J. Wildl. Manage. 62(2):1998 LONG-LEGGEDMYOTIS * Ormsbee and McComb 603

owl, appendix A, forest ecosystem management: ham, editors. Bats and forests symposium. British an ecological, economic, and social assessment. Columbia Ministry of Forests, Victoria, British Report of the Forest Ecosystem Management As- Columbia, . sessment Team, Portland, Oregon, USA. --, AND R. M. R. BARCLAY.1996. Roost-site se- Kalcounis, M.C.VALES, and D. J., R.M.AND F. L. BUNNELL.Brigham. 1985. Compari- 1998. Secondarylection and roosting use ecology of of forest-dwelling cavities sons of methods for estimating forest overstory bats in southern British Columbia. Canadian by tree roosting bigcover. Ministriesbrown of Environment bats. Journal and Forests Re- of WildlifeJournal of Zoology Management 74:1797-1805. 62(2): 603-611. port IWIFR-20. WARNER, R. M., AND N. J. CZAPLEWSKI. 1984. Long- Made availableVONHOF, courtesy M. J. 1996. of Roost-site the Wildlifepreferences of big Society: legged myotis.http://joomla.wildlife.org/ Mammalian Species 224:1-4. brown bats (Eptesicus fuscus) and silver-haired bats (Lasionycteris noctivagans) in the Pend Received 14 May 1997. D'Oreille Valley in southern British Columbia. Accepted 30 September 1997. Pages 62-80 in R. M. R. Barclay and R. M. Brig- Associate Editor: Lochmiller

SECONDARYUSE OF ASPEN CAVITIESBY TREE-ROOSTINGBIG BROWN BATS

MATINAC. KALCOUNIS,,2 Department of Biology,University of Regina, Regina,SK S4S 0A2, Canada R. MARKBRIGHAM, Department of Biology,University of Regina, Regina, SK S4S 0A2, Canada

Abstract: To further explore natural roost-site selection by temperate bats, we examined the use of tree roost sites by big brown bats (Eptesicus fuscus) in the West Block of Cypress Hills Provincial Park, Saskatchewan, an area where the number of human structures is limited. In this area, we found big brown bats roosted exclusively in cavities of trembling aspen trees ( tremuloides), despite the availability of cavities in snags of trees. Most cavities had been excavated and previously used by yellow-bellied sapsuckers (Sphyr- apicus varius). The use of by sapsuckers relates to the relatively soft and susceptibility to heart rot of these trees, which provide ideal conditions for nesting: decayed heartwood with a firm sapwood shell. Orientation of cavity entrances was close to due south. The width of bats and the width of cavity entrances differed, suggesting that bats are not using roosts for protection from predators or exclusion of competitors. Bats showed fidelity to a particular group of roost trees because, despite roost switching, bats reunited in subsequent roost sites. During the day, temperatures in aspen cavities were approximately 50C cooler than in cavities of conifer snags. Microclimate differences, including temperature, may be why aspen cavities are selected over available cavities in conifer snags. All of the randomly selected cavities in aspen that we searched showed evidence of use by bats, which suggests roost sites for big brown bats in southwestern Saskatchewan may be a limiting resource. JOURNALOF WILDLIFEMANAGEMENT 62(2):603-611

Key words: bats, Eptesicus fuscus, habitat complexity, maternity colony, roost, switching, temperature, tree- cavity, trembling aspen.

In comparison to use of nest sites by birds, man structure than in a natural site such as a relatively little is known about use of roost sites tree cavity. As a result, factors that influence by bats, especially use of natural roosts such as selection of roost sites by bats under natural tree cavities. The majority of ecological and be- conditions where human structures are not havioral research on temperate insectivorous abundant are poorly understood (for review see bats is from (Chiroptera: Vespertilionidae) spe- Brigham and Barclay 1996). cies in human because it is roosting structures, Selection of specific roost sites by breeding easier to find and access bats in a hu- roosting female bats has consequences to survival and reproductive success (Entwistle et al. 1997), Present address: and Evolution 1 Ecology Group, and there is mounting evidence that bats choose Department of Zoology, University of Western Ontar- io, London, ON N6A 5B7, Canada. roost sites for reasons of microclimate, protec- 2 E-mail: [email protected] tion from predators, avoidance of parasites, and Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

