Campbell River Bat Project: Inventory and Habitat Enhancement

Transcription

1 Campbell River Bat Project: Inventory and Habitat Enhancement Prepared for: B.C. Hydro Bridge-Coastal Fish and Wildlife Restoration Program Burnaby, BC V3N 4X8 November 2002 Prepared by: Mandy Kellner Pacific Slope Consulting 1868 Norwood St. Prince George, BC V2L 1X6 and Sal Rasheed Parks Canada Agency th Ave SE Calgary, AB T2G 4X3

2 i EXECUTIVE SUMMARY Bats are an integral part of healthy forested ecosystems. Bat habitat is primarily located in valley-bottom riparian corridors. Unfortunately, such habitat is often negatively impacted by human activities, such as flooding from hydroelectric development. In the Campbell River watershed of Vancouver Island, there was extensive flooding and loss of low-lying riparian habitat due to the construction of three dams in the 1950s. A large fire in 1938 also significantly altered the surrounding forest. Such occurrences have the potential to reduce the quality and availability of bat habitat, and so impact bat populations. Perhaps because of limited natural habitat, colonies of bats began to roost in the hydroelectric generating stations of the Campbell watershed. Bats roosting in generating stations may startle workers, creating unsafe work conditions. Bat colonies could also produce large guano accumulations and unsanitary work conditions. The historic impacts of reservoir creation and the current presence of bats inside the stations prompted BC Hydro to consult the BC Ministry of Water, Land, and Air Protection, and led to initiation of this 2-year study. The goals of this study were: 1) to provide context for the situation in the generating stations through a watershed inventory of bat species and bat habitat, 2) to investigate the bat colonies roosting in the stations, 3) to offer advice on exclusion, and 4) to provide mitigation and habitat enhancement by installing artificial roost structures. In order to determine species presence, distribution, and reproductive status, we sampled bats, using mist-nets and detectors, at 34 locations around the Campbell watershed. We captured 401 bats of 4 species: Little Brown bats (Myotis lucifugus), Yuma bats (M. yumanensis), long-eared bats (Western Long-Eared or Keen s Myotis (M. evotis/keenii)), and California Myotis (M. californicus). Both at and away from the generating stations, captures in the lower watershed were dominated by Little Brown and Yuma bats, but also included long-eared bats. In the upper watershed, Little Brown and Yuma bats made up a smaller proportion of captures, the rest being long-eared and California bats. The lower watershed housed a large proportion of reproductive female bats. To assess hydroelectric impacts on foraging habitat, we examined activity levels of bats at reservoirs and natural water bodies. Reservoir sites, characterized by fluctuating

3 ii water levels, had significantly less bat activity than did natural sites. Reservoirs in their present condition are not as attractive to bats as are natural ponds. Therefore, preferred foraging habitat (in the form of natural valley-bottom wetlands and riparian areas) was lost when reservoirs were inundated, and this habitat loss has not been compensated for through the creation of larger water-bodies. Mitigation efforts involving wetland creation and shoreline re-vegetation at reservoir edges will likely benefit bats. Our findings on roost selection show that bats are using natural structures and habitats that are rare in the lower watershed, as well as using assorted human structures (see photos in Appendix 1). We radio-tagged 15 bats and located 7 roosts in human structures (including a bat house, cedar roofing and siding on buildings, wooden bridges, and a generating station), 7 rock roosts in cliffs, road cuts, and quarries, and 19 tree roosts. Tree roosts were all snags in decay classes 2 through 6, in open areas of forest. Snags were either veteran trees, standing alone or in remnant old forest patches, or larger-than-average snags in thinned second-growth forest. Although human structures may provide ideal roost habitat for some species of bats, the selection of veteran trees and old-growth patches highlights the importance of protecting remaining old forest habitat and habitat elements in managed landscapes. Inside the generating stations, we found a heated environment housing maternity colonies of 2 species, with colonies comprised of an estimated 73 Yuma bats in Ladore station, and 97 Little Brown and Yuma bats in Strathcona station. There is a night roost site behind John Hart station, which does not impact station operations. At Ladore and Strathcona stations, we identified and described numerous entrance locations. Past efforts at exclusion have illustrated that, to be effective, a concerted effort is needed to seal and monitor all of the known entrance locations on an ongoing basis. We recommended that exclusion measures be carried out in winter 2001/2002. The exclusion was not completed, and monitoring and worker reports in June 2002 showed that bats were still using both stations. In light of the age of the buildings involved, the number of access points, and the level of commitment needed to effectively exclude the bats, we recommend a revised approach to managing bats in the stations. Exclusion work to-date should be maintained, with particular attention paid to the sliding roof covers, including maintaining their seals and closing the covers securely. When resources permit, further exclusion work should be carried out in wintertime. Finally,

4 iii workers should continue to be educated regarding the real and perceived risks of having bat colonies in the stations. Education should focus on the precautions that can be taken to reduce the already-low risk of contracting diseases, including rabies, from bats or guano. A regular clean-up program should be initiated to deal with the accumulating guano. With the current colony sizes in the stations, guano piles can easily be swept up once or twice a year, or when workers need to be in the stations. As mitigation for both the exclusion effort and the historic impacts of dam creation, we provided bat houses for habitat enhancement. We constructed and installed 6 bat houses of 2 styles. Two houses, 1 of each style, were placed at each generating station. We monitored the use by bats and the interior temperatures of the houses in Two of the maternity-style houses were used on occasion. This was encouraging in light of the lack of exclusion from the stations. In summary, hydroelectric development in the Campbell River watershed has had, and continues to have, an influence on bat populations. The watershed supports reproductive populations of at least 4 species of bats. Two of these species use generating stations as maternity colony sites. Excluding these colonies would require much effort and many resources. Instead, we recommend an approach of maintaining (and ideally expanding on) the exclusion work accomplished to-date, and providing ongoing worker education regarding bats and health issues. To ensure healthy populations of bats throughout the watershed, managers should continue to protect, and if possible, enhance habitat for these important animals. Enhancement may be accomplished through the creation of foraging habitat, such as wetlands, and provision of roosting habitat, such as snag creation and installation of bat houses. Our recommendations derived from this work are as follows: RECOMMENDATION 1 - FORAGING HABITAT ENHANCEMENT: Develop wetlands and riparian habitat around reservoirs, by stabilizing water levels in artificial ponds and encouraging the revegetation of shorelines with native plant species, in order to mimic natural conditions.

