Presented February 1999, At Risk - Species and Habitats at Risk Conference, Kamloops, B.C.

BAT USAGE OF CAVE SYSTEMS ON NORTHERN VANCOUVER ISLAND

 

Martin Davis , Alisa Vanderberg, Island Karst Research, Tahsis, B.C.
Trudy Chatwin, B.C. Ministry of Environment, Lands and Parks, Nanaimo
Monica Mather, 3165 Rock City Road, Nanaimo, BC

Roosting male long-eared bat, probably Myotis keenii.
photo : Mark Kaaremaa

Funded by:
- B.C. Habitat Conservation Trust Fund
- Bat Conservation International
- B.C. Ministry of Environment, Lands and Parks
- Forest Renewal B.C.
- Science Council of B.C.

ABSTRACT

Bat use of the Weymer Creek karst area was studied from 1996 to 1998, following the discovery of a Myotis bat hibernaculum. We used a variety of monitoring techniques backed up by careful observations to develop an understanding of how bats use the study area and to determine the climatic characteristics of subsurface bat habitat. We hope to produce recommendations that will lead to sound management of Myotis bats and their habitat within the west coast rainforest karst environment.

The Weymer caves are used for swarming and hibernation by Myotis volans, M. lucifugus, M. yumanensis and the rare and endangered M. keenii. In addition, Myotis californicus and Eptesicus fuscus use the surface habitat for feeding. Caves used for swarming and hibernation vary in elevation from 500 to 900 m, with the largest concentrations of hibernating bats being found above 800 m. Most hibernating bats are found in deep passages with stable environments of 3-4°C and 100% relative humidity. Caves underlying clearcuts are also used by bats and have stable winter temperatures.

Within the study area, the female Keen's Long-eared Myotis uses low elevation sites (<400m) in summer for feeding and raising young. They join males in cave swarming activity at higher elevations (>650m) between August and September. Our observations suggest that the entire range of habitats from sea level to high elevation and both cave and surface habitats are vital to the life history and survival of at least 6 species of bats.

INTRODUCTION

Nine species of Myotis bats inhabit British Columbia in summer, yet until 1993, no winter aggregation sites were known for any Myotis in British Columbia (Nagorsen and Brigham, 1993). Any new information regarding critical hibernacula sites and characteristics of these sites is vital to bat management and conservation. The Province of British Columbia is initiating legislation and guidelines to regulate forest practices and ensure protection of forest-dependent endangered species at this time, however, due to the paucity of information for forest-using bats, recommendations are based on information from other jurisdictions. Thus, in 1993, when cavers discovered hibernating Myotis lucifugus, Myotis volans and the endangered Myotis keenii in Labyrinth Cave under a Vancouver Island coastal montane forest slated for logging, measures were taken to protect and study the Weymer Creek cave systems.

Since the Labyrinth hibernaculum had its main entrance at 900 m above sea-level, we hypothesized that bat hibernation in coastal areas may be related to the constant cool temperature over the hibernation period in the higher elevation caves. If this proved true, then efforts to locate and conserve bat hibernacula in coastal karst areas might be focused on higher elevation sites. The goal of our study was to determine critical habitat characteristics for bats in this coastal rain forest ecosystem in order to make recommendations for conserving bats and bat habitat within this environment.

Our specific objectives were:

1. To document cave usage by bats in the Weymer karst area through discreet visitation, electronic monitoring, guano collection, and skeleton collection;
2. To determine the physical characteristics of bat hibernacula through year-round monitoring of temperature and humidity patterns in caves of different elevation ranges;
3. To document summer habitat use patterns of bats in caves and the overlying forest; and
4. To determine the impact of clear-cut logging on cave physical characteristics and bat use patterns.

STUDY AREA

The Weymer and Green Creek drainages, located northern Vancouver Island, ranges in elevation from sea level to 1126 m. The study area contains over 150 caves at various elevations, between 0 and 1000 m, and was partially clearcut over 25 years ago (Davis, 1995). The study area is primarily in the Coastal Western Hemlock Biogeoclimatic Zone (CWH, subzones vm1 and vm2) and Mountain Hemlock Biogeoclimatic Zone (MH, subzone mm1). Lower elevations have small stands of old-growth Douglas-Fir (Pseudotsuga menziesii), while red alder (Alnus rubra) and bigleaf maple (Acer macrophyllum) predominate in areas that were logged before 1950.

The cave systems are mostly protected within a 305ha park, while the remainder is within Tree Farm License # 19.

