnt tlcnron-stration tliat a sinrilar spccies, thc northern Lrat (M. stpttrttriotrolis ), also roostecl occasionally unclcr bark in this wetlanci ([]ostcr r993), ncccssitatcd tlrat wc re -cvaluatc thc clata rrsing only trccs locatccl [ry irctually ratlio-tracking Indiana lrats. ll.aclio-transnritters uscd irt r99z ancl r993 wcrc tcntpcr.rturc.scrrsitive, and all<trved us to deterrrine thc surface tcmperature of'thc bat while it was in thc roost. Data wcre rrot taken in a systenrirtic mirnner in 1992, but in r99j, body tcrlrpcratLlrc rvas rccorclcd cvery 3tl rninutes bctwccn sunrise .rncl sultsct. During botli years, arlbicnt tentperaturc in thc shirde was sinrultancc'rusly rneasured ',vitl.r a nrcrcury thermonlctcr. lo avoid distur, bancc, tcnrpcrirtLrrcs'"vcrc rccorclccl rnclrc tltan roo nr flr'rru the roost. Although lve lvould Iike to have rnonitored roost ambient temperature at the same tirle as body temperature, this was not practical because, as clcscribed latcr, the bats lrequcntly changed roosts and uscd many diffcrent trees. All nteans are follorvcd by * 6'.'. standard error anc.l thcn the range in parenthescs. Cornparisons betrveen t\vo groups were perfornted using the nolmal approxirlation to the Wilcoxon Two-sarnplc Tcst, whereas tests involving nror.e than two categories relied on the chi-scluared approximation to the Kruskal-Wallis Test. l'ests involving enumeration data (contingency tables) were pcrformed using chi-squared tests or, when expected frequencies fell below five, Fisher's Exact Test. We used Spearman's Rank -l'cst to examine potential correlation between variables. All tests were twotailed and uscd a significancc level of o.o5. RESUTTS We radio-tlacked z3 lndiana bats and located the roost trce on i3o of l4t clays that the transn'rittcrs \vere functional. The bats included six pregnant inclividuals, sevcn lirctating bats, five post-lactating/non-reproductive fcnrales, and five juvenilcs. l'hc bats roosted primarily under the exfoliating bark of z3 differenl trces. All roost trees were dead or nearly dead grccn ash that hacl lost nrost or all of thcir rnajor branches; any living blanches thirt rernaincd rvcre <l cln in dianreter. Average onir of roost trees was 4o.9 ! r.z (3o-52) cnt. Mean roost-tree height was z5.r * r.6 (B-+l) rn, and rnean cxir height was 9.9 t o.9 (r.a-r8) m. Iloost trees s ECTro N ttt Roostrng Analysls of WoodY Vegetatlon Evldence for Roost SelectlvltY Roost Loyalty Wlthln Years were on avcrage )8.7 + /.t (vr+/) rlt away lrorn lnother trcc ttse..l irs a roost in thc samc ycar ald V3 t 4.5 (r-rrr) m away from the nearest tree used as a roost during any of the thrce years. There were no significant differcnces among years in ouH, height, exit height, or nearest-neighbour clistance (all p >o.rz). All roosts were in an unshacled, 5-ha wetland and surrounded by up to r nr of standing wiiter. Our vcgetation analysis yielded nine non-overlapping plots that contailed 99 green ashes, 34 silver maples, nine American elms ( U/mus anrericann), and trvo barkless, unidentifiable trunks. Although this area was forestecl p.lsture at one time' it has been persistently inundated fbr r5 ycars, leading to the death of most trees and the irnminent death of the rest. Consecluently, many trees aPPeared suitable for Indiana bats (i.e., loose and peeling