• Chiroptera;
  • habitat fragmentation;
  • roost humidity;
  • roost selection;
  • roost temperature


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Cavity quality is important for the productivity and survival of many species of tree-dwelling wildlife. Intensive land management practices, such as logging and agriculture, frequently reduce cavity availability and potentially affect the long-term viability of populations.
  • 2
    The New Zealand long-tailed bat Chalinolobus tuberculatus selects roosts in small knot-hole cavities with specific structural properties relative to available cavities. They also change roosts daily among a large pool of different roosts. Such behaviour is likely to make C. tuberculatus vulnerable to human-induced deterioration in roosting habitat.
  • 3
    This study represents a case study of the degree of sophistication sometimes required to assess availability and quality of roost sites, by testing whether roosts selected by C. tuberculatus also have specific microclimates.
  • 4
    Selection for microclimate was demonstrated by comparing temperature and humidity inside unoccupied maternity roosts with available, apparently unused, knot-hole cavities, large trunk-hollows and ambient conditions.
  • 5
    Compared with ambient conditions, roost and available knot-hole cavities had stable microclimates displaying only small ranges in temperature and humidity. Temperature inside cavities was lower than ambient temperature in the day and was warmer (and peaked) at night. Humidity in cavities was constantly high. Mean temperatures within trunk-hollows (not known to be used by C. tuberculatus) were cooler than mean ambient and roost temperatures, and temperature ranges in hollows were large and fluctuated similarly to ambient temperatures.
  • 6
    Compared with available cavities and hollows, roost cavities had higher minimum temperatures, and maximum temperatures occurred significantly later in the day and continued for significantly longer. Humidity ranges were less and high humidity was maintained for longer.
  • 7
    The results suggest that C. tuberculatus selects maternity roost sites with microclimatic conditions that are likely to accrue substantial energetic benefits. Predicted energy savings for adult bats using roost cavities compared with available knot-holes were 1·1–3·3%, and compared with hollows 3·4–7·3%. Greater energy savings would occur at night and benefit non-volant young.
  • 8
    In order to evaluate adequately and mitigate the full impacts of land-use practices, there is a need for wider tests to provide direct evidence of interactions between habitat management, cavity provision and survival of cavity-dependent wildlife.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Cavities in trees are important for many bird and mammal species for diurnal and nocturnal shelter and as sites for breeding (Nilsson 1984; Raphael & White 1984; Bennett, Lumsden & Nicholls 1994; Newton 1994; Robb et al. 1996). Reproductive success and survival can be influenced directly by the quality of a roost site (Li & Martin 1991; Du Plessis & Williams 1994; Zahn 1999). High quality cavities are generally higher from the ground, allowing for easy detection of, and escape from, predators. They are close to available food and provide a thermal environment resulting in energetic benefits to the occupants (Kendeigh 1961; Stains 1961; Nilsson 1984; Rendell & Robertson 1989; Li & Martin 1991; Du Plessis & Williams 1994; Robb et al. 1996). A large proportion of hole-using species are ‘secondary cavity users’, i.e. they do not excavate or greatly modify their own cavities and have to rely on those that are available (Newton 1994). Therefore, the selection of sites with appropriate qualities is critical, and many species are highly selective in their choice of roost trees and cavities, using sites with specific structural qualities relative to available sites (McComb & Noble 1981; Li & Martin 1991; Elliott, Dilks & O’Donnell 1996). The availability of tree cavities is thought to be a limiting factor for some populations of several hole-using species (Meredith 1984; Newton 1994; Pell & Tideman 1997).

Processes of cavity creation and cavity loss determine availability of usable wildlife cavities within a forest. Cavities are created through wood decay, lightning strike and fire, insect activity and excavation (by primary cavity users), and are lost by deterioration and tree fall (Mackowski 1984; Sedgwick & Knopf 1992). In undisturbed forests these processes achieve an approximate equilibrium (Sedgwick & Knopf 1992). Generally, as individual trees age, the amount of dead cavity-forming wood increases and the number of usable wildlife cavities in a forest increases (Newton 1994; Kirby et al. 1998). Land management practices disrupt the natural processes of cavity turnover. Harvesting, thinning and pruning of trees result in the removal of existing or potential cavity-bearing trees, and along with grazing also reduce the recruitment of trees into cavity-bearing cohorts (Mackowski 1984; Bennett, Lumsden & Nicholls 1994).

Cavity-dwelling bats in temperate zones are more likely to be adversely affected by reduction in abundance and quality of roost sites than many other animals. Bats use cavities for roosting, breeding and hibernating, spending over half of their lives within their roost environment (Kunz 1982). Most bats are secondary cavity users and are not known to modify their roosts structurally (Kunz 1982). Several species are highly selective of the structural characteristics of roost trees and cavities (Vonhof & Barclay 1996; Brigham et al. 1998; Sedgeley & O’Donnell 1999a,b) and apparently the thermal environment within roost cavities (Kalcounis & Hecker 1996; Vonhof & Barclay 1997).

The quality of roost microclimate can have a profound effect on bats, as the thermal regime of their environment is one of the most important factors influencing their energy expenditure (McNab 1982). During summer months, bats form aggregations known as maternity colonies, dominated by breeding females and young (Kunz 1982). The energy demands of breeding females are extremely high during pregnancy and lactation (Speakman & Racey 1987; Kurta, Kunz & Nagy 1990). Breeding females tend to use daily torpor as an energy conservation mechanism less than other sex and reproductive classes of bats (Kurta, Johnson & Kunz 1987; Audet & Fenton 1988; Hamilton & Barclay 1994) because its frequent use can result in a reduction in reproductive success and, ultimately, reduce maternal and juvenile fitness (Racey & Swift 1981; Tuttle & Stevenson 1982; Kunz 1987). Therefore, the selection of a warm maternity roost site that has a physical structure that enables bats to cluster together and derive additional energetic savings of behavioural thermoregulation (Trune & Slobodchikoff 1976) may be one of the most important mechanisms by which breeding females can reduce energy expenditure while minimizing cost to reproductive success (Racey 1982; Kunz 1987; Zahn 1999).

The New Zealand long-tailed bat Chalinolobus tuberculatus Gray 1843 (Vespertilionidae) is a moderately small (8–11 g) insectivorous bat that roosts primarily in trees (Daniel & Williams 1984; Sedgeley & O’Donnell 1999a). Three parallel studies of roosting behaviour and roost selection have shown that, despite changing roosts almost daily among a large pool of different roosts (O’Donnell & Sedgeley 1999), C. tuberculatus is highly selective in its choice of roost trees and roost cavities (Sedgeley & O’Donnell 1999a,b). This behaviour is likely to make C. tuberculatus particularly sensitive to anthropogenic habitat modifications, such as the current and proposed logging of indigenous forest in New Zealand, which targets trees of the size and age classes that commonly contain cavities used as roosts (reviewed in O’Donnell 1999).

Many of the features that C. tuberculatus selects for (e.g. open tree canopy around the roost, large roost-tree diameter, roosts that are high from ground, thicker walls of roost cavities) should result in warmer and more stable microclimate conditions within roosts (McComb & Noble 1981; Calder, Golding & Anderson 1983; Vonhof & Barclay 1997). Therefore, Sedgeley & O’Donnell (1999b) predicted that the knot-hole cavities used as maternity roosts by C. tuberculatus would be of higher quality, with warmer and more stable microclimates than other available knot-holes. They further predicted that large trunk-hollows that were abundant in the study area, but not used by C. tuberculatus, were of lower quality and would not provide microclimates suitable for maternity roosts.