5 iv RECOMMENDATION 2 ROOSTING HABITAT ENHANCEMENT: Protect remnant forest patches and individual veteran trees and snags, both in old-growth and second-growth forest stands. Create canopy gaps and groups of snags, rather than isolated structures, in thinned, mature forests. RECOMMENDATION 3 - MONITORING: Monitor the 6 installed bat houses by visual inspections each summer, to document use over several years. RECOMMENDATION 4 - EXCLUSION: After 2 years of study of both the bat populations and BC Hydro operations, we felt that complete exclusion of bats from the generating stations was not viable. Instead, accept a low number of bats inside the stations, continue to maintain and expand on exclusion work accomplished to-date, and provide sound worker training and education for working around the bats. ACKNOWLEDGEMENTS This project received funding from the BC Hydro Bridge Coastal Fish and Wildlife Restoration Program. We also received support from the staff at the John Hart BC Hydro office, the Ministry of Water, Land, and Air Protection in Campbell River, and Les Peterson and the Ministry of Forests Research Branch. Vanessa Craig, Patrick Garcia, Leigh-Anne Isaac, Corinna Wainwright, Russell Atkins, and many volunteers provided hard work in the field. Vanessa Craig (Ecologic Consulting) generously performed the quantitative analysis of bat calls using a neural net call-recognition program. The Bairstows have provided much information and were involved with both the construction and installation of the bat houses. We kindly received permission from BC Hydro, TimberWest, and BC Parks to conduct work on their leases.

6 v TABLE OF CONTENTS EXECUTIVE SUMMARY...I ACKNOWLEDGEMENTS... IV TABLE OF CONTENTS...V 1. BACKGROUND AND RATIONALE STUDY AREA METHODS CAPTURE RADIOTELEMETRY AND ROOST DESCRIPTION ACOUSTIC SAMPLING STATION INSPECTIONS TEMPERATURE PROFILES RESULTS SPECIES INVENTORY Species captured Capture success Population structure Species accounts HABITAT ASSESSMENT RESERVOIRS AS FORAGING HABITAT HABITAT ASSESSMENT ROOSTING HABITAT Roost structures Roost tree characteristics Roost site location BATS IN GENERATING STATIONS Station inspections Exclusion MITIGATION THROUGH HABITAT ENHANCEMENT MONITORING Use of bat houses Temperature monitoring Monitoring bat activity levels across the watershed EXTENSION/COMMUNICATION DISCUSSION BAT SPECIES AND DISTRIBUTION ROOSTING HABITAT HYDROELECTRIC DEVELOPMENT AND BAT HABITAT BATS IN GENERATING STATIONS...39

7 vi 6. SUMMARY LITERATURE CITED...42 APPENDIX 1. ROOST SITES PHOTO GALLERY...46 APPENDIX 2. SAMPLING SITES - LOCATIONS AND DATES...48 APPENDIX 3. ROOST SITE SUMMARY...50 APPENDIX 4. EXAMPLES OF SONOGRAMS OF BIG BAT CALLS...52 APPENDIX 5. ACOUSTIC MONITORING DATA FOR BAT ACTIVITY IN THE CAMPBELL WATERSHED...53 APPENDIX 6. TEXT OF INTERPRETIVE SIGNS INSTALLED AT THE STRATHCONA DAM CAMPSITE...55

8 1 1. BACKGROUND AND RATIONALE Bats are an important component of the Campbell River ecosystem, where these insectivorous mammals eat vast quantities of insects nightly. Bat foraging and roosting habitat is often located in riparian corridors (Holroyd et al. 1994, Firman et al. 1995, Rasheed and Holroyd 1995, Wilkinson et al. 1995, Vonhof 1996, Vonhof and Barclay 1996, Grindal et al. 1999, Waldien and Hayes 2001). Hydroelectric development may alter or eliminate riparian corridors, due to flooding of low-lying portions of a landscape. In the Campbell River watershed of Vancouver Island, there was extensive flooding and loss of valley-bottom riparian habitat due to the construction of three dams in the 1950s. The resulting reservoirs inundated 5208 ha of upland forest, 417 ha of riparian habitat (within 30 m of a watercourse), and 936 ha of wetland (BCFWRP 2000). This loss of habitat was compounded by the fire history of the watershed in 1938, a large (30,000 ha) fire burned much of the lower watershed, creating the current even-aged, homogeneous forests. These occurrences may have significantly reduced the availability of bat habitat, and in particular, of natural roosting sites. Bats on northern Vancouver Island and elsewhere in British Columbia commonly roost under exfoliating bark and in cracks in wood (Rasheed & Holroyd 1995, Vonhof 1996, Vonhof & Barclay 1996, Kellner 1999). These types of roosts are generally associated with large-diameter snags or old trees, both of which are rare habitat elements across much of the watershed. Perhaps as a result of habitat loss, bats began roosting in several of the generating stations in the Campbell River watershed (Kellner pers. obs., Rasheed pers. obs., Grindal 1996). In order to understand why bats would seek out and roost in the noisy, bright generating stations, it is necessary to understand the dependence of bats on temperature. Bats in B.C. are heterothermic they can adjust their body temperature. By not having to maintain a warm body core at all times, they can save considerable energy (Altringham 1996). However, there are costs in terms of reproduction because growth is temperature dependent. Fetuses will develop more slowly at a lower body temperature, so gestation time will increase. If breeding season temperatures are too cool, then fetuses may even be reabsorbed and females may forgo reproduction for a year. Once the young are born, their growth rate continues to fluctuate with temperature, as does the ability of their mothers to produce milk. In general, warmer temperatures mean a shorter reproductive period from gestation to weaning and flight of the young bats, and thus more time in late