604 BIGBROWN BATS * Kalcounisand Brigham J. Wildl. Manage.62(2):1998

social interaction (this study; also see reviews in terms of flight, but predation risk may differ Lewis 1995, 1996). Roost type (building, tree between simple and complex habitats. When in- cavity, rock crevice, cave) may be unimportant dividuals emerge from a maternity colony, many to bats, provided roost type satisfies other re- individuals may exit in a relatively short period, quirements such as characteristics of the roost which makes the event conspicuous (Speakman or the habitat surrounding the roost. Physical et al. 1992). Emergence is also highly correlated characteristics of a roost can influence the pro- with sunset and is therefore predictable in time tection afforded by the roost and the microcli- (Kalcounis and Brigham 1994). Thus, if bats mate within the roost, whereas amount of hab- perceive a predation risk, they may select roosts itat surrounding the roost can influence flight surrounded by complex habitat to conceal pre- performance capabilities of both juvenile and dictable and conspicuous emergence events. adult females and can influence predation risk The purpose of this paper is to describe the to individuals in the roost. suite of physical characteristics, including roost Nonetheless, the influence of roost microcli- microclimate, that define big brown bat roost mate on use of roost sites is not clear. For ex- trees in Cypress Hills Provincial Park, Saskatch- ample, the leaf-nosed bat (Hipposideros speo- ewan, where this species exclusively roosted in ris) exhibits high fidelity to dark roost sites with- the cavities of trembling aspen (Kalcounis in caves (Usman 1988), whereas the bat 1995). We also test the hypotheses that (1) to (Myotis sodalis) uses tree-roost sites in unshad- avoid predators and competitors, big brown bats ed locations (Kurta et al. 1993a,b). Similarity in choose roost sites with cavity entrances not internal temperatures between different roost much larger than themselves; (2) the complexity types suggests that selection of roost sites may of the habitat directly in front of the roost-cavity involve locating a specific range of thermal con- opening is lower than that in other directions; ditions. For example, in South Africa, bats from and (3) the thermal microclimate of cavities the family Rhinolophidae select roosts in hollow used by big brown bats (in aspen) will differ baobab trees (Adansonia digitate), Vespertilion- from cavities not used by bats (in conifer snags). idae select roosts in hollow trees (Col- mopane STUDYAREA ophospermum mopane), and Molossidae select roosts in building attics because all roosts offer This study was conducted during the sum- similar temperature regimes (Fenton and Rau- mers of 1993 and 1994, within the West Block tenbach 1986). Further, the orange horseshoe of Cypress Hills Provincial Park (49034'N, bat (Rhinonycteris aurantius) in India occupies 109053'W), approximately 60 km southwest of a narrow range of available humidity in its cave Maple Creek, Saskatchewan. The Cypress Hills roost sites (Churchill 1991). run about 50 km in an east-west direction and The ontogeny of juvenile flight (Adams 1997) form the highest elevated plateau in the Cana- and the decreased maneuverability of pregnant dian prairies (Sauchyn 1993). Lodgepole pine bats (Norberg and Rayner 1987) may make less (Pinus contorta) forest occupies dry sites above complex habitat around entrances to roost trees 1,300 m, and white spruce () forest a preferred characteristic. Juvenile bats in tree occurs in cool, moist areas near wetlands and hollows, in contrast to bats roosting in buildings, on north-facing slopes. The understory of are generally unable to perform practice flights spruce forest is relatively diverse and includes before venturing outside for the first time. several species with distributions disjunct When juvenile little brown bats (Myotis lucifu- from mountain populations. Trembling aspen is gus) take their first flights, they remain close to found growing with white spruce near streams the roost, avoid acoustically complex habitat, on south-facing slopes. and do not insects In- pursue (Buchler 1980). METHODS creases in body mass also affect habitat use by Data little brown bats whereby heavier bats forage in Collection less complex habitats (Kalcounis and Brigham Locating Roost Trees.-We located roost 1995). Pregnant female bats may therefore se- trees by tracking radiocollared bats or by lect roost sites where the number of objects to searching randomly selected cavities for the ev- detect and avoid in front of the cavity entrance idence of use by bats. For the purpose of ra- is low. diotelemetry, we initially used mist nets to cap- Not only is habitat complexity relevant in ture bats. Upon capture, we affixed 0.8-g trans- Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