bark was present on the trunks). Within our plots,66 green ashes rvere suitafle (16 actual roosts and 50 apparently unused), as wcre zt silver maples, and two American elms' Because of the srnall sample of elms, we excluded them from further analyses. Were Indiana bats selective in choice of roost tree? We examined three potential modes of selectivity-by tree species, DBH, and the amount of exfoliating bark present on the trunk' Within our plots, 760/o of the 87 potential roost trees (excluding elm) were green ashes, whereas z\o/o were silver maples. Using these values as expected proportions, we determined that the probability of all z3 roosts being green ash as the result of chance was less than o.oo5 (X2 = 7'32; d'f'= t), and we concluded that the bats u'ere choosing green ash over silver maple. Were Indiana bats more likely to use a tree with a large amount of loose and peeling bark? following Gardner et al' (r99r), we ranked lz roost trees as high, nine as medium, and two as low. This distribution differed from random (X2 = 6'87; d'f'= '; p <o'or)' and superficially supports the hypothesis. Ilowever, within our sample of 5o suitable green ashes that were not used as roosts, the distribution of high, medium, and low was zr, 13, and 16, respectively' There was no significant difference (X'= 4.73; d.f.= ,; p >o.o9) between roost and non-roost samples, and hence, no evidence that the amount of loose bark was a factor in roost selection. Although the arnoult of loose bark was not important, the size of the trce was. The cliarreter o( roost trees (mean = 4o.g * r.z cm; n = 23) was significantly gfcarer (z = j.B5; p <o.oool) and significantly less variable (1.-,,. .,, = j.03; | = o.oo6) than the diameter of suitable, non-roost green ashes (nteiln - 3l4 t r 4 crn;'l = 5o) within our plots. Diameter of nonroost trees varred frorn ll to 70 cm, whereas trees used by flemales and young werc tightl,v groupcd betrvee n 3o and 52 cm' Individuals wcre not highly faithiul to a particular tree. Seventeen oI the 2l bats srvitc[c{ trecs at least ottce, yielding a total of 37 roost changes. Orr the day tltat a bat rvas first locatcd, rve did not know whether it had changccl roosts of ha1| rcntained in thc Sanle tlcc; therefore, we deleted the first clay l-r'ortr or-rr alalyses. 1-hirty scven roost changes over the rclraining ro7 6bscrvation clays inclicatecl t[at Indiana bats typically 105 Llt.utllcrl r()().\ts r'\,ct), r.9 tl.rt's. r\.lotl.rl nurrrlrcr ol roosI clrrrrrlic.s/trat rvas olll)' t\\,o, brrt chirngcs/[r.rt variccl fl<rnr zclo to six. Nulrrtrcr of-changcs rvas p<lsitivcly c<lrl'clatccl ',vitlt rttttn[rcr of'clays tlrat thc lrat was rnorritorcrl (r, -: o./o; // - zl; /) = o.ooo2; l:igrrrc r). Sorrrc charrqcs \vcrc to tl-ccs tlrirt thc ltut hitrl [)r,cviously roostcd in, and otltcrs wcrc t() trccs thirt rvc hacl rrot trackccl tlrrrt irrcliviclual to bcforc. Morlal nurntrcr o{'dil'lcrcnt trccs rrscrl lrv caclr lrat also was only two, trut inclivititrals ttsctl firrrr onc to lirur. Nrrrrrl>cr of-clil'lcrcrrt trccs uscd by cac6 [rat als. was corrclirrcd witlr rrtrr'trcr til'obscrvati.r riays (r, = o.zr; tl - 2.1; /.r = tl.<ltlctt; Iritrrrc r ). Avcragc distance fbl all j7 roost changcs was 24.2 t 1.6 (l_r7B) nt. Arnong arlults, clistancc rrrovccl rvas inclcPcrrrltrrt ol-rcproclrrctivc conclition (Xr = o.78; tl.l'. - tt p = <'t.