9 2 summer/ fall for females and their pups to store fat reserves before winter hibernation (Holroyd 1993). Therefore, the temperature in a maternity roost has far-reaching consequences for the reproductive success of the colony. Bats in northern climates will select very warm roost sites in an effort to increase the chances of successfully rearing young. Having a heated maternity roost, such as a generating station, may have large reproductive payoffs for the local bat population. Although artificial roosts such as stations could therefore be important to bat populations, use of these roosts may bring bats and people into conflict. In the Campbell watershed, the potential for bats to create unsanitary working conditions and to startle people working in the dangerous interior of the stations led to safety concerns and the desire to remove the bats. Exclusion of the bats from the generating stations and concurrent construction of alternate roosting structures was deemed necessary. Special care and appropriate timing are required when excluding bats, because several facts about bat biology make bat populations vulnerable to disturbance by humans. Bats are exceedingly long-lived, often living to be over 20 years of age. They reproduce very slowly, usually have only 1 offspring each year, and if disturbed at a maternity colony, may abandon their single pup. Thus, bat populations are easily impacted, particularly by disturbances at maternity colonies. In order to remove the bats while minimizing negative impacts, an initial assessment of the stations was conducted in 1996 (Grindal 1996) and recommendations for mitigation were given. In 2000, exclusion measures were initiated by B.C. Hydro, with the sealing of the major access points used by bats. Nevertheless, bats continued to roost in the generating stations. The Campbell Watershed Bat Project was initiated in 2001 to provide context for, and information on, the ongoing occupancy of the stations by bats. To investigate if alternate habitat was available and provide background information on species presence and distribution, we conducted an inventory of bat species and an assessment of bat habitat in the Campbell River watershed. We also investigated the situation in the generating stations, offered advice on exclusion, and installed artificial roosts for habitat enhancement and mitigation.

10 3 The specific objectives of this project were to: 1. Conduct an inventory of bat species in the watershed, to determine relative abundance and distribution, 2. Examine use of reservoirs as foraging habitat, 3. Examine natural roosting behaviour through a radio-telemetry study, 4. Provide an assessment of critical habitat areas and recommendations for protecting bat populations in the watershed, 5. Study the bats using the BC Hydro generating stations in the Campbell watershed, to determine species, estimate colony size, locate roost sites and access points, 6. Recommend measures to exclude bats from the generating stations, 7. Provide artificial roosting habitat for bats, and 8. Monitor use of the artificial roost structures. The research will contribute to the conservation-oriented management of bats in an industrial setting.

11 4 2. STUDY AREA The study area encompassed the Campbell River watershed, upstream of the town of Campbell River, on Vancouver Island, British Columbia (Fig.1). Elevations ranged from 40 to 305 m above sea level. A portion of our work was conducted at the Strathcona and Ladore generating stations, which are operated by BC Hydro and create the Upper and Lower Campbell Lakes, respectively. Away from the stations, we sampled at diverse habitats including natural ponds and wetlands, slow-flowing creeks, edges between forest and clearings, drawdown areas along the edges of large reservoirs (Buttle, Upper Campbell, Lower Campbell, and John Hart Lakes), and open water on these reservoirs. Figure 1. The Campbell River watershed and associated hydroelectric developments (From BCFWRP 2000.) Sampling sites were all valley-bottom sites, in the very dry Maritime Coastal Western Hemlock subzone (CWHxm) or the submontane moist maritime CWH variant (CWHmm1). The CWHxm has warm, dry summers, with May to September precipitation

12 5 levels of mm. The primary tree and shrub species are Douglas-fir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylla), western redcedar (Thuja plicata), salal (Gaultheria shallon), and dull Oregon-grape (Mahonia nervosa). Also significant are red alder (Alnus rubra), shore pine (Pinus contorta), ocean spray (Holodiscus discolor), and baldhip rose (Rosa gymnocarpa). The CWHmm1 sites are generally higher sites, above the CWHxm, and are dominated by western hemlock, Douglas-fir, amabilis fir (Abies amabalis), red huckleberry (Vaccinium parvifolium), Alaskan blueberry (V. alaskaense), salal, and dull Oregon-grape (Green and Klinka 1994).

13 6 3. METHODS 3.1 CAPTURE Objective 1. Conduct an inventory of bat species in the watershed, to determine relative abundance and distribution. Objective 4. Provide an assessment of critical habitat areas and recommendations for protecting bat populations in the watershed. Between June 2001 and August 2002, we captured bats using mist-nets and on one occasion, a harp trap. All capture sites are detailed in Appendix 2. Netting effort and capture data were recorded on RIC-format datasheets (Resources Inventory Branch 1998). For consistency and comparability, netting effort was described in terms of netnight equivalents. One net-night was defined as a 6 m wide, 4-tier net open for at least 1.5 hours post-sunset. We divided the watershed into upper and lower sections based on habitat differences (Fig. 2). The lower watershed includes Strathcona Dam and below most of this area was burned in 1938, and is dominated by low-elevation, rolling topography and second-growth forest stands. The upper watershed Upper Campbell Lake, Buttle Lake, and the Elk River valley is much more mountainous, with older forest and numerous rocky outcrops. Figure 2. Location of mist-netting sites in the Campbell Watershed (adapted from BCFWRP 2000). The black line indicates the division between the upper and lower watersheds. We mist-netted on 77 nights at 34 sites around the watershed (267.5 net-nights) (Table 1). The majority of our netting effort was focused on the lower watershed, where the generating facilities are located. We stratified our sampling and our results to include