The aims of this study were to test the predictions that C. tuberculatus is selective of roost cavity microclimate by: (i) examining the temperature and humidity of cavities used as maternity roosts by C. tuberculatus in relation to external ambient temperature and humidity; (ii) evaluating roost microclimate selection by comparing maternity roosts with structurally similar (but apparently unused) knot-hole cavities; (iii) examining differences in microclimate between maternity roosts and larger unused trunk-hollows; (iv) evaluating the potential energetic consequences of using cavities with differing qualities of microclimate; and (v) examining the conservation implications of the potential effects of human-induced habitat modification on the availability of high quality roost sites.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study area

The study was conducted in the lower Eglinton Valley, Fiordland National Park, South Island, New Zealand (44°58′ S, 168°01′ E). The valley is of glacial origin, with steep sides and a flat floor, 0·5–1·5 km wide, at c. 250–550 m a.s.l. Tussock grasslands cover much of the valley floor. Temperate southern beech Nothofagus forest covers gentle glacial terraces and outwash fans on the lower hill slopes and the valley walls, which rise steeply to the timberline at 1000–1200 m a.s.l. Red beech N. fusca Hook. f. and silver beech N. menziesii Hook. f. dominate the forest on the valley floor. Forest composition varies, ranging from pure stands of silver beech c. 20 m tall along the forest margin to stands of red beech up to c. 60 m tall further into the forest. Mountain beech N. solandri var. cliffortioides Hook. f. occasionally contributes to the canopy at low altitudes, and becomes more common with increasing altitude. Rainfall averages 2300 mm year−1 in the centre of the valley, but increases markedly in a gradient to > 5000 mm year−1 further up the valley. Climate is cold temperate, with mean monthly temperatures ranging from 2 to 8 °C. In summer, dusk temperatures range from 2 to 18 °C and minimum overnight temperatures from −2 to +13 °C (O’Donnell 1999).

Location of roost cavities

Day roost cavities were located over 4 years (late spring to early autumn, October to March 1993–97) by following bats fitted with radio-transmitters (BD2A 0·7 g; Holohil Systems, Carp, Ontario, Canada). Chalinolobus tuberculatus were caught in canopy-height mist nets (Dilks, Elliott & O’Donnell 1995) set within the forest but close to the forest–grassland edge. Seventy-three bats in total were radio-tracked [14 adult males, 34 breeding females, nine nulliparous females (non-breeders that had not previously given birth) and 16 juveniles]. Bats were followed for as long as the transmitters remained attached and functional [average (± SD) 11·9 ± 6·4 days, range 1–28 days]. Transmitters were on average 5·6–7·9% of the bats’ body mass, higher than the 5% of body mass recommended by Aldridge & Brigham (1988) but below the 10% recommended by Bradbury et al. (1979). Subsequent monitoring detected no adverse affects of transmitters on the health or roosting behaviour of C. tuberculatus (Sedgeley & O’Donnell 1996; O’Donnell & Sedgeley 1999).

One-hundred and forty-nine day roost cavities were found: 84 in live tree trunks, 33 in dead trunks and 32 in large branches (Sedgeley & O’Donnell 1999b). All roosts were knot-hole cavities. Roosts were identified as maternity roosts or otherwise by examining the age, sex and reproductive status of bats caught in harp-traps that were positioned directly outside roost exits (O’Donnell & Sedgeley 1999). Because a high proportion of emerging bats (88·0 ± 18·9%) was captured using this technique, the composition of bats caught in traps accurately reflected the composition of bats using roosts. Additionally, roosts were examined after trapping (by climbing trees) for the presence of non-volant young (O’Donnell & Sedgeley 1999). Maternity roosts were defined as those used predominately by pregnant and lactating females and young. Trapping directly at roosts did not appear to affect roosting behaviour of C. tuberculatus. There were no significant differences in rates of occupancy and reuse, emergence times or survival of young between bat colonies occupying trapped or untrapped roosts (Sedgeley & O’Donnell 1996).

Temperature and relative humidity were measured in unoccupied roost cavities chosen at random from 65 maternity cavities located in live tree trunks (cavities in dead trunks and large branches were generally inaccessible or the trees were unsafe to climb). Chalinolobus tuberculatus in the study area frequently changed roost sites. Breeding females with dependent young switched roosts virtually every day. Levels of roost reuse were low, and roosts were seldom reused during the same season (O’Donnell & Sedgeley 1999). This behaviour made it impossible to measure temperature and humidity in roost cavities while bats occupied them. Because the presence of bats directly influences temperature and humidity inside roosts (reviewed in Kunz 1982), a more valid comparison of the microclimate between roosts and non-roosts could be made if the bats were not present.

Location of available knot-hole cavities and trunk-hollows

The microclimate of maternity roost cavities was compared with that of potentially available cavities of two different types: knot-hole cavities that were structurally similar to cavities used as maternity roosts, but not known to be used by bats, and larger trunk-hollows. To locate these unused sites, the nearest cavity-bearing tree of the same species and with a similar stem diameter as each roost tree was identified. The tree was climbed using a single rope technique and any cavities present were assessed. As sampling was retrospective, minimum and maximum measurements of bat roost cavity characteristics were used as a guideline to assess whether cavities were potentially available to bats (for full details see Sedgeley & O’Donnell 1999b). Cavities were considered to be available to bats if they were dry inside, had relatively uncluttered entrances and were 6–34 m from the ground. Available cavities were classified as either knot-holes that were structurally similar to roost cavities (small–medium sized, well-defined internal cavities, approximately 20 × 20 × 70 cm, probably formed by branch death) or as trunk-hollows with dimensions exceeding those of roost cavities (entrance holes opening into an open hollow tree trunk, internal dimensions of approximately 50 × 50 × 300 cm or greater, probably formed as a result of stem/heart rot). If a tree had multiple cavities, only one cavity (chosen randomly) was included in the data set.

If a random cavity was unoccupied by bats, and there were no droppings present, it was assumed it had not been used by bats. Ten new roosts were identified by the presence of bat droppings in random cavities. These cavities were not included in the data set as it was not known if they had been used as maternity roosts. To assess whether droppings became harder to detect over time, 30 randomly chosen known roosts were examined. Bat droppings were detected in all of these roosts (and some had interiors stained with faeces, urine or body oils) even though bats may not have occupied some of them for 3 years. It was not possible to ascertain if there were droppings in some of the larger hollows. However, as none of 301 roosts found in previous studies (O’Donnell & Sedgeley 1999; Sedgeley & O’Donnell 1999a,b) was located in hollows, it was assumed that C. tuberculatus did not use them.

Temperature and humidity sampling

Small temperature data loggers [Onset Stowaway XTI02, range −39 to +122 °C, accuracy (0·5 °C) with external probes (PB35–60)] and relative humidity (RH) data loggers (Stowaway RH humidity logger, 5–90% RH non-condensing; Hastings Data Loggers, Port Macquarie, New South Wales, Australia) were used to record temperature and relative humidity. A recording station was established within the forest, central to the study area and at the average distance of roosts from the forest edge (120 m; Sedgeley & O’Donnell 1999a), to measure ambient conditions external to cavities. Data on external ambient temperature and humidity were collected continuously from the same location throughout the study. The loggers and probes were screened to shield them from direct sunlight and suspended below the canopy at 17 m above the ground (average roost height; Sedgeley & O’Donnell 1999b) using a rope and pulley system.

The microclimate inside roost and available cavities was measured by inserting temperature probes and humidity loggers directly into the unoccupied cavities and positioning them at the highest point (the place where bats most frequently roost). As roost reuse by C. tuberculatus is so infrequent, it was assumed no bats returned to occupy roosts while measurements were being taken. Data were collected during austral summer months between November 1996 and March 1997. This sampling period corresponded with occupation dates of maternity roosts during pregnancy and lactation recorded during previous seasons (O’Donnell 1999). Data loggers recorded temperature or humidity once every 5 min over a 5-day period. Data were not collected from all cavities at the same time, but as each successive roost cavity was sampled a paired sample from either an available knot-hole or trunk-hollow was collected simultaneously. The data loggers were removed at the end of each 5-day period, downloaded (Onset LogBook for Windows 2·03), and then moved to a new site and introduced into a different roost and equivalent available knot-hole cavity or trunk-hollow. Recording began at 12:00 h and finished on the fifth day at 12:00 h. External ambient temperature was compared with temperature inside 24 maternity roost and available knot-hole cavity pairs, and 11 maternity roost cavities and available trunk-hollow pairs; and external ambient humidity was compared with 11 roost and available knot-hole cavity pairs.

Data were averaged per hour and categorized into hourly intervals over each 24-h period: data collected between 12:01 and 13:00 h were assigned to h-13, etc. Hourly data were used to calculate variables that would illustrate pattern and variation in temperature and humidity for each 24-h period: mean, maximum, minimum; range (maximum minus minimum); mean rate of change per hour (h-2 minus h-1, h-3 minus h-2, etc.); time of day maximum temperature occurred; length of time maximum temperature was maintained; and lag time of cavity maximum in relation to ambient maximum (cavity time of maximum minus ambient time of maximum). These results were then averaged over the recording period to generate mean variables for a 24-h period, which were then compared among roost, available and ambient sites. Day roosts were also used at night, primarily by young bats while females were out foraging. Therefore data were also assigned to two time blocks reflecting day (07:00–21:00 h), and night (22:00–06:00 h) and additional analyses were performed. These time intervals were chosen to correspond with the mean roost exit time and mean return time of radio-tracked and video-monitored C. tuberculatus (O’Donnell 1999).