J. Wildl. Manage.62(2):1998 BIG BROWNBATS * Kalcounisand Brigham 605

CAVITYOPENING DIRECTION mitters (BD-2T, Holohil Systems, Woodlawn, Ontario, Canada) via Skinbond cement. Individ- (A) uals carrying radiotransmitters were tracked to their roosts the morning following capture via Merlin 12 (Custom Electronics, Urbana, Illi- nois, USA) portable telemetry receivers and 450 hand-held Yagi antennae. When we located a roost tree, it was observed at dusk to ensure it was used by bats and to determine colony size. / We defined a random site as an aspen tree A with a cavity opening of >4.0 cm because all RoostTree big brown bats roosted in aspen trees and re- quired an opening greater than their body width to enter the cavity. We selected random sites via randomly generated distance and di- rection tables. To access cavities (to check ran- dom cavities and take measurements on roost cavities), we climbed trees with tie-on tree- climbing steps and a standard climbing harness (B)FRONT and ropes. We checked randomly selected cav- ities for evidence of use by bats by scraping the 14.59 12.51 bottom and sides of the cavity with a long-han- + dled spoon to retrieve bat droppings. Characteristics of Roost Sites.-Characteris- tics recorded for each roost tree included tree condition (live or dead), height and diameter at breast height (dbh), percent cover by neighbor- ing trees (number of branches from neighbor- 18.26 9.23 ing trees that fell within 1 m of the tree), and + number of cavity openings. We also recorded the presence or absence of fruiting bodies of heart-rot fungus (hoof-shaped conks) on tree trunks because fruiting bodies are external in- dicators of false tinder rot (Fomes igniarius; Pe- Fig. 1. (A) 12-m-radiuscircular plot divided into quadrants for terson and Peterson measuring and calculatingHabitat Complexity Indices. (B) 1992). HabitatUse Indices(mean + 1 SD) forquadrant in frontof the Characteristics recorded for each cavity in- roost openingand the 3 otherquadrants. cluded height, entrance orientation, width (di- mension perpendicular to trunk axis) and length (dimension parallel to trunk axis) of the cavity of the cavity entrance, we measured width of entrance, light intensity as a proportion of am- bats from elbow-to-elbow because this distance bient light intensity (Model 840006 light meter; is the widest part of a bat, and thus the dimen- Sper Scientific, Tempe, Arizona, USA), percent sion most likely to constrain selection of tree relative humidity as a proportion of ambient cavities. Using calipers, we made 3 measure- percent relative humidity (Model 880 humidity ments of width for each individual. To maximize meter; General Eastern, Woburn, Maryland, sample size, we measured individuals from a USA) at solar noon, and percent of cavity open- colony of over 100 big brown bats roosting in ing covered by its own tree foliage and by the attic of Elm Street School in nearby Med- neighbor tree foliage (determined by estimating icine Hat, Alberta (50002'N, 111040'W). how many branches, foliage, or both fell within Habitat Complexity.-To quantify habitat 1 mn of the entrance). We took all measures of complexity around roost trees, we determined light intensity and percent relative humidity be- distance (from the roost tree) and estimated tween 1 and 14 August 1994 to minimize vari- height (as a percent of canopy height) for all ation due to canopy development. trees within a 12-m-radius circular plot cen- To generate predictions regarding the width tered at the roost tree (Fig. 1A). We established Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

606 BIGBROWN BATS * Kalcounisand Brigham J. Wildl. Manage.62(2):1998

the position of every tree within the circular partial profiles (>12 hr/24-hr period) were ob- plot by determining the direction and distance tained for 4 days. of each tree from the roost tree. A Habitat Statistical Complexity Index (HCI) was calculated using Analysis the following equation: All data on tree and cavity characteristics are presented as mean ? 1 standard deviation. = Z HCI [log(h). 1/log(d)], Comparisons of body and cavity width and HCI indices between front and back habitat were where h is the height of the tree (expressed as made with parametric t-tests. We used an anal- percent of canopy height), and d is the distance ysis of variance (ANOVA) to test for HCI equa- (m) of the tree from the roost tree. This index tion variables (number, dis- reflects the of the habitat individually height, average complexity tance of trees). Entrance orientation data were around the roost tree such that a higher index corrected to true north and analyzed with a indicates more close, tall trees. relatively test (Zar 1984) to determine whether We divided the circular into Rayleigh plot quadrants entrance orientations were randomly distribut- such that 1 quadrant was bisected by the en- ed around a circle. To trance orientation We calculated an compare temperature (Fig. 1A). profiles between aspen and conifer snag cavi- HCI for the (hereafter, cavity-opening quadrant ties, we divided each sampling day into 4 time front habitat) and for the 3 remaining quadrants blocks reflecting night (0000-0559), morning back We scaled the back (hereafter, habitat). (0600-1159), afternoon (1200-1759), and eve- HCI for the differences in area between the ning (1800-2359). Temperatures recorded in front and back habitat the HCI of by dividing each time period were pooled over the 10 sam- back habitat 3. The HCI index con- by may pling days. We performed ANOVA and, where found the effects of the variables in the equa- appropriate, post hoc comparisons of means tion and increase the likelihood of thereby Type (Tukey HSD) on the temperature data. Tem- II errors. Therefore, we also each analyzed perature data are reported as mean ? 1 stan- characteristic (i.e., number of trees, separately dard error. We used a = 0.05 as a rejection height and distance from the roost tree). criterion for all statistical tests. Thermal Microclimate.--We used tempera- ture-sensitive radiotransmitters (PD-2T and RESULTS Holohil BD-2T, Systems, Woodlawn, Ontario, We marked 11 adult female big brown bats to ambient with Canada) compare temperature with radiotags, and we found 27 roost trees (16 of the used temperature profiles cavity types by from radiotracking, 11 from random cavity bats (aspen) and the available cavities not used searches). Of the 16 roost trees found from ra- bats of We defined conifer by (snags ). diotracking, 10 were found in 1993 and 6 were as dead white or snags spruce lodgepole pine found in 1994. We watched the emergence of trees with cavities and tree broken off. We tops colonies from roosts on 19 nights. Colony size transmitters in cavities placed unoccupied by ranged from 2 to 43 individuals with a mean (? climbing the trees and suspending the trans- SD) of 20.7 ? 13.1 individuals. The colonies mitter in the via monofilament cavity fishing were maternity colonies composed of juveniles line. We selected 2 from randomly cavities the and pregnant, lactating, and postlactating adult group of roost trees, 2 cavities in aspen trees, females. Individuals did not remain in a single and 2 cavities from conifer snags. The trans- roost tree for the entire season, and at least 4 mitter measuring ambient temperature was sus- radiocollared bats stayed in roost trees for no pended, in midair, between 2 trees near the re- more than 1 night. Two of the 10 roosts found ceiver (see below). in 1993 were used again by big brown bats in To measure temperatures, we used an auto- 1994. At least 1 radiocollared bat returned to a mated telemetry receiver (Lotek SRX 400 te- previously used roost site within 1 year. On 4 lemetry receiver with Event Log Version 2.62 occasions, radiocollared bats (which were W18 logging software; Lotek Engineering, Au- known to have roosted together) reunited in rora, Ontario, Canada). We programmed the re- subsequent roost sites, suggesting that bats ceiver to record the temperature of each trans- show fidelity to a particular group of roost trees. mitter hourly from 13-26 July 1994. Complete In 2 roosts with >1 cavity opening, red squirrels 24-hr profiles were obtained for 6 days, while (Tamiasciurus hudsonicus) were seen entering Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