14 7 sampling at colonies (Ladore, Strathcona, and 3 bridges) and sampling at sites away from known colonies in the lower and upper regions of the watershed. Table 1. Sampling effort by location. Location # sites # nights # net-nights Lower non-colony colony Upper All sites With the exception of females in the later stages of pregnancy, we placed captured bats in cloth bags for at least ½ hour before processing, to allow time for excretion of any recent ingested food items. Bats were processed at the end of the capture session to avoid recaptures. We then determined species, sex, age class, forearm length, and weight, and released the bats. In the early summer, females that showed no sign of reproducing were described as not obviously pregnant. If they showed no evidence of pregnancy or lactation by late July, they were then classed as non-reproductive. 3.2 RADIOTELEMETRY AND ROOST DESCRIPTION Objective 3. Examine natural roosting behaviour through a radio-telemetry study. Objective 4. Provide an assessment of critical habitat areas and recommendations for protecting bat populations in the watershed. Select bats (heavier than 6.0 g and not in the later stages of pregnancy) were radiotagged with Holohil LB g transmitters before release. These bats were tracked to their roosts daily, beginning the following day, until the transmitter fell off or stopped working. Roosts were described according to RIC standards (Resources Inventory Branch 1998), and roost location was recorded using a hand-held GPS unit (Appendix 3). When bats used trees as roosts, we recorded detailed tree and habitat characteristics at the roost and at a paired random plot. Paired plots were 100 m away in a random direction, with the tree nearest the 100 m mark chosen as the plot centre. Plots were 2 sizes: 0.05 ha circular plots (7 plots) and 0.01 ha square plots (31 plots) to best reflect the local ecosystem type. Each plot was described with site and wildlife tree forms, for

15 8 which we measured all trees with a diameter at breast height (dbh) of at least 10 cm. Decay classes for appearance, crown, bark, and wood were assigned based on the BC Wildlife Tree classification. 3.3 ACOUSTIC SAMPLING Objective 2. Examine use of reservoirs as foraging habitat. Objective 4. Provide an assessment of critical habitat areas and recommendations for protecting bat populations in the watershed. We had 2 sampling schemes involving acoustic sampling of bat activity levels (all sites are described in Appendix 2). Firstly, to investigate relative use by bats of natural and reservoir-influenced foraging sites, we used a randomized-block design. Each block consisted of a) the edge of a pond not affected by hydroelectric development, b) the reservoir edge, and c) a site out on the reservoir. When sampling out on the reservoir, we were at least 100 m from shore in all directions. Within a block, sites were sampled on the same night and were located as close together as possible (usually within 1 km), with the farthest sites 5.5 km apart (Figure 3). We sampled at 7 blocks of sites around Lower and Upper Campbell Lakes and Buttle Lake. Each site was monitored for the first hour after dusk. Figure 3. Schematic diagram of the randomized-block sampling scheme. There are 3 sites (pond edge, reservoir edge, and open water) within each of 7 sampling blocks.

16 9 Secondly, as part of our monitoring plan, we recorded bat echolocation calls at 32 sites around the watershed. These sites were stratified as: a) at generating stations, b) near stations within a 5 km radius, and c) far from stations greater than 5 km away. Sites were sampled twice in 2001 and once in Acoustical sampling was done with automated Anabat bat detectors, and all bat calls were recorded on tape. Calls were divided into commuting passes and feeding buzzes (Resources Inventory Branch 1998). A pass is a series of two or more discernable echolocation calls, and indicates the presence of a flying bat. A buzz is the obvious increase in call rate as a bat approaches and attempts to capture an insect, and indicates that a bat is attempting to feed in that location. Foraging rate was calculated as number of buzzes divided by number of passes. Tapes were analyzed by ear and with Anabat software to determine the number of passes and buzzes of little bats (likely Myotis species) and big bats (likely Lasionycteris noctivagans, Lasiurus cinereus, and possibly Eptesicus fuscus). 3.4 STATION INSPECTIONS Objective 5. Study the bats using the BC Hydro generating stations in the Campbell watershed, to determine species, estimate colony size, and locate roost sites and access points. Objective 6. Recommend measures to exclude bats from the generating stations. At the generating stations, we (two observers and a safety representative from BCH) used interior inspections and exit counts to determine roost sites, numbers of bats, and entrance/exit locations. Interior inspections revealed visible live and desiccated dead bats and bat sign, including guano, urine staining, squeaking noises, and shredded fibreglass insulation. During exit counts, we monitored both the inside and outside of the stations to determine where bats were exiting from the buildings. We also captured bats at the exits or immediately outside of the stations. A sample of bats from Strathcona (n = 20, white bands) and Ladore (n = 7, yellow bands) were banded with bright plastic, numbered, split-ring arm bands to allow us to definitively document use of bat houses by bats from generating stations (Resources Inventory Branch 1998). Several of these bats were re-sighted in subsequent visits in 2001; however, restricted access to stations in

17 resulted in only 1 interior inspection, during which no banded bats were observed in the stations or in the bat houses. 3.5 TEMPERATURE PROFILES Objective 4. Provide an assessment of critical habitat areas and recommendations for protecting bat populations in the watershed. Objective 5. Study the bats using the BC Hydro generating stations in the Campbell watershed, to determine species, estimate colony size, and locate roost sites and access points. Objective 7. Provide artificial roosting habitat for bats. We recorded temperatures inside the generating stations and other colony sites in For comparison, we also recorded the ambient temperature outside in the shade. This data provided a target temperature range for the bat houses, which were constructed for habitat enhancement and mitigation. Temperatures in the bat houses were monitored in 2002, and can be manipulated in the future by changing the location of the houses and the amount of solar radiation they receive. All data was obtained using Hobotemp data loggers recording every ½ or 1 hour. We calculated average temperature for the entire time that the loggers were deployed (Julymid-September 2001 and June August 2002). We also calculated the average daily variation in temperature, based on a 24-hour day. This measure reflects the variability of the temperature at a site, and we included it because bats should prefer stable as well as warm temperatures.