Data were normally distributed (tested with Wilks–Sharpiro statistic W). Two-way analysis of variance (Statistix for Windows; Analytical Software, Tallahassee, FL) was used to test for differences in temperature and humidity variables among roosts, available knot-hole cavities and ambient conditions, and among roosts, trunk-hollows and ambient conditions. As each roost and available knot-hole cavity or trunk-hollow pairing was located at a different site, a term for site was included to reduce the amount of variance in the model and to allow within-site comparisons to be made once site had been statistically controlled for.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Comparison of ambient and knot-hole cavity temperatures

External ambient temperature (described hereafter as ambient temperature) fluctuated greatly during the 5-day period relative to temperatures within cavities, which had more stable regimes (Fig. 1). All ambient temperature variables (except minimum) were significantly greater in the day than the night (two-sampled t-tests, P < 0·001), whereas roost cavities and available cavities exhibited little variation between day and night. The temperature range of roost cavities (t = 5·48, d.f. = 35·3, P < 0·001) and available cavities (t = 3·29, d.f. = 40·1, P < 0·01) was significantly smaller at night than in the day.


Figure 1. Average hourly temperatures recorded in roost cavities (n = 24) and available knot-hole cavities (n = 24) compared with external ambient temperature over a 5-day period. Recording began at 12:00 h and data points are shown in 1-h increments, where 13 = 12:01 to 13:00 h, etc.

Download figure to PowerPoint

Mean, minimum and maximum temperature and the time of day that maximum temperature occurred varied significantly among sites (two-way anova, F = 2·61 – 45·8, d.f. = 2, 23, P < 0·001). The average mean temperature among individual roosts ranged from 8·6 to 18·3 °C, or from 2·9 °C below to 2·3 °C above ambient temperature. Once variation due to site was taken into account there was also significant within-site variance. Mean temperature variables differed significantly among ambient, roost and available cavities during 24-h, day- and night-time blocks (Table 1). The average mean roost temperature was significantly higher than the ambient temperature over 24 h (Tukey's pairwise comparison of means test, P < 0·05) but the average mean temperature for available cavities did not differ from roost temperature or ambient temperature (Table 1). During the day mean roost temperature averaged 1·2 °C below ambient temperature, and available cavities 1·5° below ambient temperature. Roost and available cavity temperatures were significantly cooler than ambient temperature (Tukey's test, P < 0·05) but not different from each other. At night, ambient, roost and available cavity temperatures all differed from each other (Tukey's test, P < 0·05), with roost cavity temperatures averaging 2·3 °C above ambient temperature, and available cavities 1·5 °C above ambient temperature (Table 1).

Table 1.  Average temperature (± 1 SD) recorded in roost cavities (n = 24) and available knot-hole cavities (n = 24) compared with external ambient temperature. Temperature was recorded over a 5-day period, but results are presented as mean temperature in three time blocks: 24 h (midday–midday), day (07:00–21:00 h) and night (22:00–06:00 h). All statistical differences are summarized using two-way anova (in all cases d.f. = 2, 23) except for difference in temperature lag between available knot-hole cavities and roost cavities, which is compared over 24 h using a paired t-test (d.f. = 23)
VariableExternal ambientAvailable cavitiesRoost cavitiesF/tP
24 h     
Mean (°C)11·95 ± 2·5312·32 ± 2·5212·43 ± 2·75  3·55< 0·05
Maximum (°C)17·37 ± 2·9613·05 ± 2·4113·21 ± 2·84160·69< 0·001
Minimum (°C) 8·56 ± 2·3410·67 ± 2·7911·26 ± 2·73108·53< 0·001
Range (°C) 8·81 ± 1·36 2·38 ± 1·42 1·94 ± 0·89264·81< 0·001
Rate of change h−1 (°C) 0·74 ± 0·13 0·18 ± 0·11 0·12 ± 0·06 25·33< 0·001
Time maximum occurred (h of day)16·91 ± 0·5819·12 ± 2·6721·54 ± 2·55 43·33< 0·001
Length of time maximum held (h) 1·09 ± 0·14 3·94 ± 3·23 5·75 ± 2·43 29·31< 0·001
Lag (cavity time of maximum – ambient time of maximum) (h)   – 2·20 ± 2·55 4·54 ± 2·46 −4·98< 0·001
07:00–21:00 h (day)     
Mean13·48 ± 2·5612·00 ± 2·4412·27 ± 2·71 33·45< 0·001
Maximum (°C)17·37 ± 2·9613·03 ± 2·4013·10 ± 2·79168·36< 0·001
Minimum (°C) 8·77 ± 2·2810·68 ± 2·7611·27 ± 2·73 86·90< 0·001
Range (°C) 8·60 ± 1·36 2·34 ± 1·39 1·83 ± 0·82252·70< 0·001
Rate of change (mean h−1) (°C) 0·86 ± 0·15 0·19 ± 0·12 0·13 ± 0·07299·27< 0·001
22:00–06:00 h (night)     
Mean10·39 ± 2·5111·87 ± 2·7212·69 ± 2·83 50·85< 0·001
Maximum (°C)12·46 ± 2·6912·50 ± 2·5813·02 ± 2·82  3·60< 0·05
Minimum (°C) 8·95 ± 2·5311·28 ± 2·8712·24 ± 2·83 90·62< 0·001
Range (°C) 3·50 ± 0·97 1·21 ± 0·93 0·78 ± 0·44 81·90< 0·001
Rate of change h−1 (°C) 0·54 ± 0·12 0·17 ± 0·12 0·10 ± 0·05130·75< 0·001

Ambient temperature exhibited greater extremes and changed more rapidly than the temperature in roost or available cavities (Table 1). Over the 24-h time period and during the day, ambient average maximum temperature was significantly higher than in roosts and available cavities (Tukey's test, P < 0·05). At night there was no difference among ambient, roost and available cavity maximum temperatures (P > 0·05). Ambient minimum temperature was consistently lower than roost and available cavity minima, and roost cavity minimum was consistently higher than ambient and available cavities for all time blocks (Tukey's test, P < 0·05).

Ambient maximum temperature was reached significantly earlier than maximum temperatures in either roost or available cavities (Tukey's test, P < 0·05). Roost cavities reached maximum temperature significantly later than available cavities (paired t-test, P < 0·001; Table 1). Roost cavities reached maximum temperature either late in the day or at night, whereas ambient temperature never reached maximum at night, and available cavities only rarely (Fig. 2). Roost cavities retained maximum temperature for the greatest length of time (Tukey's test, P < 0·05; Table 1).


Figure 2. Comparison of the time of day maximum temperature occurred in roost cavities (n = 24) and available knot-hole cavities (n = 24) compared with external ambient maximum temperature over a 24-h period (mean of 5 days).

Download figure to PowerPoint

Comparison of ambient and knot-hole cavity humidities

External ambient humidity (described hereafter as ambient humidity) fluctuated greatly and was significantly lower than humidity within roost and available cavities, which were extremely stable (Fig. 3 and Table 2). Humidity in cavities was often maintained at 100% for 1 or more consecutive days. Roost cavities stayed at 100% humidity for 25 (45·5%) of the total 55 days sampled (11 sites, 5 days each), and available cavities for 20 days (36·4%). By contrast, ambient humidity remained at a constant 100% for only 2 days (3·6%). Maximum humidity in roosts was maintained for longer than ambient maximum humidity (Tukey's test, P < 0·05; Table 2).