METHODS


Temperature and Humidity Data Logging
To characterize the physical habitats suitable for bat hibernation we deployed OnSet temperature and humidity data loggers in caves and surface control sites at low elevation (0-400 m), mid-elevation (401-650m), and high elevation (651-950m) (Table 1). For low elevation, finding replicate caves of sufficient length was not always possible, so some shorter caves were used. Data was also collected from caves under clear-cuts versus those under forested areas in each elevation category. The pair of temperature and humidity loggers were placed 10 m within the cave entrance and deep within the cave (where possible > 100 m from the entrance). Surface temperatures were monitored in each treatment category. Data was recorded at 1 hour intervals.

We compared temperatures from data loggers in 28 locations (Table 1). To compare temperature differences and variations throughout the winter, we used data collected between October 17, 1996 and March 5, 1997. This time span was determined by the complete data set available to us. Analysis of temperature data included calculating seasonal means and standard errors at logger locations. To compare temperature variability, we calculated the coefficient of variation. The coefficient of variation is a dimensionless quantity that measures the amount of variability relative to the value of the mean. It is calculated by dividing the standard deviation by the mean. We compared mean temperatures and coefficients of variation among caves at different elevations and depths to determine if temperatures were constant and within ranges suitable for hibernating bats. We also compared temperatures and variability in caves under clear-cuts and those in forested areas. Analysis of Variance (ANOVA) and t-tests were used for analysis of categorical variables. We compared mean temperatures and variability with elevation using simple regression analysis.

Bat Activity
We delineated the sampling periods based on seasonal patterns of bat use and the net nights when we first and last noted them: hibernation (mid-October to late May), emergence from hibernation and feeding in early summer (May 20 to July 8), and swarming in late summer (July 29 to September 13).

To document surface and subsurface habitat use by bats we used both direct and indirect inspection methods:
· Hibernation sites were confirmed by cave visits and exploration by cavers;
· Skeletal remains and recently-dead bats were collected from caves to confirm bat use and identify species;
· Guano collection mats were deployed in caves to confirm current bat use;
· Remote ultra-sonic detectors were deployed to monitor and record bat passes at entrances and selected surface sites; and
· Mist-netting was used at forest edges, ponds, and outside cave entrances to identify species and habitat use.

In 1998, we netted at various elevations and habitat types from May 20 to October 10, on 82 net nights. Bats were handled according to R.I.C. standards (Resource Inventory Committee 1998)

Bone identification of bat species, especially Myotis keenii was based on skull measurements and characteristics defined in van Zyll de Jong and Nagorsen (1994).

RESULTS

Winter Temperature: Means
The average winter temperature at our 28 logger sites ranged from 0 to 8°C. Mean temperatures were significantly lower at high elevations (linear regression, df=27, F=28.6, R²=52%, p<0.05, Fig. 2a). Temperatures were also significantly higher deep within caves (approx. 100m) compared with those outside caves (ANOVA, df=24, F=5.14, p<0.05). There was no significant difference between temperatures near caves in clear-cuts versus those in forested areas (t-test, t=-1.0, df=23, p>0.05, Fig. 1a).

Winter Temperature Variation
The mean coefficients of variation were greater at higher elevations (simple regression, df=26, F=4.25, R²=15%, p<0.05, Fig. 2b). The coefficient of variation was also significantly different among cave depth categories. Outside caves and at entrances (0-10m) had higher temperature variations than did deep (50-200m) within caves (ANOVA, df=24, F=3.44, p<0.05, Scheffe's pairwise comparisons, depth category 0-10m differs from 15-45m, and from 50-200m p<0.05). There was no significant difference in coefficient of variation between caves in clear-cuts and forested areas (t-test, t=0.06, df=23, p>0.05, Fig. 1b).

Hibernacula Temperatures and Humidity
We found evidence of bats hibernating in 5 deep cave locations between 592-843m (Table 1). The average winter temperatures ranged from 2.6-3.9°C. The average coefficients of variation ranged from 0.02-0.24 (Fig. 1a,b). Since all hibernacula were found in deep cave sites we looked at the differences between mean temperature and variation in these sites at different elevations, to determine consistent temperature characteristics of bat hibernacula.

The deep cave sites at low elevations had much higher mean temperatures than did those at mid elevation and those at high elevation (ANOVA, df=11, F=28, p<0.05 , Fig. 3a). The coefficient of variation in deep caves did not vary with elevation (ANOVA, df=10, F=2.16, p>0.05). The mean temperatures within deep caves with hibernating bats were significantly lower than those without (t-test, df=9, t=-3.1, p<0.05, Fig. 3b), however the coefficient of variation was not different for deep sites with and without hibernating bats (t-test, df=9, t=0.75, p>0.05, Fig. 3b). We found no differences in mean temperatures or coefficients of variation between deep cave sites under clearcuts and under forests (Mean: t-test, df=9, t=-0.45, p>0.05, Coefficient of Variation: t-test, df=9, t = -0.61, p>0.05, Fig. 3b).

Relative humidity at the largest hibernacula (Labyrinth) was a constant 100% in winter. Other sites have not been analyzed yet.