18 11 4. RESULTS 4.1 SPECIES INVENTORY Species captured We captured 401 bats of 4 species (Little Brown bats (Myotis lucifugus), Yuma bats (M. yumanensis), California Myotis (M. californicus), and Long-eared bats (M. keenii/evotis)). We used fur colour and texture / appearance, with forearm length as a consideration, to distinguish between adult M. yumanensis and M. lucifugus. Juveniles of these 2 species were only assigned a species if captured in conjunction with an identifiable adult. Longeared bats, either Keen s Myotis (M. keenii, a provincially Red-listed species) or the Western Long-eared Myotis (M. evotis), are indistinguishable in the field (Nagorsen and Brigham 1993). As illustrated by Fig.4, captures in all areas were dominated by M. lucifugus. Figure 4. Captures by species in the upper and lower watersheds and at colonies. For a small percentage of animals, we were unable to distinguish between M. lucifugus and M. yumanensis. We also identified species through analysis of recorded echolocation calls. We recorded calls at 32 sites as part of our monitoring program. The majority of calls were from Myotis species, for which we did not attempt to discriminate between species. In order to confirm the presence of non-myotis species in the study area, we selected and analyzed random samples of calls from non-myotis species. These calls were analyzed

19 12 qualitatively (sensu O Farrell et al. 1999) and quantitatively. Qualitative analysis identified three commonly seen types of call patterns calls with a minimum frequency around 22 khz, calls with a minimum frequency around 25 khz, and calls with a minimum frequency around 27 khz (Appendix 4). Although we did not capture any non- Myotis species, and thus have no local reference calls, published literature suggests that the first call pattern belongs to Lasiurus cinereus, which is reported to have a minimum frequency near 22 khz (O Farrell et al. 1999), and the remaining 2 patterns to Eptesicus fuscus (minimum frequency of khz, Betts 1998) and/or Lasionycteris noctivagans (minimum frequency khz, Betts 1998, O Farrell et al. 1999). Quantitative analysis with a neural-net call recognition program (performed by V. Craig, Ecologic Consulting) gave different results. The neural net program classified 5 of 10 random call samples as L. noctivagans. These calls had ending frequencies of 25 and 27 khz. Two calls were identified as E. fuscus, with ending frequencies of 22 and 27 khz. The remaining 3 samples, with frequencies of 22 and 25 khz, were classified as unknown. No calls were classified as L. cinereus. Based on our Anabat call recordings, it seems probable that either all or a subset of L. cinereus, L. noctivagans and E. fuscus were recorded in the watershed. Other species that may potentially occur in the study area, and about which we did not record any information, are M. volans (Long-legged Myotis), and Corynorhinus townsendii (Townsend s Big-eared bat) Capture success Apart from when netting at colonies, we were more successful at catching bats in the upper watershed compared to in the lower (Fig.5). On any given night, we were most likely to catch M. lucifugus and M. yumanensis, either in the upper or lower watershed. As expected, when we were netting at known colony sites (including generating stations), our capture success almost tripled compared to our success elsewhere (Fig. 5).

20 13 Bats per net-night myca myke/ev upper lower colony Bats per net-night upper lower colony mylu myyu juvenile mylu/yu Figure 5. Capture effort by location, for different species. Juvenile M. lucifugus and M. yumanensis were indistinguishable in the field, and are grouped together Population structure Both sexes and ages of bats were captured in both regions of the watershed. There is an obvious female sex bias in the lower watershed, even away from known maternity colonies. This bias is reduced in the upper watershed (Fig. 6). Not only were there more females in the lower watershed, but a higher proportion of them were reproductive (see individual species accounts). Number of individuals upper lower colony female male Figure 6. Sex bias of bats captured at different locations. In summary, the lower watershed supports a lower diversity of bat species, but this area is very important for reproductive females. For all species except M. californicus, reproductive females made up a greater proportion of captures in the lower versus the upper watershed. These results are detailed in the following individual species accounts.

21 Species accounts Myotis lucifugus - Little Brown bats M. lucifugus was the most commonly captured bat in the study area, and was captured at 23 out of 34 netting sites (68%) (Fig. 7). Both upper and lower regions of the watershed supported reproductive populations (Fig. 8). M. lucifugus form colonies in snags and cliffs (Nagorsen and Brigham 1993, Kellner 1999, this study), and in human structures. Several colonies were identified in human structures in both regions, including at Strathcona generating station, under a wooden bridge, under loose house siding, under cedar shakes on a roof, and in a rock quarry (see Section 4.2.1). Number of adults repro non upper lower colony Figure 7. Location of M. lucifugus captures. Figure 8. Population structure of captured adult M. lucifugus. Repro = reproductive females (possibly pregnant, pregnant, lactating, post-lactating). Non = males, not-obviously-pregnant and non-reproductive females. Dates for pregnancy (Table 2) agree with those reported for bats in the Okanagan (Nagorsen and Brigham 1993). However, we had a notably early capture of a volant juvenile M. lucifugus or M. yumanensis, on June 19, We suspect it was a M. lucifugus, based on a relatively large forearm size (34.8 mm) compared to its low weight and juvenile appearance. The bat, weighing 4.0 g, was captured at a relatively high pass (Drum Lakes, elev. 300 m), at a time when we were just beginning to catch pregnant adult female bats.