Table 2.  Average humidity (± 1 SD) recorded in roost cavities (n = 11) and available knot-hole cavities (n = 11) compared with external ambient humidity. Humidity was recorded over a 5-day period, but results are presented as mean humidity in three time blocks: 24 h (midday–midday), day (07:00–21:00 h) and night (22:00–06:00 h). Statistical differences in ambient, available knot-hole cavity and roost cavity humidity during three time blocks (24 h, day and night) using two-way anova (in all cases d.f. = 2, 10)
VariableExternal ambientAvailable cavitiesRoost cavitiesFP
24 h     
Mean (RH percentage)77·2 ± 8·892·8 ± 8·395·4 ± 6·620·39< 0·001
Maximum (RH percentage)93·9 ± 5·097·5 ± 4·397·2 ± 5·4 1·82NS
Minimum (RH percentage)54·5 ± 11·686·2 ± 14·493·2 ± 7·947·20< 0·001
Range (RH percentage)39·4 ± 8·411·2 ± 10·8 4·0 ± 4·195·94< 0·001
Rate of change h−1 (RH percentage) 3·7 ± 0·9 1·1 ± 1·06 0·4 ± 0·484·56< 0·001
Length of time maximum held (h) 3·3 ± 3·4 9·4 ± 9·412·6 ± 10·3 4·20< 0·05
07:00–21:00 h (day)     
Mean (RH percentage)72·2 ± 9·692·0 ± 9·095·3 ± 6·529·65< 0·001
Maximum (RH percentage)93·3 ± 5·497·3 ± 4·497·0 ± 5·6 2·17NS
Minimum (RH percentage)54·5 ± 11·686·3 ± 14·493·3 ± 7·847·46< 0·001
Range (RH percentage)38·9 ± 8·611·1 ± 10·8 3·8 ± 3·993·87< 0·001
Rate of change h−1 (RH percentage) 4·3 ± 0·9 1·3 ± 1·2 0·5 ± 0·484·95< 0·001
22:00–06:00 h (night)     
Mean (RH percentage)85·5 ± 7·894·2 ± 7·295·6 ± 6·8 6·78< 0·01
Maximum (RH percentage)92·1 ± 7·096·3 ± 5·196·6 ± 6·0 1·91NS
Minimum (RH percentage)75·9 ± 9·791·7 ± 9·894·6 ± 7·516·59< 0·001
Range (RH percentage)16·2 ± 6·9 4·6 ± 5·1 2·0 ± 2·242·70< 0·001
Rate of change h−1 (RH percentage) 2·7 ± 0·9 0·8 ± 0·8 0·4 ± 0·417·09< 0·001

Figure 3. Average hourly relative humidity recorded in roost cavities (n = 11) and available knot-hole cavities (n = 11) compared with external ambient humidity over a 5-day period.

Download figure to PowerPoint

Neither roosts nor available cavities showed any significant variation in humidity for all variables measured between day and night (two-sample t-tests, P > 0·05), whereas ambient humidity variables with the exception of maximum humidity varied significantly between day and night (two-sampled t-tests, P < 0·05). In the day, mean humidity in roost cavities averaged 23·1% higher than ambient humidity, and humidity in available cavities was 19·8% higher. At night mean humidity in roosts averaged 10·6% above ambient humidity and humidity in available cavities averaged 8·7% above ambient humidity (Table 2). Ambient humidity had the greatest range and rate of change over all time blocks, and during the day humidity in roost cavities had the smallest range and rate of change (Tukey's test, P < 0·001; Table 2).

Comparison with trunk-hollow temperatures

Trunk-hollows had significantly larger entrances and internal volumes than roost cavities (Mann–Whitney U-tests, P < 0·05). In contrast to roost and available knot-hole cavities, temperatures within trunk-hollows were much less stable, often fluctuating similarly to ambient temperature (Fig. 4).


Figure 4. Average hourly temperature recorded in roost cavities (n = 11) and available trunk-hollows (n = 11) compared with external ambient temperature over a 5-day period.

Download figure to PowerPoint

Roost and ambient average mean temperatures were not significantly different from each other over a 24-h period, but trunk-hollows were significantly cooler than both ambient and roost temperatures (Tukey's test, P < 0·05; Table 3). During the day, mean ambient temperature was significantly higher (Tukey's test, P < 0·05) than mean roost or trunk-hollow temperature, which were not different from each other. Roost cavity temperatures averaged 0·9 °C below ambient temperature, and trunk-hollow temperatures averaged 1·7 °C below ambient temperature. Mean temperature at night differed significantly among all three samples, with ambient temperatures being the coolest and roost-cavity temperatures the warmest (Tukey's test, P < 0·05). Roosts averaged 2·8 °C above ambient temperature and trunk-hollows 1·0 °C above ambient temperature (Table 3). Ambient temperature ranges and rates of hourly change in ambient temperature were significantly greater than in trunk-hollows and roosts. Roosts were significantly more stable than both ambient and trunk-hollows during all time blocks (Tukey's test, P < 0·001).

Table 3.  Average temperature (± 1 SD) recorded in roost cavities (n = 11) and available trunk-hollows (n = 11) compared with external ambient temperature. Temperature was recorded over a 5-day period, but results are presented as mean temperature in three time blocks: 24 h (midday–midday), day (07:00–21:00 h) and night (22:00–06:00 h). All statistical differences summarized over three time blocks using two-way anova (in all cases d.f. = 2, 10) except for difference in temperature lag between roosts and trunk-hollows, which is compared over 24 h using a paired t-test (d.f. = 10)
VariableExternal ambientAvailable hollowsRoost cavitiesF/tP
24 h     
Mean (°C)13·96 ± 2·1613·26 ± 2·0314·41 ± 2·1911·38< 0·001
Maximum (°C)19·63 ± 3·2215·63 ± 2·7115·52 ± 2·4339·67< 0·001
Minimum (°C) 9·15 ± 1·9510·85 ± 2·0112·73 ± 2·3078·91< 0·001
Range (°C)10·47 ± 2·72 4·77 ± 1·42 2·78 ± 1·5666·97< 0·001
Rate of change h−1 (°C) 0·90 ± 0·23 0·39 ± 0·14 0·19 ± 0·1483·24< 0·001
Time maximum occurred (h of day)17·00 ± 0·6319·09 ± 1·5722·09 ± 3·3618·97< 0·001
Length of time maximum held (h) 1·09 ± 0·10 1·69 ± 0·96 5·29 ± 2·7222·61< 0·001
Lag (h)   – 2·09 ± 1·64 5·00 ± 3·34−3·17< 0·05
07:00–21:00 h (day)     
Mean15·14 ± 2·4413·42 ± 2·1714·22 ± 2·1714·49< 0·001
Maximum (°C)19·63 ± 3·2215·61 ± 2·7315·37 ± 2·3639·11< 0·001
Minimum (°C) 9·20 ± 1·9710·88 ± 2·0412·73 ± 2·3075·20< 0·001
Range (°C)10·43 ± 2·75 4·73 ± 1·44 2·63 ± 1·5465·14< 0·001
Rate of change h−1 (°C) 1·03 ± 0·26 0·43 ± 0·16 0·21 ± 0·1788·59< 0·001
22:00–06:00 h (night)     
Mean11·99 ± 1·7813·00 ± 1·9014·74 ± 2·2546·03< 0·001
Maximum (°C)14·70 ± 2·1114·25 ± 1·9515·22 ± 2·35 4·52< 0·05
Minimum (°C) 9·89 ± 1·7711·73 ± 1·9014·08 ± 2·2568·19< 0·001
Range (°C) 4·81 ± 1·57 2·52 ± 0·92 1·14 ± 0·9045·37< 0·001
Rate of change h−1 (°C) 0·69 ± 0·20 0·33 ± 0·12 0·15 ± 0·1160·97< 0·001

Ambient maximum temperature and maximum temperature in trunk-hollows occurred at similar times of day, but roost maximum temperature occurred significantly later in the day (Tukey's test, P < 0·05; Table 3). Maximum temperature occurred in trunk-hollows on average 2 h after ambient maximum temperature, and roosts 5 h after ambient maximum (paired t-test, P < 0·05; Table 3). Maximum temperature in roosts was held for significantly longer than for trunk-hollow or ambient maximum temperatures (Tukey's test, P < 0·05), which did not differ from each other.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Microclimate characteristics of roost cavities

Numerous factors contribute to a roost microclimate (reviewed in Bakken & Kunz 1988) and many have not been considered in this study. For example, the microclimate within maternity roosts can be modified substantially by the metabolic heat generated by roosting bats, increasing the temperature inside occupied roosts 5–10 °C above that of unoccupied roosts (Burnett & August 1981; Kunz 1987; Kalcounis & Brigham 1998). There are also limitations in the techniques used in the present study. Only one temperature probe was used in each roost, and while roost cavities were relatively small there may have been variation in temperature within the cavity. Temperature readings would be affected if the probe touched the wall of the cavity. Humidity data logger accuracy can be reduced above RH 90% (D. Rolfe, personal communication; Hastings Dataloggers), consequently readings of a constant 100% humidity for long periods of time may be inaccurate. There were also differences in temperatures among the 24 C. tuberculatus roosts, but these data were collected at different times during the breeding season. To test if different roosts had different thermal properties (and different energetic advantages) temperature would need to be measured in all roosts simultaneously.