(t7), and thcrc lvirs no dilli'rcnce bctwccn aclult lcrrralcs art<l juverrilcs (z = o.z9; p = o.l7); jista6cc rngvcd rvas incleperr, tlcnt of-ycar (Xr = .1.o,1; l.f. = ,- p = o.zz). Wcrc [ats sir'ply .tovlns ro tltc tlcarcst irvail.l[rlc trcc? I1'so, tlrc tlistr.ibrrtion of r'oost changc distanccs slitlLrlci [lc thc satttc as thc clistriLlution <tl (list.utccs [rctwccn.,rch roosr ancl its lreirrcst rrcighboLrr in that ycar ([]igLrrc z).'l'he trvo, lrowcver, diffcrecl significantly (Fislrer's Exact li:st: p = (.).ool). Although 67(yo of "vithirr-ycar, neirrcst-rrciglibclur distanccs wcrc (5o rn, <tnly t7tilt of roost change clistarnces rverc that .small. we cornparcd the ter.rclency to changc roosts ilnrong adult fcnrales using a -l x 2 contingcncv tablc; tlrc thrcc colurrrns rc[)rcscntccl rcproductivc condition, and the rows represented the number o[ occasions that a bat was observecl in the sarne roost as on the previous day versus a roost diffcrent frorn that uscd thc prcvious observation day. Roost-switching rvas not randonrlv clistribLrtcd anrong rcprocluctivc conditions (Xr = 6.tiz; d.f.= t; P <o.ol). For pregnant bats, we recorded r9 days on which a roost changc occurrcd and zo ciays on which it ditl not, indicating .t 49(ilt chance that a pregnant bat would not bc in the samc roost on consecutrve days. For lactating and post-lactating/non-reproductive fenrales, we recorded 7 changes versus zB non-changes (zoo/o), and 6 changes versus 13 non-changes (3zo/o), respectively. Juveniles switched roosts on 5 days and remaincd i'the samc tree on ro days (rl%), a frequency virtually identical to that of the post-lactating females (X2 = o.or; d.f.= ,; p >o.i). '[b investigate whether roost-switching was related to ambient tempera_ ture, we cxamincd the correlation between roost-cmcrgence counts and minimum arrbient temperature the day of tlie obscrvation (usually .ccurring ncar clau'n rvhcn lrats are selecting il ro()st) .rnti nr.rxrr-r]un) tcnrpcraturc on thc clay bcfore thc cor,rr.rt (rvhich might cletcrrninc rvhether bats re turn the next rrorning). 'fr:rnperaturcs werc obtaincd frorn a recording station at challotte, l4 knr ESE of'the study sitc. \,Ve restricted our analysls to trees at rvhich we nraclc )ro emergencc counts in any one year, after 14 i\4ay, but belore 17 July, which rvas the carliest clate that volant juveniles \\,erc cal)trlrcd. \Vc Pcrlot'ntcd five analyses on four roost trecs, but there was no consistcnt relationship lretwccn the environmcntal variables and loost population sizc. 'frvo trccs shorved significant correlations bctween population sizc irrrcl both tcrnPcraturcs, two roosts hacj no significant corr.clations, atlcl rltlc clispllyccl a significant rclati,rrrship rvitI rlinimunt tern, pcraturc, [rut not rnaxir.r..rr-rnr ('lablc I ). sECTroN rrr l?ocrs/irrc loo Roost-changing Behaviour r = 0./0 P = 0.0002 n = 23 bats a C]r c c U r = 0.71 p < 0.0001 n = 2l bats o @ F 4567 Observation Days Roost-chonging behoviour. (Top) The relotionship between number of roost chonges ond number of doys that the bot was locoted. (Bottom) The relotionship between number of different trees used ond number of doys thot the bot wos locoted. ro7 - o l o i Roost-change versus Nearest-neighbour Distances 20.0 17 .5 r 5.0 