22 15 Table 2. Timeline of reproduction for M. lucifugus. Reproductive stage Date first observed Date last observed Pregnancy 11- Jun 13 Aug Lactation 5 Jul 15 Aug Post-lactation 17 Jul Volant juvenile 1 20 Jul (19 Jun) 2 1 Either M. lucifugus or M. yumanensis Jun capture of a juvenile seems anomalous. Myotis yumanensis - Yuma bats Reproductive populations of M. yumanensis were found in both regions of the watershed (Figs. 9 and 10). Reproductive bats used snags and human-made colony sites, including a wooden bridge, the Ladore and Strathcona generating stations, and bat houses (see Section 4.3.1). Dates of reproduction were identical to those reported for M. lucifugus (Table 2). Number of adults repro non upper lower colony Figure 9. M. yumanensis capture locations. Figure 10. Population structure of captured adult M. yumanensis. Myotis keenii/evotis - Long-eared Myotis Long-eared bats were captured at 2 sites in the upper watershed and 6 sites in the lower (Fig. 11). Capture success for long-eared bats was significantly less than either M. lucifugus or M. yumanensis. In the upper watershed, much less netting effort was required to catch Long-eared bats than the lower watershed. However, although both regions supported reproductive populations (Fig. 12, Table 3), there was a greater proportion of reproductive females in the lower watershed. Radio-tagged non-

23 16 reproductive females roosted in south-facing cliffs and in snags. We did not locate any maternity colonies for these species. Number of adults repro non upper lower colony Figure 11. Capture locations for M. keenii/evotis. Figure 12.Population structure of captured adult M. keenii/evotis. Table 3. Timeline of reproduction for M. keenii/evotis. Reproductive stage Date first observed Date last observed Pregnancy 16 Jul Lactation 22 Jul Post-lactation 3 Sept Volant juvenile 1 Myotis californicus - California Myotis M. californicus was only captured in the upper watershed, where a large proportion of the captures was comprised of reproductive females (Figs. 13 and 14, Table 4). Two of the M. californicus captured at Drum Lakes, Strathcona Park, were unusually coloured; they were strikingly golden orange, with very dark contrasting ears and wings. This colouration is usually found in the interior of B.C. Seven other M. californicus had the usual dark coastal colouration.

24 17 Number of adults repro non upper lower colony Figure 14. Population structure of captured Figure 13. Capture locations for adult M. californicus. M. californicus. Table 4. Timeline of reproduction for M. californicus. Reproductive stage Date first observed Date last observed Pregnancy 19 - Jun Lactation 10 - Aug 5 - Sept Post-lactation Volant juvenile Sept

25 HABITAT ASSESSMENT RESERVOIRS AS FORAGING HABITAT To investigate the use of reservoirs by bats as foraging or drinking sites, we sampled at 7 blocks of sites. Each block included a site at the edge of a natural pond unaffected by large water fluctuations, the edge of a reservoir, and out on the reservoir (Fig. 15) (see Methods, Section 3.3). Figure 15. Potential bat foraging habitat: a natural beaver pond (left) and Upper Campbell Lake reservoir edge (right). Activity levels and foraging rates were lower at both types of reservoir sites than at the natural pond sites. Total bat activity, including all passes and buzzes, was significantly different between the 3 habitat types (Kruskal Wallis H = 7.641, P = 0.02; Fig. 16). Specifically, activity levels decreased from ponds to edges to open water. Multiple comparisons between habitat types showed a higher number of bats were present at the natural ponds than at open water habitats (Dunn s Q = 2.714, P = 0.02).

26 19 Figure 16. Bat activity recorded at pond, reservoir edge, and reservoir open water sites (n = 7 for each site). Foraging activity, which was low at all sites, also had a decreasing trend across ponds, edges, and open water (Fig.17), although we did not detect any statistical differences (H = 2.387, P = 0.27). To better address this question, we recommend a more substantial sampling scheme, to gather more data on foraging activity. Figure 17. Foraging rate recorded at pond, reservoir edge, and reservoir open water sites (n = 7 for each site).

27 HABITAT ASSESSMENT ROOSTING HABITAT Roost structures Radio-tagged bats To examine where bats were roosting, we located bat roosts using radio-telemetry, visual searches, and local reports. We radio-tagged 15 bats of 3 species (Table 5), captured at various locations in the watershed. All structures selected as roosts were in sunny, open areas, such as trees in forest gaps, south-facing cliffs, building walls, and roof slopes. Rock crevice roosts included natural rock bluffs and man-made areas in road cuts and quarries. The wooden bridges used as roosts were made of large logs with narrow horizontal crevices between the logs. In the concrete bridge roost, bats used the space between the concrete span and the supporting wood/concrete pillars at either end. Roost structure, location, and dates used are detailed in Appendix 3. Table 5. Roost structures used by 15 radio-tagged bats. Location Species Sex Reproductive Roost site condition Lower Campbell M. yumanensis F Pregnant Building (generating station) Lower Campbell M. yumanensis F Lactating Snags, bathouse (TimberWest Camp8) Other bats Reproductive females Lower Campbell M. lucifugus F Possibly pregnant Building (cedar siding) Buttle Lake M. lucifugus F Not obviously pregnant Cliff, building (cedar shake roof) Lower Campbell M. lucifugus F Pregnant Cliff (road cut), snag, wooden bridge Lower Campbell M. lucifugus F Pregnant Snags Lower Campbell M. lucifugus F Pregnant Cliff (quarry) Lower Campbell M. lucifugus F Lactating Wooden bridge Lower Campbell M. lucifugus F Lactating Wooden bridge Elk River valley M. keenii/ evotis F Not rep. Cliff Elk River valley M. keenii/ evotis F Not rep. Cliff Lower Campbell M. keenii/ evotis F Not rep. Cliff, snag Elk River valley M. lucifugus F Post-lactating Snag, cliff Lower Campbell M. lucifugus F Post-lactating Snags Lower Campbell M. lucifugus M Not rep. Snags