Despite these factors this study clearly shows that cavities provide a different microclimate to external ambient conditions, and maternity roost cavities have distinctive microclimatic characteristics. While there are data on the microclimates of tree roosts used by other bat species (Humphrey, Richter & Cope 1977; Alder 1994; Kurta & Williams 1994; Hosken 1996; Vonhof & Barclay 1997; Kalcounis & Brigham 1998), there are few data gathered from cavities used by maternity colonies and generally sample sizes are small (< 10 roosts measured). Comparisons are difficult because in previous studies (i) there are limited data for maternity roosts, (ii) cavity microclimate was examined while roosts were occupied, and (iii) studies took place in geographical areas with very different climates. However, several findings from previous studies are consistent with those reported here. Cavities used as maternity roosts are insulated against temperature extremes and have significantly smaller temperature and humidity ranges relative to external ambient conditions. Generally, tree cavity or under-bark day roosts are cooler than ambient conditions during the day and warmer at night, with maximum temperatures maintained for relatively long periods (this study; Myotis daubentoni, Alder 1994; Nyctophilus geoffroyi, Hosken 1996; Turbill 1999; M. lucifugus, Kalcounis & Hecker 1996; Desmodus rotundus, Wilkinson 1985). This pattern was also observed in a night-roost in a building (M. lucifugus, Barclay 1982). Roost cavities were not only insulated from extremes in ambient conditions, but also provided a unique thermal climate compared with available cavities (this study; M. evotis, M. Vonhof, unpublished data; Vonhof & Barclay 1997; M. lucifugus, Kalcounis & Hecker 1996). In contrast, M. sodalis roosts under bark provided little insulation, and temperatures inside were elevated above ambient (Kurta & Williams 1994).

Structural characteristics influencing microclimate

Several studies have shown that the physical structure of a roost can directly influence the microclimate within (Vaughan & O’Shea 1976; Bell, Bartholomew & Nagy 1986). For example; roosting surface, thickness of walls and amount of solar radiation will affect heat transfer by convection, radiation, conduction and evaporation, and thus influence the microclimate and energy savings of bats roosting on these surfaces (Bakken & Kunz 1988). Most of the differences in structural features that distinguished C. tuberculatus roost trees and cavities from non-roost sites (Sedgeley & O’Donnell 1999a,b) are those most likely to result in warmer and more stable microclimate conditions, for example amount of open tree canopy around the roost, roost cavity entrance size and roost tree trunk diameter. A relatively open tree canopy will expose roosts to higher levels of solar radiation, which will contribute to cavity warming (McComb & Noble 1981; Calder, Golding & Anderson 1983) and smaller cavity entrances may reduce convective heat loss (Calder, Golding & Anderson 1983). The insulative properties or thermal inertia of wood and bark will increase with diameter, and cavity wall or bark thickness (Derby & Gates 1966; Nicolai 1986), and this in turn results in reduced temperature ranges and lags in temperature change within cavities compared with the ambient temperature (Stains 1961; Sluiter, Voûte & van Heerdt 1973; Maeda 1974; Calder, Golding & Anderson 1983; Alder 1994; Vonhof & Barclay 1997).

Trunk-hollows measured in the present study were structurally very different from roosts. Trunk-hollows had significantly larger entrances and internal volumes than roost cavities, and walls tended to be thinner. These physical characteristics result in more variable internal temperatures that are close to ambient temperatures, and may explain why C. tuberculatus has never used trunk-hollows in the Eglinton Valley, or elsewhere in New Zealand in different habitat types and under different climatic conditions (Gillingham 1996; Griffiths 1996). Results from this study therefore support predictions that physical characteristics can provide meaningful indicators of the thermal suitability of cavities for use as maternity roosts.

Importance of high quality roost microclimates

Insectivorous bats in temperate zones meet their increased energy demands by increasing the frequency or duration of foraging or by conserving energy by using daily torpor (Hamilton & Barclay 1994; Grinevitch, Holroyd & Barclay 1995). Foraging is energetically expensive (Kurta et al. 1989) and may not always be cost effective when weather conditions are cold or wet or insect availability is low. Generally, time spent foraging is reduced under these adverse conditions (Barclay 1985; Racey & Swift 1985; Wilkinson & Barclay 1997) and use of daily torpor is more frequent (Audet & Fenton 1988). Frequent use of torpor by pregnant females results in slowed foetal development and extended gestation (Racey 1973; Racey & Swift 1981). When used by lactating females and by young it leads to depressed postnatal growth rates (Tuttle 1976; Tuttle & Stevenson 1982). Delays in the time it takes for young to become independent may result in both females and young having less time to prepare for hibernation, resulting in lower over-winter survival (Tuttle 1976; Kunz 1987). Consequently, breeding females tend to use torpor less than other sex and reproductive classes of bats (Kurta, Johnson & Kunz 1987; Audet & Fenton 1988; Hamilton & Barclay 1994).

Maternity roosts with a warm stable temperature should allow breeding females to reduce their energy expenditure whilst remaining active and homeothermic for longer periods. This in turn results in an increased rate of gestation, postnatal growth and, ultimately, over-winter survival (Racey 1982; Kunz 1987; Zahn 1999). As breeding female C. tuberculatus do not appear to compensate for the increased energy demands of reproduction by foraging any longer than other reproductive classes (O’Donnell 1999), selection of warm roost sites may be their main mechanism for energy conservation. Female C. tuberculatus select roosts that reach maximum temperatures late in the day and retain high temperatures throughout the night, thus the greatest energy savings are likely to occur at night. Female C. tuberculatus forage throughout the night, leaving young alone in roosts for > 61% of the night (O’Donnell 1999). The thermal benefits of clustering may be reduced after the departure of the females and, as suckling bats are poikilothermic and do not thermoregulate in the early postnatal period (Kunz 1987), high roost temperatures at night are likely to be important for reducing hypothermia and consequent slowed development in young bats.

Use of roost cavities with high humidities by C. tuberculatus appears important. Water as well as energy is a factor that contributes to timing and success of reproduction (Kurta, Kunz & Nagy 1990). Water balance in small bats is sensitive to both temperature and humidity (Herreid & Schmidt-Nielson 1966). Maternity roosts and hibernacula frequently have high humidities (Twente 1955; Baudinette et al. 1994). High humidity reduces evaporative heat loss (Bakken & Kunz 1988; Webb 1995) and prevents dehydration (Van der Merwe 1973).

Energetic advantages of knot-hole cavities

The mean hourly temperature of C. tuberculatus roosts was 0·5 °C warmer than non-roost knot-hole cavities over 24 h, or 0·3 °C warmer in the day and 0·8 °C warmer at night. Can such small increases in temperature result in energetic savings? As resting metabolic rate increases nearly linearly with decreasing temperature (Henshaw 1970), differences in cavity temperature should have a direct effect on the energy expenditure of roosting bats. Entwistle, Racey & Speakman (1997) predicted that a temperature difference of 1·2 °C between maternity roost and non-roost sites (in buildings) of Plecotus auritus would result in an energy saving of 1·0 kJ day−1, or 4% of total roosting energy expenditure. Chalinolobus tuberculatus should also derive significant energy savings from roosting in small cavities with higher temperatures.

Estimates based on rates of energy expenditure by resting adult C. tuberculatus across a range of ambient temperatures using open-flow respirometry (P. I. Webb, unpublished data) suggest that, by using roosts rather than available knot-holes, these bats would achieve a reduction in energy expenditure of approximately 1·1% when roosting during the day and 3·3% when roosting during the night. There are greater temperature differentials between roosts and trunk-hollows; mean hourly temperature in roosts over 24 h was 1·2 °C warmer than trunk-hollows, or 0·8 °C warmer in the day and 1·7 °C warmer at night. Consequently predicted energy savings are greater, and bats using roosts rather than trunk-hollows would obtain energy savings of 3·4% during the day and 7·3% at night. Adult female C. tuberculatus may therefore benefit from the stable temperatures within roosts during the day, and young bats will benefit at night. Respirometry data are not available to calculate rates of energy expenditure by young bats.