Cave Habitat Use by Bats

During the summer/early fall season in 1996 and 1997 we captured 91 bats of 5 species over 21 net nights: Myotis volans, 46; M. keenii/evotis, 6; M. lucifugus, 29; M. lucifugus/yumanensis, 6; M. yumanensis, 3; and M californicus, 1 (Davis 1997).

In 1998 we caught 70 bats of 5 species over 87 net nights (Table 2). All but Eptesicus fuscus and Myotis californicus have been captured at cave entrances.

At mid and high elevation caves in early summer, no bats were caught. Emergent bats were documented at high elevation cave entrances between May 23 and 27th. However, we were surprised to find 2 bats, (probably M. volans), hibernating in Wormhole Cave on June 21, over 300 metres from the nearest known entrance. This date was later than has previously been observed in the study area. No bats were seen at this site during an October 10 return visit, but a bat was roosting near the entrance.

There was no recorded activity at low elevation caves in early summer during a single night's netting effort.

By late July, there was swarming activity (multiple bats flying in and out of caves that are used for hibernation) by male Myotis of 4 species (Myotis volans, M. keenii, M. lucifugus, M. yumanensis) at the mid to high elevation caves (capture ratio: 1.413 bats/net night) (Table 2). The greatest swarming activity occurred at entrances to the longer caves with the most stable temperatures, such as the Weymer System, Ursa Major System, and Wormhole Cave. Guano collection sheets record the highest rates of deposition in caves at high elevation (229 droppings on 18 sheets). At mid elevation caves, we documented most bat activity at entrances which are lower exits to lengthy, deep cave systems that originate at higher elevations and have stable, cool airflows, such as the Slot Canyon entrances to the Weymer System. Caves with all entrances at mid elevation, such as Fallen Giant, Headwall, and Whistling, yielded few bat captures and detections: 1 capture out of 5 net nights and 18 passes out of 8 detector nights over 2 years. Guano collection sheets also show little deposition at this elevation (15 droppings on 14 sheets).

Swarming activity at cave entrances in August began at 23:30 hours, whereas other forest activity began at dusk (about 21:30). Swarming commenced progressively earlier as the daylength shortened . By Sept. 2, bats were being caught by 21:20. The swarming activity increased through August and by early September appeared to peak. Our highest capture ratio, 5.25 bats/net night, occurred on September 2. In 1996 and 1997, however, the swarming peaked in mid-August (Aug. 11-14). Females and juveniles did not appear with the swarming males until early September in 1998. The relative abundance of females for this period is only 3% that of males, over 3 years of netting. The overall capture ratio of bats at cave entrances during late summer is 1.28 bats/net night.

The capture ratio for low elevation caves during late summer is 0.429. Two significant low elevation caves (Boneyard and Liquid Sky) exhibited no bat activity. However, a 100 m long cave, Knoll Hole, showed activity during netting on August 11, 1998 between 20:30 and 21:00. Two post-lactating female Myotis keenii and 1 juvenile male M. keenii were captured at the cave; 2 more long-eared bats were briefly caught, but escaped. Apart from this site, no other cave-related bat activity has been detected below 550 m elevation.

Surface Habitat Use by Bats

Early Summer
Some bat use was documented (through bat detection and netting) at all elevational ranges. Bats were seen and detected flying along the littoral zone of Tahsis Inlet and the Leiner River estuary from February 1998 to September 5, 1998. In a quiet back eddy of the Leiner River, several Myotis sp. bats were detected based on call frequencies ranging from 50-80Khz. Eptesicus fuscus was heard flying along the Leiner River (based on calls from 20-30Khz). One lactating female Myotis keenii was captured at the Green Creek estuary on July 8. The capture ratio for low-elevation surface netting is 0.333 bats per net-night.

At mid to high elevation there is occasional feeding activity in the forest/clearcut margins and marshes. The capture ratio in the mid-high elevation surface is 0.1 bats per net-night.

Late summer/early fall
As late as September 4, calls consistent with Myotis bats were heard regularly over Tahsis Inlet and the adjacent Leiner estuary. There was feeding activity along the forest margins and marshes at high elevation but no bats were captured. Bats were also seen flying along trails in the forest and heard feeding over dolines.

Skull Identifications
All bat skulls were collected from cave passages above 730 m elevation (Table 3), with the notable exception of 1 specimen (Myotis sp.) from Liquid Sky Cave at 380 m elevation. This skull was coated with calcium carbonate, common to caves, which suggests an older specimen. It was not identifiable at the species level. Myotis lucifugus, M. yumanensis, M. keenii, M. volans, and possibly M. evotis were identified from skulls, including recently deceased (<2 years) individuals.