12.5 r 0.0 7.5 5.0 2.5 0.0 101-1s0 Dinance (m) FrcuRE z Roost-chonge distonces compored to nearest-neighbour distances within thot yeor. Neorest-neighbour distances were meosured only for trees to which lndiono bots were octuolly rodio-trocked in that yeor. We made emergence counts on 16r dates, observing l-Z roosts on any given night, for a total of 283 observations. On u3 occasions (4oolo) no bats left the roost; occasionally the lack of bats was due to cold or wet weather, but we attributed most instances to the bats' tendency to concentrate activity at some trees and rarely occupy others (see below). During zo of the remaining r7o nights (rzo/o), only a single bat emerged. Twelve obsqrvations of solitary e mcrgence were of seven bats carrying trans1'Arr-E r Ronk correlotions between number of bots leavtng o tree ond the moximum ombient temperoture recorded on the previous doy ond the minimum ambient temperoture recorded on the morning of the count. Populatlon Slze and Relatlve lmportance of Roosts l{oost/year n Rangc of dates Maximum tenrperature (oC) Minimum ternpcrature ("C) 24,t1992 r4 zElt992 l3 2Alt993 25 3Bl 1993 l(r 3Ft1993 27 30 May-15 July I 16 Juncl0 )uly r 19 May-10 July r I |une15 )uly r 14 func-I4 fuly r = -0.27; p = 0.34 r = 0.50; p = 0.08 r, = -0. l{t; p = 0.38 r, = 0.70; p = 0.003 r, = -0.56; p = 0.007 r\ 0.02.; p = Q.95 0.56; P = 0.047 -0. 15; p=Q.43 0.84; P=0.000 I -0.59; p = 0.004 Fisfrer's Exact Test p-0007 I cnung" fl Nearest sECTroN rtr Roosttng lo8 mitters, so we k.ew that at least one juvenile and three pregnant, one lactating, and two post-lactating {'ernales roosted alone on at least one night. The other l5o emergcrces involvecl z-45 bats, and most (g9%o) involved z-zr bats (Figure 3). -fio cieterrnine whcther roost populations rvere consistently larger after young became volant, rvc conrparcd emergencc counts made before and after r7 July, in r99z and 1993, rhe two years during which field work sp.rnnccl thc cntirc spring-suntrncr roosting pcriod (Figurc 4). To mini_ rnize cornplications resulting from migration, we also limited the data to counts n.rade betweerr 14 May and 15 August. There was no sienificant dif_ ferencc in population size before anci after ry July, in eithcr r99z lz - o.92; p = c36) or r99l (z = t.o6; p = o.z9). This suggested that Indiana bats occupied a greater number of trees within the wetland soon after the young became volant, or that sorne dispersed to other areas. Based on emergence counts, we estimated"the relative importance of roost trees by calculating the number of "bat-days" that each was occupied; o'e bat-day equals one bat using a tree for one day. Indiana bats did not use each tree to the same extent, but seemed to prefer two to four particular trees every year (Figure 5). In 1993, for exarrple, four roosts (ro, 2,a., 3n, and 3r), out of r8 that were occupied, accounted for 579 (8o%) of Tztbat-days; in contrast, eight other trees (rn, re, zF,3D> 3r,3L 3t-, and 3v)contributed a total of only 4 eo/o) bat-days. During the Emergence Counts (199'l -93) 22-31 32-41 42-51 Croup Size r;r(;Lrrrr, 3 Histogrom showinq t'requency of roost populotion sizes CAN f, o a 109 TITERATURE CITED Anonynrous. r995. Arc Antcrican Lrats losing grotrncl? Bats r3:,1. Barbour, Il.W. a'd W.ll. Davis. r9(r9. Birts .