J. Wildl. Manage. 62(2):1998 BIG BROWNBATS * Kalcounis and Brigham 607

Table 1. Characteristicsof roosttrees and cavities used by big brownbats in Cypress HillsProvincial Park during the summers of 1993 and 1994.

Characteristic n t SD Range Tree Height (m) 27 25.6 7.7 13.6-51.8 Dbh (cm) 27 35.8 7.2 22.8-57.1 Percent cover 27 33.5 23.4 0-75.0 Number of entrances 27 2.0 1.3 1.0-5.0 Cavity Height (m) 26 8.4 2.9 4.3-14.6 Relativehumidity 24 1.3 0.3 0.6-1.9 Light intensity 24 0.8 0.3 0.2-1.0 Percent cover by self 26 11.4 18.9 0-50.0 Percent cover by neighbours 26 10.0 23.5 0-75.0 Length (mm) 21 255.4 606.5 40.0-2,700.0 Width (mm) 21 45.7 11.6 25.8-70.0

the cavity from 1 opening while bats emerged HabitatComplexity from another. There was no difference between the HCI in Characteristicsof RoostSites the front and back of roost trees (F1,52 = 1.51, P > 0.05; Fig. Similarly, there was no dif- All roost trees were 24 were 1B). trembling aspen: ference between the mean number of trees, the live 3 were dead All live trees, and trees. roost or the distance from the roost tree in trees had of heart rot in the height, symptoms fungal the area in front and back of roost trees (P > form of conks. Most (at least 73%) cavities were 0.05; Table 2). originally excavated by sapsuckers or other woodpeckers. On several occasions, yellow-bel- ThermalMicroclimate lied sapsuckers were seen excavating nest cavi- Over the mean tem- ties in aspen trees that had entrance dimensions 10-day sampling period, consistent with those of the roost cavities. Some peratures differed during the time blocks of P cavities (at least 23%) originated as branch scars night (F2,189 = 21.22, < 0.001), morning = P afternoon = or splits in the trunks of trees. (F2,189 33.18, < 0.001), (F2,230 P and = Tree and cavity characteristics of roosts are 4.80, < 0.01), evening (F2,192 10.26, P For given in Table 1. We were unable to reach the < 0.001; Fig. 2). the post hoc mean com- cavity of 1 live roost tree or 2 of the dead roost parisons, an experimentwise error rate (oa) of trees, and therefore could not measure their 0.05 was used, and the critical q-value for all cavity characteristics. The mean orientation of comparisons was 3.314 (Zar 1984). Aspen cavi- cavities (169.39 ? 16.230) was close to due ties were warmer than ambient temperature but south, and cavity orientations were not uniform- cooler than conifer snag cavities at night (ql8s,:3 ly distributed (Z26 = 2.97, P < 0.001). The = 5.33, P < 0.001). In the morning, aspen cav- mean body width of big brown bats (39.82 ities were cooler than ambient temperature 0.66 n = was smaller than the width?_ mm, 21) (qs89,:3 = 9.30, P < 0.001) and conifer snag cav- of cavity entrances (45.37 ? 2.09 mm; F1,45 =- ities (q189,3 = 9.51, P < 0.001), which did not 5.32, P < 0.03). differ from each another (q189,3 = 0.18, P >

Table 2. Habitatcharacteristics in frontand back of roosttrees used by big brownbats in Cypress HillsProvincial Park during the summersof 1993 and 1994.