28 21 Reproductive female bats used snags, natural and blasted rocks, human structures, and a bat house. The bat house was used by a lactating M. yumanensis. This bat was captured in 2002 within 1 km of the former TimberWest Camp 8 site, which had housed large colonies of bats until fall 2001 when the buildings were demolished. The bat used several snags in conjunction with the bathouse. Non-reproductive radio-tagged bats were located only in snags and natural rock cliffs (although see comment on the use of steel cranes as roosts of un-tagged bats, Table 6). Other identified colony sites We also found numerous bat colonies through visual searches and local reports (Table 6). These sites contribute to our knowledge of bat habitat use. Unfortunately, we were rarely able to tell the reproductive status of these colonies. Table 6. Colony sites located without radio-tracking bats. Roost site Bat houses maternity style - at TimberWest Camp 8 site confirmed use by maternity colonies, including use by radio-tagged lactating M. yumanensis (see table 5 above) Bat houses maternity style - at Ladore and Strathcona generating stations Large Ladore bat house, and gap between house and generating station Concrete/wood bridge on Hwy 28 maternity colony (captured juveniles) TimberWest Camp8 building roofs now demolished maternity colony (observed juveniles) Metal roof of dynamite shed Buildings cedar shake roofs and loose siding Cedar shake roofs on BC Parks interpretive signs Steel cranes at Strathcona station not maternity colony - used in late summer by many post- and non-reproductive bats Roost tree characteristics Seven of the 15 radio-tagged bats roosted in trees or trees and other structures. In total, we located 19 roost trees. Three reproductive females (1 lactating M. yumanensis, 2 pregnant M. lucifugus) used 7 of the trees. Two post-lactating female M. lucifugus, 1 non-reproductive female M. keenii/evotis, and 1 male M. lucifugus used the other 12 trees. For analyses of tree characteristics, we compared roost trees to available trees, defined as trees of the same size (over 4 m tall) and decay classes (2 through 6) as

29 22 known roost trees. These criteria provided 101 available trees. We compared roost to available trees in terms of tree species, decay class, and bark class using chi-square analyses, and height and diameter using Mann-Whitney U tests. We compared roost to random plots for percent cover of different vegetation strata, using paired t-tests. Tree species Bats roosted in Douglas-fir (10 roosts), western hemlock (4 roosts), amabilis fir (3 roosts) and western white pine (2 roosts) trees (Fig.18). Other available tree species were western redcedar, red alder, and western yew (Taxus brevifolia). Bats selected for Douglas-fir (Χ 2 = 5.5, P = 0.019), avoided western redcedar (Χ 2 = 7.86, P = 0.005), and used other species as available, although the amabilis fir were used entirely by 1 male M. lucifugus roost available Proportion Douglas-fir Western hemlock Amabilis fir Western white pine Western redcedar Other Figure 18. Species of roost (n = 19) and available (n = 101) trees. Decay and bark class Roost trees were in decay classes 2 through 6. Bats tended to avoid trees of class 2 (Χ 2 = 3.32, P = 0.069) and select trees in class 4 (Χ 2 = 3.24, P = 0.070), while other decay classes were used as available (Fig. 19). Bark class may be even more important that overall decay class, because loose bark is the structural element commonly used as a roost (Kellner 1999, Nagorsen and Brigham 1993). Bats roosted in trees of bark classes 2 through 5 (Fig. 20), selecting trees in bark classes 3 (Χ 2 = 5.85, P = 0.016) and 4 (Χ 2 = 5.60, P = 0.018), which are characterized as having % of their bark remaining, often in the form of large loose flakes.

30 roost available roost available Proportion Proportion Decay class Bark class Figure 19. Decay classes of roost and available trees. Figure 20. Bark classes of roost and available trees. Size of roost trees Roost trees were larger than available trees in terms of both diameter at breast height (dbh) and height (Fig. 21). Roost trees had a mean dbh of 36.7 ± 19.3 cm. Available trees had a mean dbh of 19.0 ± 9.6 cm (U = 214.5, n = 120, P < 0.001). Roost trees had a mean height of 19.0 ± 11.8 m, and available trees, 12.4 ± 6.4 m (U = 543.5, P = 0.003) roost availabl Proportion Height (m) Figure 21. Height (m) of roost and available trees. Vegetation cover We recorded percent cover of different vegetation strata at each plot, and compared roost to paired random plots (Fig. 22). Roost plots had significantly less tree cover (21.8 ± 17.7 vs ± 24.7 %; t = 3.36, P = 0.003), and denser tall shrub (25.8 ± 15.5 vs.15.5

31 24 ± 16.1 %; t = 1.939; P = 0.068) and low shrub/herb layers (71.5 ± 33.7 vs ± 25.9 %; t = 2.876, P = 0.010), reflecting the presence of canopy gaps and the associated well-developed understory. However, there was no significant difference in the stem density at roost (727 ± 355 stems / ha) or random (853 ± 668) stands (U = 180, n = 19, P = 0.98) roost random % cover trees > 10 m shrubs 2-10 m < 2 m Vegetation height Figure 22. Vegetation cover at roost and random plots (n = 19) Roost site location The snags selected as roost trees were most often isolated remnant trees or were in remnant patches of old forest, surrounded by younger forest (58 % of trees) (Table 7). Roost snags were also commonly in gaps in mature, thinned, second-growth forests (37 %). Table 7. Habitat description for tree roosts (n = 19) used by radio-tagged bats. Habitat description # tree roosts % of tree roosts Contiguous old-growth (Strathcona park) 1 5 Old-growth remnant wetland edge 3 16 Old-growth remnant or single veteran tree in 8 42 younger stand Gap in mature, thinned second growth 7 37 The distance between capture and roost sites provides a rough idea of how far bats may travel from a roost when they are foraging at night. The average capture roost distance was 1.5 ± 1.3 km, with a maximum distance of 4 km travelled by a pregnant M.

32 25 yumanensis who was captured near Fry lake and roosted in the Strathcona generating station. Bats used from 1 to 5 roosts during their respective monitoring periods, staying in a roost from 1 to 10 days. For bats who used greater than one roost, the average distance between sites was 1.0 ± 0.99 km, and a maximum of 3.2 km travelled by a pregnant M. lucifugus.