{' A'crica. University oI Kcntucky l)rcss, Lcxington, Ky. Brigham, R.M. r99r. Flcxibility i. fbragi.g a'd roosrirg bchaviour by thc big brown ltttt (L:.ptcsicus fusctts ). (.an. l. Zool. 6g:rt7-21. Calahan, Ir.v., III. r99j. Indiana lriit surnrne l habitat rcquiremcnts. M.s. 'l'hesis, Univcrsity of Missouri, Columbia, Mo. clawson, It. r9it7. I.cliana bats: clown lbr thc count. I]ndangered species Tcchnical Bulletin 22..9-11. Foster, R. r991. Roost sitc sclcction by the northcrn bat, A4yotis sep_ tcntrionulis. Bat Res. News -14:ro8. Gardncr,,.8.' J.D. Garner, and J.E. Ilofinann. r99r. Sumrner roost selection and roosring bchavior of Myotis sodalis (lndiana bat) in IIIin.is. U'publ. rcp. Illinois Narur.rrl Hisrorv Survcy, champaign, Il IJart' J.s. r97r. Rodents. In conrparativc phvsiclogy of thermoregulation, Vol. z. G.C. Whittow (cditor). Academic Press, Nov york, N.y., PP. r-r49. He nshaw, R.E,. and G.E. Folk, Jr. ry66. Relarion o[ thermoregulation to seasonally changing rnic'oclimate in trvo species of bats (Myotis Iucifugus and M. sodalis). physiol. Zool. 39:24-36. Humphrey, S.R., A.R. Richter, and j.B. Cope. ry77. Summer habitat and ecology o[ tl.re enda'gered Indiana bat, Myotis sodalis. I. Mammal. 58334-36. Kunz, T.H. 1982. Roosti'g ecology of bars. In Ecology o[ bats. T.H. Kunz (editor). Plenum Press, New york, N.y., pp. r_jr. Kurta, A. r995. Bark roost oIa male big brown bat (Eptesicus fuscus). Bat Res. News 35:63. Kurta, A. and R. Foster. 1995. The brown crecper (Aves: Certhii<1ae): a competitor of bark-roosting bats? Bat Rcs. News 36:6_7. xurta, A. and T.H. Kunz. 1987. Size of bats at birth and maternal investrnent in pregnancy. Symp. Zool. Soc. i.ondon 57:79_ro7. London, U.K. Kurta, A. and K.J. Williams. r994. Thcrmal aspccrs of Indiana bats roosr_ ing in trees. Bat Res. News 35:to4. Kurta, A., D. King, J.A. Teramino, I.M. Srriblcy, and K.J. Williams. r993a. Summer roosts of the endanger.ed Indiana bat (Myotis sodalis) on the northern edge of its range. Arner. Midl. Nat. iz9:r3z_3g. Kurta, A., I. Kath, E.L. Srnith, Il. Fostcr, N4. Orick, and R. Ross. r993b. A maternity roost of the e'dangered I'diana bat (Myotis sodalis) in ar unshaded, hollow, sycamore tree (platanus occidentali.s). Amer. Midl. Nat. r3o:4o5-o7. s ECTI o N ttt Roosting Lewis, S.E. 1993. Effect of climatic variarion on reproduction by pailid bats (Antrozous pallidus). Can. ). Zool. 7vt4z9-33. 1995. Roost fidelity of bats: a revicw. I. Mammal. 76:4gt*96. Lomolino, M.V. and R. Channeli. r995. Splendid isolation: patterns of geographic range collapse in endangered mammals. J. Mammal. 76335-47. Marshall, A.G. r982. Ecology of insccts parasitic on bats. In Ecology of bats. T.H. Kunz (editor). Plenum Press, Nerv york, N.y., pp.369_ 402. Merilzi, J. and K. Allander. 1995. Do grear tits (parus major) prefer ectoparasite-free roost sites? An cxperiment. Ethology 99:y_,6o. Richter, A.R., S.R. Humphrey, J.B. Cope, and V. Brack, Jr. 1993. Modified cave entrances: thermal effect on body mass and resulting decline of endangered Indiana bats (Myotis sodalis). Conservation Biology 7i4o7-r5. Studier, E.H. r98r. Energetic advantages of slight drops in body temperature in little brown bats, Myotis lucifugus. Comp. Biochem. physiol. 70Ai537-40. Thomson, C.E. t982. Myotis sodalis. Mammalian Species View publication statsView publication stats