Front Back Characteristic . SD I SD df F P Number of trees 7.74 5.52 7.09 3.77 1, 52 0.20 0.62 Height of trees (%) 88.76 14.31 87.33 9.60 1, 51 0.18 0.67 Distance of trees from roost (m) 8.98 1.81 8.37 1.02 1, 52 2.35 0.13 Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

608 BIG BROWNBATS - Kalcounis and Brigham J. Wildl. Manage. 62(2):1998

O AMBIENT U ASPEN SNAG duces extensive decomposition of heartwood as a 27 while sparing sapwood, which remains tough, living outer shell (Kilham 1971). In con- 25 - trast, white spruce and lodgepole pine do not Du I-- have the same decay characteristics as aspen w 22 and Peterson 1992). In conifers, a _ (Peterson heartwood and both more I-- sapwood decay rap- >- 20 idly, which precludes the formation of a solid * I et al. 0 outer shell of sapwood (McClelland 1979). z 17 U In the Cypress Hills, suitability and use of as- I i pens for nesting and subsequent roosting sites 15I I seem determined both the availability of as- NIGHT MORNING AFTERNOON EVENING by as the dominant hardwood and by its over 10 of pen only Fig. 2. Mean (?+SE) temperatures days sampling characteristics. in the 4 time blocks reflectingnight (0000-0559), morning decay (0600-1159), afternoon (1200-1759), and evening (1800- Cavity orientation can enhance microclimate Means an asteriskare differ- 2359). separated by significantly effects, and many birds derive thermoregulatory ent (P < 0.05; Tukey'stest). benefits from specific orientations of nest sites by avoiding direct wind exposure (Austin 1976, Williams 1993, 1995) and sun (Inouye 0.05). In the afternoon, aspen cavities were Haggerty et al. 1981, Korol and Hutto 1984), or facilitat- cooler than ambient temperature (q230,3 = 4.07, direct to sun 1976, Smith P < 0.05) but did not differ from conifer snag ing exposure (Inouye et al. 1987). Sapsuckers and other woodpeckers cavities (q230,3 = 1.06, P > 0.05). Both aspen tend to excavate nests on the south-facing side (q192,3 = 6.03, P < 0.001) and conifer snag of trees (Reller 1972, Crockett and Hadow (q189,3 = 5.25, P < 0.001) cavities were warmer 1976), which is an than ambient temperature in the evening. The 1975, Inouye probably adap- tive to the difference between aspen cavities and conifer response by sapsuckers prevailing east-west wind and cool tem- snag cavities was greatest in the morning. relatively spring peratures in the Cypress Hills. The predomi- DISCUSSION nant southerly orientation of cavity entrances Characteristicsof Roost Sites suggests that cavity orientation confers some benefit to the sapsuckers that excavate the cav- Although most commonly found roosting in ities and to the big brown bats that subsequent- buildings (Brigham and Fenton 1986), big ly roost in the cavities. brown bats have been found roosting in natural The width of bats was significantly smaller hollows of beech (Fagus grandifolia; Kurta than the width of roost entrances. Throughout 1980), (Pinus ponderosa pine ponderosa; Brig- their excavate nest cavities ham 1991), and western redcedar range, sapsuckers (Thuja pli- with circular entrances consistent di- cata; Vonhof and 1996), as well as rock having Barclay mensions. In British Columbia (Erskine and crevices (Brigham 1991). In the Cypress Hills, McClaren 1972) and 2 sites in Colorado (In- big brown bats are secondary cavity roosters 1976, Winternitz and Cahn 1983), the that use both live and dead aspen trees, despite ouye mean diameters of sapsucker nest entrances the availability of cavities in conifer snags. Most were between 3.5 and 4.1 cm. In our study, the of the aspen cavities had been excavated and mean diameter of entrances was 4.5 cm. The used by sapsuckers as nest sites. As secondary between the mean diameter of en- cavity users, bats choose cavities from among discrepancy trances in this and those already excavated and are therefore con- study previous investigations can be attributed to of around strained by the preferences of the primary cav- decay sapwood the entrance hole in the older cavities ity excavators for nesting sites and by the decay sapsucker characteristics of the tree and cavity. The use of that were used secondarily by bats. The differ- ence between the width of bats and the width aspens by primary cavity excavators such as sap- suckers is related to the aspen's relatively soft of roost entrances, along with the observation wood and susceptibility to heart rot. False tin- of red squirrels entering cavities, suggests bats der rot is a major cause of decay in aspens (Bas- do not select cavities that preclude entrance by ham 1958, Peterson and Peterson 1992) that in- predators, competitors, or both. Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