Roost selection and implications for conservation

Anthropogenic-induced habitat loss, fragmentation and deterioration is recognized as a key issue facing conservation of biological diversity world-wide (reviewed in Bennett 1998). Several studies indicate that human land-use practices frequently reduce roosting or nesting habitat and adversely affect cavity-dependent wildlife (Scott 1979; Bennett, Lumsden & Nicholls 1994; Bennett 1998). Anthropogenic threats to cavity availability and quality include harvesting, thinning and pruning of trees (Lindenmayer et al. 1990; Mackowski 1984; O’Donnell 2000a), exotic plantation forests overgrowing indigenous woodlands (O’Donnell 2000b), grazing and intensive agriculture (Saunders, Smith & Rowley 1982; Bennett, Lumsden & Nicholls 1994), and competition from introduced cavity nesting birds and mammals (Pell & Tideman 1997).

The present work represents a case study illustrating the degree of sophistication sometimes required to assess roost quality. Identification of high quality cavities is considerably more complex than simply identifying areas of preferred habitat or physical characteristics of roosts. Bennett, Lumsden & Nicholls (1994) suggested that measures of cavity abundance, or even absolute counts of total available cavities, may not accurately reflect the carrying capacity of cavity-dependent wildlife. Several species have exacting roosting or nesting requirements, preferring to use cavities of particular dimensions and physical characteristics or, if influenced by behavioural spacing mechanisms, may require multiple sites within their home range or particular densities and distribution of cavities. Therefore, number of cavities may still be a limiting factor for a population even when cavities are apparently abundant. The present study and other research on the roosting requirements of C. tuberculatus support this view. Sedgeley & O’Donnell (1999b) estimated that 22% of random trees contained cavities available to bats, but only 1·3% contained cavities with physical characteristics that fell within the range of those used by C. tuberculatus. The present study shows that microclimates of structurally similar used and unused knot-hole cavities varied significantly, implying that not all available knot-hole cavities provide microclimates that are optimal for use as maternity roosts. Lower quality cavities, which have microclimates that are suboptimal for maternity colonies, could be used by solitary bats and non-breeding groups that also use knot-hole cavities but have differing thermoregulatory requirements (O’Donnell & Sedgeley 1999; Sedgeley & O’Donnell 1999b). However, it is likely that negative effects on growth, productivity and survival (Brigham & Fenton 1986; Zahn 1999) accrue with long-term use of low quality roost sites (Richter et al. 1993).

Given the high degree of specialization in cavity selection evident in this study and others, it is recommended that land managers adopt more sophisticated techniques for evaluating risks of management practices to cavity-dependent wildlife. There is a need for more rigorous tests of the hypotheses that (i) practices such as logging target the same trees that cavity users prefer to roost in, thereby reducing the availability of good quality roosts, and (ii) that loss of quality roosts will result in reduced fitness and consequently a reduction in long-term viability of populations. It is important to determine if enough cavities of adequate quality are sustained through different logging cycles and at what thresholds negative impacts would occur. Tests would be possible using (i) experimental logging, (ii) modelling the persistence of high quality cavities through different logging regimes (e.g. Lindenmayer & Possingham 1995; Rodréguez & Andrén 1999), and (iii) enhancing habitat through provision of artificial cavities (McComb & Noble 1981; Calder, Golding & Anderson 1983; Boyd & Stebbings 1989).

Experimental logging may be possible for some species of cavity-users, but for threatened species such as C. tuberculatus the risk of losing critical cavities is too high to justify. Modelling is problematic because of the precision required to predict the rate of natural and management-induced cavity turnover and the recovery rates of cavity-bearing trees and cavities (Efford 1999). Enhancement and restoration of roosting sites can be used to mitigate the effects of deterioration and loss of quality due to human influence (Richter et al. 1993; Baudinette et al. 1994). Data on the thermal characteristics of high quality roosts can be directly applied to the design of artificial roosts and the restoration of existing sites. For example, restoration of physical structure and microclimate in caves resulted in increases in numbers of bats using hibernation sites (Myotis sodalis, Richter et al. 1993; Rhinolophus ferrumequinum, R.D. Ransome, unpublished data). Significant increases in juvenile female growth rates and survival occurred at a maternity roost where heaters were used to raise thermal regimes to 27 °C (Rhinolophus ferrumequinum, R.D. Ransome, unpublished data). Before undertaking enhancement or restoration of roost sites it is important to gather sufficient data on roosting requirements and preferences of the target species (McComb & Noble 1981; Calder, Golding & Anderson 1983; Tideman & Flavel 1987). Data should include information on variables such as inter- and intraspecific and seasonal differences in requirements. If information on preferences such as roost microclimate is not incorporated into artificial roost design, the resulting roosts are likely to be of poor quality, and subsequent conservation measures are likely to be ineffective. It is probable that wildlife will not use the sites provided (Neilson & Fenton 1994), or to only a limited degree (Boyd & Stebbings 1989), and long-term use of suboptimal sites, in the absence of high quality sites, may not be beneficial (Brigham & Fenton 1986).