DISCUSSION

Temperature monitoring deep in caves showed that mean winter temperature increases with elevation decline (Fig. 2a), whereas the coefficient of variation (CV) is low for deep cave sites at all elevations. We also found that the CV outside caves is significantly higher than inside (bar graph ANOVA significant). The increased scatter of points on the regression graph (Fig 2b) at higher elevations is due to the range of CV in high elevation locations. All the points showing high CV are at cave entrances or outside. At low elevations, temperature variation was similar among cave depths, possibly due to the constant influence of the sea temperature in some low caves, particularly one with a tidal sump.

The coefficient of variation increases nearer to entrances because of high temperature variations outside the cave entrances and fluctuations in air flow direction between the surface and the stable deep cave environment. We found no apparent effect of clear-cutting on winter temperature and variation in and near caves.

Our sampling showed that bats hibernate in Labyrinth, Wormhole, Deer Drop, Fracture and Slot Canyon Caves. These caves range in elevation from 592 m to 843 m. They are all deep and long (>100 m). Their deep-cave mean winter temperatures range from 2.64-3.98° C, and the CV is very low: 0.020-0.242. These conditions are best for long, undisturbed periods of torpor and low oxygen and fat consumption (McManus 1974, Tuttle and Stevenson 1978, Nagorsen and Brigham 1993).

No bats were found hibernating below 592 m elevation. Although deep, low elevation caves appear to have stable temperatures, the mean winter temperatures and metabolic costs may both be too high for hibernation (Hock 1951, Stones 1965, Davis 1970, McManus 1974, Tuttle and Stevenson 1978).

Bats use the entire study area at all times of year. The mid to high elevation deep caves are used for hibernation by Myotis keenii, M. volans, M. lucifugus, and M. yumanensis from mid October to mid-June, then for swarming from late July to late September. Low elevation riparian areas and littoral zones are used for feeding. At low elevation streams and Knoll Cave, we captured only female and young Keen's Long-eared Myotis. It appears that female Myotis keenii bats are feeding and raising young at low elevations and then moving to higher elevation caves for swarming and hibernation. This suggests that the study area is very important to Keen's Myotis for two critical aspects of its life history: rearing young and hibernation.

Climate has been shown to affect timing of reproduction in bats (Grindal 1992), and probably affects timing of hibernation also. Thus, the late emergence of the two Myotis bats in Wormhole cave may be the result of a cool, wet spring. Likewise, the annual variation in swarming behavior might be linked to weather patterns as discussed by Fenton (1969).

Bats use caves under both clearcuts and forests for hibernation and swarming. We noted that Myotis volans used a 15 m wide, 25 m deep pit (Skypot) in a clearcut for feeding. This was the only recorded summer feeding use within a clearcut. However, we are not sure if the use of this cave for feeding is related to its unique thermal characteristics which may or may not be related to clearcutting, as we did not monitor temperatures here.

In summer, males predominate at higher elevation cave and surface sites. Females and young of the year appear with swarming males at the caves in mid August. This difference in sexual distribution over an elevational and temporal range may be due to a variety of physiological and abiotic factors such as variable energy demands between the sexes, prey availability, and climate. Females require warmer areas with greater insect abundance to meet the demands of raising young (Thomas 1988, Barclay 1991). Furthermore, the low number of bats documented feeding at high elevation bogs and marshes could be explained by a relatively wide distribution of more productive aquatic feeding areas at lower elevations. Green Creek and Leiner River at sea level, Malaspina and Perry Lake at 100 m elevation provide good feeding areas. However, we do not know where bats roost during the early or late summer period.

Skeletal remains confirmed the presence of Myotis keenii, M. volans, M. lucifugus, and M. yumanensis in the caves. However, one of the long-eared skulls collected has measurements that fall within the Myotis evotis range. To clarify taxonomic uncertainties, DNA sampling and analysis is currently being conducted. Until taxonomic uncertainties for field identification of M. keenii are resolved, we will be unable to determine species identity with absolute certainty (Firman and Barclay 1993).

At Weymer, diverse habitats at a wide range of elevations combined with deep caves make for excellent conditions for hibernation, feeding, and raising young. Our findings regarding bat hibernation will be useful for locating other hibernacula on Vancouver Island.

CONCLUSIONS


1. Myotis bats hibernate in caves with stable temperatures between 2.4 - 4°C and 100% relative humidity. In this coastal montane environment, these are deep caves (>100m long) and occur above 500m elevation. The largest aggregations are found above 800m elevation.

2. Clearcutting appears not to have an effect on winter temperatures deep within caves.

3. Myotis bats use the mid to high elevation caves year-round except for the period from late July to mid-June. Swarming occurs from late July to late September. Hibernation occurs from mid-October to mid-June.

4. Low elevation forest and riparian areas are used by Keen's Long-eared Myotis for feeding and raising young.

ACKNOWLEDGMENTS

LITERATURE CITED