* J. Wildl. Manage. 62(2):1998 BIG BROWNBATS Kalcounis and Brigham 609

HabitatComplexity snag cavities may render conifer snag cavities unsuitable as roost sites for brown bats. Relative to other temperate Vespertilionidae, big big brown bats are large and have a low echo- MANAGEMENTIMPLICATIONS location call frequency (Norberg and Rayner 1987). Because of these characteristics, big From a conservation perspective, roosts are brown bats should be restricted to flying in the most important resource for many bat spe- more open habitat (see Norberg and Rayner cies (reviewed in Barclay and Brigham 1996). 1987). In our study, however, the complexity of The pervasive use of human structures by big habitat surrounding the roost site was not a fac- brown bats in (Williams and tor influencing use of a roost site. One problem Brittingham 1997) suggests that either suitable associated with field studies that attempt to cor- natural sites no longer exist or that buildings are relate habitat selection to habitat complexity in preferred roost sites. In the Okanagan Valley of bats is the difficulty in determining how com- British Columbia, big brown bats roosted in plex a habitat must be before a species is no trees despite the presence of human structures, longer maneuverable enough to fly there. Al- which supports the hypothesis that use of build- though all roost trees in our study were in the ings may be the result of a loss of natural sites middle of mixed aspen-spruce forests, adult (Brigham 1991). Although roost sites are not and juvenile big brown bats had flight capabil- limiting for building-roosting big brown bats in ities that did not preclude them from using the Ontario, individuals experimentally excluded surrounding habitat. from roosts often attempted to find alternative entrances to their original roost site and had ThermalMicroclimate decreased reproductive success in alternate sites and Fenton Different decay dynamics should influence (Brigham 1986). If we assume bats select roosts that maximize temperature regimes within cavities of conifer their then mature snags and aspen trees. A study by Burnett and reproductive success, aspen trees with old nest cavities are critical August (1981) on the energy budgets of a ma- sapsucker for the survival and fitness of the brown bat ternity colony of building-roosting little brown big in Hills. Given the wide- bats offers insight as to why the 5?C daytime population Cypress distribution of brown bats and difference between aspen and conifer snag cav- spread big aspen this conclusion have ities may be an important criterion in the selec- trees, may general appli- because all of the ran- tion of roost sites by big brown bats. At noon, cability. Furthermore, selected cavities in our unoccupied little brown roosts are 30'C as com- domly aspen study showed evidence of use roost sites for pared with 350C for occupied roosts. The ther- by bats, moneutral zone for little brown bats is between big brown bats may be limited. the 32.5 and 37.5'C, which puts the temperature of Where bats are secondary cavity users, of sites exca- occupied roosts within the thermoneutral zone. selection used by primary cavity However by occupying a roost, little brown bats vators probably influences the population num- bers and of increase the temperature by 5?C. Hence, a roost community composition bats, just much warmer than 30TC would be unsuitable as the local distributions and abundances of sec- for little brown bats because, once occupied, ondary cavity-nesting birds are enhanced by the the temperature within the roost would ap- availability of old nest cavities (Daily 1993). In proach or exceed the upper limit of the ther- western Canada, little is known about biodiver- moneutral zone for bats. In a tree cavity, the sity relations of aspen mixed-wood forests be- effect of roost occupation by a maternity colony cause most studies have concentrated on coni- of big brown bats would be greater because the fer-dominated canopies (Stelfox 1995). Howev- volume of a tree cavity is much smaller than er, it is clear that primary cavity excavators such that of an attic in a building. In addition, we as the yellow-bellied sapsucker prefer old, as- found relative humidity of roost cavities to be pen-mixed-wood stands (Schieck and Nietfeld 1.3% higher than ambient, and any effect of 1995). Thus, harvesting practices which pro- roost occupation on roost temperature would be mote the maintenance of large, old stands of enhanced by a high relative humidity inside the mixed-wood forests must be implemented to roost. We suggest that the difference in morn- provide an adequate density of aspen for tree- ing temperature between aspen and conifer roosting big brown bats. Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