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Thank you to the many people who have assisted with bat-catching, radio-tracking, location of roosts and tree-climbing. I would especially like to thank Colin O’Donnell, Warren Simpson, Dave Mawer and Peter Webb. Thank you to Graeme Elliott for statistical advice, and Ingo Rieger, Allen Kurta, Christopher Turbill, Maarten Vonhof, Peter Webb and Roger Ransome for sharing comparative data. I am very grateful to Colin O’Donnell, Peter Webb, Lindy Lumsden and Ian Jamieson for providing many useful comments on earlier drafts of the manuscript. Thank you to Robert Barclay, R. Mark Brigham and Steve Ormerod, whose critical comments greatly improved the manuscript. The project was undertaken with funding from Department of Conservation Investigation 1758, the Royal Forest and Bird Protection Society (J.S. Watson Trust and Stocker Scholarships) and the University of Otago.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Alder, H. (1994) Erste Erfahrungen mit dem Data Logger: Ereigniszählung vor Baumhohlenquartieren von Wasserfledermäusen, Myotis daubentoni, bei gleichzeitiger Messung mikroklimatischer Werte. Mitteilungen der Naturforschenden Gesellschaft Schaffhausen, 39, 119133.
  • Aldridge, H.D.J.N. & Brigham, R.M. (1988) Load carrying and manoeuvrability in an insectivorous bat: a test of the 5% ‘rule’ of radio-telemetry. Journal of Mammalogy, 69, 379382.
  • Audet, D. & Fenton, M.B. (1988) Heterothermy and the use of torpor by the bat Eptesicus fuscus (Chiroptera: Vespertilionidae): a field study. Physiological Zoology, 61, 197204.
  • Bakken, G.S. & Kunz, T.H. (1988) Microclimate methods. Ecological and Behavioral Methods for the Study of Bats (ed. T.H.Kunz), pp. 303332. Smithsonian Institute, Washington, DC.
  • Barclay, R.M.R. (1982) Night-roosting behavior of the little brown bat, Myotis lucifugus. Journal of Mammalogy, 63, 464474.
  • Barclay, R.M.R. (1985) Long-versus short-range foraging strategies of hoary (Lasiurus cinereus) and silver haired (Lasionycteris noctivagans) bats and the consequences for prey selection. Canadian Journal of Zoology, 63, 25072515.
  • Baudinette, R.V., Wells, R.T., Sanderson, K.J., Clark, B. (1994) Microclimatic conditions in maternity caves of the bent-winged bat, Miniopterus schreibersii: an attempted restoration of a former maternity site. Wildlife Research, 21, 607619.
  • Bell, G.P., Bartholomew, G.A., Nagy, K.A. (1986) The roles of energetics, water economy, foraging behavior and geothermal refugia in the distribution of the bat Macrotus californicus. Journal of Comparative Physiology B, 156, 441450.
  • Bennett, A.F. (1998) Linkages in the Landscape: The Role of Corridors and Connectivity in Wildlife Conservation. IUCN, Gland, Switzerland and Cambridge, UK.
  • Bennett, A.F., Lumsden, L.F., Nicholls, A.O. (1994) Tree hollows as a resource for wildlife in remnant woodlands: spatial and temporal patterns across the northern plains of Victoria, Australia. Pacific Conservation Biology, 1, 222235.
  • Boyd, I.L. & Stebbings, R.E. (1989) Population changes of brown long-eared bats (Plecotus auritus) in bat boxes at Thetford forest. Journal of Applied Ecology, 26, 101112.
  • Bradbury, J., Morrison, D., Stashko, E., Heithaus, R. (1979) Radio-tracking methods for bats. Bat Research News, 20, 917.
  • Brigham, R.M. & Fenton, M.B. (1986) The influence of roost closure on the roosting and foraging behaviour of Eptesicus fuscus (Chiroptera, Vespertilionidae). Canadian Journal of Zoology, 64, 11281133.
  • Brigham, R.M., Vonhof, M.J., Barclay, R.M.R., Gwilliam, J.C. (1998) Roosting behavior and roost-site preferences of forest-dwelling California bats (Myotis californicus). Journal of Mammalogy, 78, 12311239.
  • Burnett, C.D. & August, P.V. (1981) Time and energy budgets for day roosting in a maternity colony of Myotis lucifugus. Journal of Mammalogy, 62, 758766.
  • Calder, T.G., Golding, B.G., Anderson, A.D. (1983) Management for Arboreal Species in the Wombat State Forest. Monash University Graduate School of Environmental Science Environmental Report 16. Monash University, Melbourne, Australia.
  • Daniel, M.J. & Williams, G.R. (1984) A survey of the distribution, seasonal activity and roost sites of New Zealand bats. New Zealand Journal of Ecology, 7, 925.
  • Derby, R.W. & Gates, D.M. (1966) The temperature of tree trunks – calculated and observed. American Journal of Botany, 53, 580587.
  • Dilks, P., Elliott, G., O'Donnell, C. (1995) Mist netting techniques. Ecological Management, 3, 2028.
  • Du Plessis, M.A. & Williams, J.B. (1994) Communal cavity roosting in green woodhoopoes: consequences for energy expenditure and the seasonal pattern of mortality. Auk, 111, 292299.
  • Efford, M. (1999) Analysis of a model currently used for assessing sustainable yield in indigenous forests. Journal of the Royal Society of New Zealand, 29, 175184.
  • Elliott, G.P., Dilks, P.J., O'Donnell, C.F.J. (1996) Nest site selection by mohua and yellow-crowned parakeets in beech forest in Fiordland, New Zealand. New Zealand Journal of Zoology, 23, 267278.
  • Entwistle, A.C., Racey, P.A., Speakman, J.R. (1997) Roost selection by the brown long-eared bat Plecotus auritus. Journal of Applied Ecology, 34, 399408.
  • Gillingham, N.J. (1996) The behaviour and ecology of long-tailed bats (Chalinolobus tuberculatus Gray) in the Central North Island. MSc Thesis. Massey University, Palmerston North New Zealand.
  • Griffiths, R. (1996) Aspects of the ecology of a long-tailed bat, Chalinolobus tuberculatus (Gray, 1843), population in a highly fragmented habitat. MSc Thesis. Lincoln University, New Zealand.
  • Grinevitch, L., Holroyd, S.L., Barclay, R.M.R. (1995) Sex differences in the use of torpor and foraging time by big brown bats (Eptesicus fuscus) during the reproductive season. Journal of Zoology, London, 235, 301309.
  • Hamilton, I.M. & Barclay, R.M.R. (1994) Patterns of daily torpor and day-roost selection by male and female big brown bats (Eptesicus fuscus). Canadian Journal of Zoology, 72, 744748.
  • Henshaw, R.E. (1970) Thermoregulation in bats. About Bats (eds B.H.Slaughter & D.W.Walton), pp. 188232. Southern Methodist University Press, Dallas, Texas.
  • Herreid, C.F. & Schmidt-Nielson, K. (1966) Oxygen consumption, temperature, and water loss in bats from different environments. American Journal of Physiology, 211, 11081112.
  • Hosken, D.J. (1996) Roost selection by the lesser long-eared bat, Nyctophilus geoffroyi, and the greater long-eared bat, N. major (Chiroptera, Vespertilionidae) in Banksia woodlands. Journal of the Royal Society of Western Australia, 79, 211216.
  • Humphrey, S.R., Richter, A.R., Cope, J.B. (1977) Summer habitat and ecology of the endangered Indiana bat Myotis sodalis. Journal of Mammalogy, 58, 334346.
  • Kalcounis, M. & Brigham, R.M. (1998) Secondary use of aspen cavities by tree-roosting big brown bats. Journal of Wildlife Management, 62, 603611.
  • Kalcounis, M. & Hecker, K. (1996) Intraspecific variation in roost site selection in little brown bats (Myotis lucifugus). Bats and Forests Symposium, October 19–21, 1995, Victoria, British Columbia, Canada (eds R.M.R.Barclay & R.M.Brigham), pp. 8191. Working Paper 23/1996. Research Branch, British Columbia Ministry of Forests, Victoria, Canada.
  • Kendeigh, S.C. (1961) Energy of birds conserved by roosting in cavities. Wilson Bulletin, 73, 140147.
  • Kirby, K.J., Reid, C.M., Thomas, R.C., Goldsmith, F.B. (1998) Preliminary estimates of fallen dead wood and standing dead trees in managed and unmanaged forests in Britain. Journal of Applied Ecology, 35, 148155.
  • Kunz, T.H. (1982) Roosting ecology of bats. Ecology of Bats (ed. T.H.Kunz), pp. 155. Plenum Press, New York, NY.
  • Kunz, T.H. (1987) Post-natal growth and energetics of suckling bats. Recent Advances in the Study of Bats (eds M.B.Fenton, P.A.Racey & J.M.V.Rayner), pp. 395420. Cambridge University Press, Cambridge, UK.
  • Kurta, A. & Williams, K.J. (1994) Thermal aspects of Indiana bats (Myotis sodalis) roosting in trees. Bat Research News, 35, 104.
  • Kurta, A., Bell, G.P., Nagy, K.A., Kunz, T.H. (1989) Energetics of pregnancy and lactation in free ranging little brown bats (Myotis lucifugus). Physiological Zoology, 62, 804818.
  • Kurta, A., Johnson, K.A., Kunz, T.H. (1987) Oxygen consumption and body temperature of female little brown bats (Myotis lucifugus) under simulated roost conditions. Physiological Zoology, 60, 387397.
  • Kurta, A., Kunz, T.H., Nagy, K.A. (1990) Energetics and water flux of free-ranging big brown bats (Eptesicus fuscus) during pregnancy and lactation. Journal of Mammalogy, 71, 5965.
  • Li, P. & Martin, T.E. (1991) Nest-site and nesting success of cavity-nesting birds in high elevation forest drainages. Auk, 108, 405418.