610 BIG BROWNBATS * Kalcounis and Brigham J. Wildl. Manage. 62(2):1998

ACKNOWLEDGMENTS site selection by Williamson and red-naped sap- suckers. Condor 77:365-368. Invaluable assistance in the field was provid- DAILY, G. C. 1993. Heartwood decay and vertical ed by D. J. Bender, D. L. Gummer, and K. R. distribution of red-naped sapsucker nest cavities. Hecker. We use of the Wilson Bulletin 105:674-679. appreciate Department A. P. A. AND R. SPEAKMAN. of of Field Station ENTWISTLE, C., RACEY, J. Biology, University Regina 1997. Roost selection by the brown long-eared the of the West and acknowledge hospitality bat, Plecotus auritus. Journal of Applied Ecology Block of Cypress Hills and Fort Walsh park ad- 34:399-408. ministrators and employees. The suggestions of ERSKINE, A. J. 1960. Further notes on interspecific C. D. A. D. competition among hole-nesting ducks. Canadian Ankney, G. Baldassarre, W. Inouye, Field-Naturalist 74:161-162. M. R. E. P. MacDonald, J. O'Connor, J. Thomp- -, AND W. D. MCCLAREN. 1972. Sapsucker son, M. G. Topping, and M. J. Vonhof improved nest holes and their use by other species. Cana- earlier versions of this manuscript. This re- dian Field-Naturalist 86:357-361. M. AND L. RAUTENBACH. 1986. A search was funded by grants from the Regina FENTON, B., I. Natural and the Theodore Roo- comparison of the roosting and foraging behav- History Society iour of three of African insectivorous bats sevelt Memorial Fund Museum of species (American (Rhinolophidae, Vespertilionidae, and Molossi- Natural History) to MCK and a Natural Science dae). Canadian Journal of Zoology 64:2860-2867. and Engineering Research Council of Canada HAGGERTY,T. M. 1995. Nest-site selection, nest de- and nest-entrance orientation in Bachman's (NSERC) research grant and University of Re- sign sparrow. Southwestern Naturalist 40:62-67. gina President's NSERC grant to RMB. MCK INOUYE,D. W. 1976. Non-random orientation of en- was supported by an NSERC postgraduate trance holes to woodpecker nests in aspen trees. scholarship while writing this manuscript. Condor 78:101-102. INOUYE, R. S., N. J. HUNTLY, AND D. W INOUYE. LITERATURECITED 1981. Non-random orientation of gila woodpeck- er nest entrances in saguaro cacti. Condor 83:88- ADAMS, R. A. 1997. Onset of volancy and foraging 89. patterns of juvenile little brown bats, Myotis lu- KALCOUNIS, M. C. 1995. Natural roost site selection cifugus. Journal of Mammalogy 78:239-246. by big brown bats (Eptesicusfuscus). Thesis, Uni- AUSTIN, G. T. 1976. Behavioral adaptations of the versity of Regina, Regina, Saskatchewan, Canada. verdin to the desert. Auk 93:245-262. --, AND R. M. BRIGHAM.1994. Impact of pre- BARCLAY,R. M. R, AND R. M. BRIGHAM,editors. dation risk on emergence by little brown bats, 1996. Bats and forests symposium. British Co- Myotis lucifugus (Chiroptera: Verspertilionidae), lombia Ministry of Forests, Victoria, British Co- from a maternity colony. Ethology 98:201-209. lumbia, Canada. -5, AND R. M. BRIGHAM. 1995. Intraspecific BASHAM,J. T. 1958. Decay of trembling aspen. Ca- variation in wing loading affects habitat use by nadian Journal of Botany 36:491-505. little brown bats (Myotis lucifugus). Canadian BRIGHAM,R. M. 1991. Flexibility in foraging and Journal of Zoology 73:89-95. roosting behaviour by the big brown bat (Eptes- KILHAM,L. 1971. Reproductive behavior of yellow- icusfuscus). Canadian Journal of Zoology 69:117- bellied sapsuckers 1. Preference for nesting in 121. Fomes-infected aspens and nest hole interrela- AND R. M. R. BARCLAY.1996. Bats and for- tions with raccoons, and other an- ,, flying squirrels, ests. Pages XI-XIV in R. M. R. Barclay and R. imals. Wilson Bulletin 83:159-171. M. Brigham, editors. Bats and forests symposium. KOROL,J. J., AND R. L. HUTTO. 1984. Factors af- British Colombia Ministry of Forests, Victoria, fecting nest site location in gila woodpeckers. British Columbia, Canada. Condor 86:73-78. AND M. B. FENTON. 1986. The influence of KURTA, A. 1980. The bats of southern lower Michi- , roost closure on the roosting and foraging behav- gan. Thesis, Michigan State University, East Lan- iour of Eptesicusfuscus (Chiroptera: Vespertilion- sing, Michigan, USA. idae). Canadian Journal of Zoology 64:1128- , J. KATH, E. L. SMITH, R. FOSTER, M. W. 1133. ORICK,AND R. ROSS. 1993a. A maternity roost BUCHLER,E. R. 1980. Development of flight, for- of the endangered Indiana bat (Myotis sodalis) in aging and echolocation in Myotis lucifugus. Be- an unshaded, hollow, sycamore tree (Platanus oc- havioral Ecology and Sociobiology 6:211-218. cidentalis). American Midland Naturalist 130: BURNETT, C. D., AND P. V. AUGUST. 1981. Time and 405-407. energy budgets for day roosting in a maternity , D. KING, J. A. TERAMINO, J. M. STRIBLEY, colony of Myotis lucifugus. Journal of Mammal- AND K. J. WILLIAMS. 1993b. Summer roosts of ogy 62:758-766. the endangered Indiana bat (Myotis sodalis) on CHURCHILL,S. K. 1991. Distribution, abundance the northern edge of its range. American Midland and roost selection of the orange horseshoe-bat, Naturalist 129:132-138. Rhinonycteris aurantius, a tropical cave-dweller. LEWIS, S. E. 1995. Roost fidelity of bats: a review. Wildlife Research 18:343-353. Journal of Mammalogy 76:481-496. CROCKETT,A. B., AND H. H. HADOW. 1975. Nest - . 1996. Low roost-site fidelity in pallid bats: Kalcounis, M.C. and R.M. Brigham. 1998. Secondary use of aspen cavities by tree roosting big brown bats. Journal of Wildlife Management 62(2): 603-611. Made available courtesy of the Wildlife Society: http://joomla.wildlife.org/

J. Wildl. Manage.62(2):1998 BIGBROWN BATS * Kalcounisand Brigham 611

associated factors and effect on group stability. SPEAKMAN, J. R., D. J. BULLOCK, L. A. EALES, AND Behavioral Ecology and Sociobiology 39:335-344. P. A. RACEY.1992. A problem defining temporal MCCLELLAND,B. R., S. S. FRISSELL,W. C. FISCHER, pattern in animal behaviour: clustering in the AND C. H. HALVORSON.1979. Habitat manage- emergence behaviour of bats from maternity ment for hole-nesting birds in forests of western roosts. Animal Behaviour 43:491-500. larch and Douglas-fir. Journal of Forestry 77:480- STELFOX,J. B. 1995. Introduction. Pages 1-12 in J. 483. B. Stelfox, editor. Relationships between stand NORBERG, U. M., AND J. M. V. RAYNER.1987. Eco- age, stand structure, and biodiversity in aspen logical morphology and flight in bats (Mammalia; mixed wood forests in Alberta. Alberta Environ- Chiroptera): wing adaptations, flight perfor- mental Centre, Vegreville, Edmonton, and Ca- mance, foraging strategy and echolocation. Pro- nadian Forest Service, Edmonton, Alberta, Can- ceedings of the Royal Society of London Series ada. B 316:335-427. USMAN, K. 1988. Role of light and temperature in PETERSON,E. B., AND N. M. PETERSON.1992. Ecol- the roosting ecology of tropical microchiropteran ogy, management, and use of aspen and balsam bats. Proceedings of the Indian Academy of Sci- poplar in the prairie provinces, Canada. Forestry ence 97:551-59. Canada, Northern Forestry Centre, Edmonton, VONHOF, M. J., AND R. M. R. BARCLAY. 1996. Roost- Alberta, Canada. site selection and roosting ecology of forest- RELLER,A. W. 1972. Aspects of behavioral ecology dwelling bats in southern British Columbia. Ca- of red-headed and red-bellied woodpeckers. nadian Journal of Zoology 74:1797-1805. American Midland Naturalist 88:270-290. WILLIAMS, J. B. 1993. Nest orientation of orange- breasted sunbirds in South Africa. SAUCHYN,D. J. 1993. Quaternary and late Tertiary Ostrich 64:40- 44. landscape evolution in the western Cypress Hills. L. AND M. C. 1997. Pages 46-58 in D. J. Sauchyn, editor. Quaternary WILLIAMS, M., BRITTINGHAM. Selection of roosts brown bats. and late Tertiary landscapes of southwestern Sas- maternity by big of Wildlife 61:359-368. katchewan and adjacent areas. Canadian Plains Journal Management B. AND H. CAHN. 1983. Nestholes Research Centre, Regina, Saskatchewan, Canada. WINTERNITZ, L., in live and dead 102-106 in W SCHIECK,J., AND M. NIETFELD. 1995. Bird species aspen. Pages J. Davis, G. A. Goodwin, and R. A. tech- richness and abundance in relation to stand age Ockenfels, nical coordinators. of the hab- and structure in aspen mixedwood forests in Al- Proceedings snag itat U.S. Forest Service berta. 115-157 in B. Stelfox, editor. Re- management symposium. Pages J. General Technical RM-99. lationships between stand age, stand structure, Report and in mixed wood forests in ZAR, J. H. 1984. Biostatistical analysis. Second edi- biodiversity aspen tion. Alberta. Alberta Environmental Prentice-Hall, Upper Saddle River, New Centre, Vegre- USA. ville, Edmonton, and Canadian Forest Service, Jersey, Edmonton, Alberta, Canada. Received 6 August 1996. SMITH, D. G., A. DEVINE, AND R. GILBERT. 1987. Accepted 22 September 1997. Screech owl roost site selection. Birding 19:6-15. Associate Editor: O'Connor