  • Lindenmayer, D.B. & Possingham, H.P. (1995) The conservation of arboreal marsupials in the montane ash forests of the central highlands of Victoria, south-eastern Australia. VII. Modelling the persistence of Leadbeaters's possum in response to modified timber harvesting practices. Biological Conservation, 73, 239257.
  • Lindenmayer, D.B., Cunningham, R.B., Tanton, M.T., Smith, A.P. (1990) The conservation of arboreal marsupials in the montane ash forests of the central highlands of Victoria, south-eastern Australia. II. The loss of trees with hollows and the implications for the conservation of Leadbeaters's possum Gymnobelideus leadbeateri McCoy (Marsupialia: Petauridae). Biological Conservation, 54, 133145.
  • McComb, W.C. & Noble, R.E. (1981) Microclimates of nest boxes and natural cavities in bottomland hardwoods. Journal of Wildlife Management, 45, 284289.
  • Mackowski, C.M. (1984) The ontogeny of hollows in blackbutt (Eucalyptus pilularis) and its relevance to the management of forests for possums, gliders and timber. Possums and Gliders (eds A.P.Smith & I.D.Hume), pp. 553567. Surrey Beatty & Sons, Chipping Norton, Australia.
  • McNab, B.K. (1982) Evolutionary alternatives in the physiological ecology of bats. Ecology of Bats (ed. T.H.Kunz), pp. 151200. Plenum Press, New York, NY.
  • Maeda, K. (1974) Eco-éthologie de la grande noctule, Nyctalus lasiopterus, à Sapporo, Japon. Mammalia, 38, 461448.
  • Meredith, C.W. (1984) Possums or poles? The effect of silvicultural management on the possums of Chiltern State Park, north-east Victoria. Possums and Gliders (eds A.P.Smith & I.D.Hume), pp. 575577. Surrey Beatty & Sons, Chipping Norton, Australia.
  • Newton, I. (1994) The role of nest sites in limiting the numbers of hole-nesting birds: a review. Biological Conservation, 70, 265276.
  • Neilson, A.L. & Fenton, M.B. (1994) Responses of little brown Myotis to exclusion and to bat houses. Wildlife Society Bulletin, 22, 814.
  • Nicolai, V. (1986) The bark of trees: thermal properties, microclimate and fauna. Oecologia, 69, 148160.
  • Nilsson, S.G. (1984) The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Ornis Scandinavica, 15, 167175.
  • O'Donnell, C.F.J. (1999) The ecology, spatial organisation and conservation of long-tailed bats, Chalinolobus tuberculatus. PhD Thesis. University of Otago, Dunedin, New Zealand.
  • O'Donnell, C.F.J. (2000a) Conservation status and causes of decline of the threatened New Zealand long-tailed bat Chalinolobus tuberculatus (Chiroptera: Vespertilionidae). Mammal Review, 30, 89106.
  • O'Donnell, C.F.J. (2000b) Distribution, Status and Conservation of Long-Tailed Bat (Chalinolobus tuberculatus) Communities in Canterbury, New Zealand. Report U00/38. Environment Canterbury, Christchurch, New Zealand.
  • O'Donnell, C.F.J. & Sedgeley, J.A. (1999) Use of roosts by the long-tailed bat, Chalinolobus tuberculatus, in temperate rainforest in New Zealand. Journal of Mammalogy, 80, 813923.
  • Pell, A.S. & Tideman, C.R. (1997) The impact of two exotic hollow-nesting birds on two native parrots in the savannah and woodland in eastern Australia. Conservation Biology, 79, 145153.
  • Racey, P.A. (1973) Environmental factors affecting the length of gestation in heterothermic bats. Journal of Reproduction and Fertility Supplement, 19, 175189.
  • Racey, P.A. (1982) Ecology of bat reproduction. Ecology of Bats (ed. T.H.Kunz), pp. 57104. Plenum Press, New York, NY.
  • Racey, P.A. & Swift, S.M. (1981) Variations in gestation length in a colony of pipistrelle bats (Pipistrellus pipistrellus) from year to year. Journal of Reproduction and Fertility, 61, 123129.
  • Racey, P.A. & Swift, S.M. (1985) Feeding ecology of Pipistrellus pipistrellus (Chiroptera: Vespertilionidae) during pregnancy and lactation. I. Foraging behaviour. Journal of Animal Ecology, 54, 205215.
  • Raphael, M.G. & White, M. (1984) Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph 86. Journal of Wildlife Management, 48 (Supplement), 166.
  • Rendell, W.B. & Robertson, R.J. (1989) Nest-site characteristics, reproductive success and cavity availability for tree swallows breeding in natural cavities. Condor, 91, 875885.
  • Richter, A.R., Humphrey, S.R., Cope, J.B., Brack, V. (1993) Modified cave entrances: thermal effect on body mass and resulting decline of endangered Indiana bats (Myotis sodalis). Conservation Biology, 7, 407415.
  • Robb, J.R., Cramer, M.S., Parker, A.R., Urbanek, R.P. (1996) Use of tree cavities by fox squirrels and racoons in Indiana. Journal of Mammalogy, 77, 10171027.
  • Rodréguez, A. & Andrén, H. (1999) A comparison of Eurasian red squirrel distribution in fragmented landscapes. Journal of Applied Ecology, 36, 649662.
  • Saunders, D.A., Smith, G.T., Rowley, I. (1982) The availability and dimensions of tree hollows that provide nest sites for cockatoos (psittaciforms) in Western Australia. Australian Wildlife Research, 9, 541556.
  • Scott, V.E. (1979) Bird response to snag removal. Journal of Forestry, 77, 2628.
  • Sedgeley, J.A. & O'Donnell, C.F.J. (1996) Harp-trapping bats at tree roosts in tall forests and an assessment of the potential for disturbance. Bat Research News, 37, 110114.
  • Sedgeley, J.A. & O'Donnell, C.F.J. (1999a) Roost selection by the long-tailed bat, Chalinolobus tuberculatus, in temperate New Zealand rainforest, and its implications for the conservation of bats in managed forests. Biological Conservation, 88, 261276.
  • Sedgeley, J.A. & O'Donnell, C.F.J. (1999b) Factors influencing the selection of roost cavities by a temperate rainforest bat (Vespertilionidae: Chalinolobus tuberculatus) in New Zealand. Journal of Zoology, London, 249, 437446.
  • Sedgwick, J.A. & Knopf, F.L. (1992) Cavity turnover and equilibrium cavity densities in a cottonwood bottomland. Journal of Wildlife Management, 56, 477484.
  • Sluiter, J.W., Voûte, A.M., Van Heerdt, P.F. (1973) Hibernation of Nyctalus noctula. Periodical Biology, 75, 181188.
  • Speakman, J.R. & Racey, P.A. (1987) The energetics of pregnancy and lactation in the brown long-eared bat, Plecotus auritus. Recent Advances in the Study of Bats (eds M.B.Fenton, P.A.Racey & J.M.V.Rayner), pp. 367393. Cambridge University Press, Cambridge, UK.
  • Stains, H.J. (1961) Comparison of temperatures inside and outside of two tree dens used by racoons. Ecology, 42, 410441.
  • Tideman, C.R. & Flavel, S.C. (1987) Factors affecting choice of diurnal roost site by tree-hole bats (Microchiroptera) in south-eastern Australia. Australian Wildlife Research, 14, 459473.
  • Trune, D.R. & Slobodchikoff, C.N. (1976) Social effects of roosting on the metabolism of the pallid bat (Antrozous pallidus). Journal of Mammalogy, 57, 656663.
  • Turbill, C. (1999) Thermal biology and roost selection of long-eared bats, Nyctophilus geoffroyi. BSc Honours Thesis. University of New England, Armidale, New South Wales, Australia.
  • Tuttle, M.D. (1976) Population ecology of the gray bat (Myotis grisescens): factors influencing growth and survival of newly volant young. Ecology, 57, 587595.
  • Tuttle, M.D. & Stevenson, D. (1982) Growth and survival of bats. Ecology of Bats (ed. T.H.Kunz), pp. 105150. Plenum Press, New York, NY.
  • Twente, J.W. (1955) Some aspects of habitat selection and other behavior of cavern-dwelling bats. Ecology, 36, 706732.
  • Van der Merwe, M. (1973) Aspects of temperature and humidity in preferred hibernation sites of the Natal clinging bat Miniopterus schreibersi natalensis (A. Smith, 1834). Zoologica Africana, 8, 121134.
  • Vaughan, T.A. & O'Shea, T.J. (1976) Roosting ecology of the pallid bat, Antrozous pallidus. Journal of Mammalogy, 57, 1942.
  • Vonhof, M.J. & Barclay, R.M.R. (1996) Roost-site selection and roosting ecology of forest-dwelling bats in southern British Columbia. Canadian Journal of Zoology, 74, 17971805.
  • Vonhof, M.J. & Barclay, R.M.R. (1997) Use of tree stumps as roosts by the western long-eared bat. Journal of Wildlife Management, 61, 674684.
  • Webb, P.I. (1995) The comparative ecophysiology of water balance in microchiropteran bats. Symposia of the Zoological Society of London, 67, 203218.
  • Wilkinson, G.S. (1985) The social organization of the common vampire bat. I. Pattern and cause of association. Behavioral Ecology and Sociobiology, 17, 111121.
  • Wilkinson, L.C. & Barclay, R.M.R. (1997) Differences in the foraging behaviour of male and female big brown bats (Eptesicus fuscus) during the reproductive period. Ecoscience, 4, 279285.
  • Zahn, A. (1999) Reproductive success, colony size and roost temperature in attic-dwelling bat Myotis myotis. Journal of Zoology, London, 247, 275280.

Received 23 December 1999; revision received 9 October 2000