33 BATS IN GENERATING STATIONS Station inspections At John Hart generating station, we focussed our efforts on the workshed behind the station. There were no reports of bats inside the station, but the workshed reportedly housed a colony until a new roof was added several years ago. We found large guano accumulations outside the workshed door, and mist-netted 5 male and female M. lucifugus and 1 male M. yumanensis in the vicinity of the shed. We observed bats using the sheds as night roosts. This behaviour is not unusual, and artificial structures are commonly used as night roosts (Kunz 1982, Perlmeter 1996). We found no evidence of day roosts in the area. Presumably, the increased shading caused by the new roof had caused the temperature in the shed to change and the colony to leave. At Strathcona and Ladore stations, there were bats inside the stations in spite of high ultrasonic noise levels, bright lights, and the exclusion measures undertaken in 2000 (sealing of sliding roof covers). Strathcona houses a mixed colony of M. lucifugus (Little Brown bats) and M. yumanensis (Yuma bats), while in Ladore there are solely M. yumanensis. Both are maternity colonies, made up of adult females and, later in the summer, their offspring (Table 8). This segregation of sexes is to be expected, since female bats generally seek out very warm environments to promote growth of their young, while males roost elsewhere, in cooler sites (Hamilton and Barclay 1994). Table 8. Date of capture, number of bats, reproductive class, and sex of bats captured at maternity colonies in generating stations, Station Date Species Adult females 1 Juveniles Total bats NOP P L PL Male Female Ladore 17/7 M. yumanensis Strathcona 18/7 M. lucifugus M. yumanensis 7 20/7 M. lucifugus M. yumanensis /8 M. lucifugus 2 11 M. yumanensis Reproductive classes: NOP = not obviously pregnant, P = pregnant, L = lactating, PL = postlactating.

34 27 Adult bats were already in the stations when we first examined the sites on June 28. It is likely that the colony forms well before this time. We first observed juvenile bats on July 12. On July 17 we were still catching pregnant females and could see newborn bats in the stations. On August 15, all the females captured at Strathcona were lactating, suggesting that any reproducing bats had given birth by this time. The average time from birth to flight is 3 weeks, so the pups should all be flying by early- to mid-september. Bats continue to live in colonies after the young are born, where the warm temperatures lead to faster growth of the young. We observed bats in the generating stations when we last checked on September 12 (Strathcona) and September 13 (Ladore), although a nearby maternity colony under a wooden bridge had already broken up and dispersed by September 2. Based on interior inspection and exit counts we estimated minimum colony sizes of 97 and 73 for Strathcona and Ladore, respectively. We believe to these to be conservative counts, because of the difficulties we encountered in observing and counting bats. Not all of the bats in a station can be seen while roosting inside during the day, due to their location in crevices and under flaps. However, exit counts alone also did not provide reliable counts for the number of bats in a station, because bats could be seen inside the stations well after dusk. This very unusual behaviour means that the bats are missing prime evening foraging time (Rydell et al. 1996). Exit counts are generally considered a reliable method to estimate colony sizes (Thomas and LaVal 1988, Resources Inventory Branch 1998), but in the case of the generating stations, they must be combined with internal counts. One reason for bats not exiting may be that the bright artificial lighting (on 24 hours each day) and the lack of windows and natural lighting make it impossible to know when it is dusk Exclusion Recommended methods and timing for exclusion of bats from stations were described in detail in the Year 1 Progress Report (Kellner and Rasheed 2001). As stated in the progress report, the long history of use and the ideal conditions inside the stations suggest that only a comprehensive, ongoing exclusion program (i.e. the sealing of ALL known access points) will be effective in keeping bats out of the stations. We advised that exclusion must be done in winter, when bats are not in the generating stations. All identified entrance locations must be sealed up, and these seals maintained throughout the summer, which is the bats reproductive period and when they will be

35 28 seeking access to their former roost site. Exclusion measures must not be undertaken after early March, as bats may begin to occupy the stations in the spring. We identified access points in 2001, and in February 2002, we inspected the stations to ensure that all bats had vacated before sealing began (activities are detailed in Table 9). BC Hydro staff were to begin sealing all access points shortly after this inspection. Unfortunately, this step was not completed. Inspections in summer 2002 showed that, while certain areas had been sealed, there were still many open access points at each station. Reports from BC Hydro staff confirmed that there were bats present inside both stations in June At this point, one-way exit flaps over the access points were proposed, but there were no BC Hydro resources to do this work. In summary, circumstances prevented effective sealing of the access points to the stations, and so bats returned in 2002 and continued to use the stations as maternity colony sites. Table 9. History of exclusion activities, Summer 2001: identified access points, Dec. 2001: delivered progress report detailing access points and sealing recommendations, Dec. 14, 2001: removed guano from stations, Feb. 5, 2002: inspected stations in winter to confirm all bats had left observed no new guano, Feb. 8: notified John Hart biologist that it was safe to begin sealing access points, Feb. 12: installed 6 bat houses, Jun. 6: requested John Hart staff to notify us if any bats were seen in stations, Jun. 7: received report of over 20 bats inside Strathcona station, Jun. 10: received confirmation that not all the known access holes were sealed, Jun. 10,11: discussed installation of one-way flaps to exclude bats, Jul. 20: Strathcona sliding roof covers opened for maintenance leading to colony eviction late in the reproductive season. However, there are still a few bats observed in the station. In July 2002, the sliding roof covers on Strathcona station were opened for several weeks for routine maintenance activities. This resulted in the displacement of the colony, with very few bats observed in the station at this time. However, the colony will likely return to the station next spring.

36 MITIGATION THROUGH HABITAT ENHANCEMENT We installed bat houses at or near the generating stations to provide mitigation for a) the impacts of excluding bats from the stations, and b) the loss of roosting habitat that occurred as a footprint impact of reservoir creation. Six bat houses were built and installed, with 2 each at John Hart, Ladore, and Strathcona stations. Bat houses were constructed using two different designs (Figs. 23, 24). The standard maternity colony style (Tuttle and Hensley 2000) is a flat, rectangular, multi-chambered style that can house bats. We constructed double-wide versions (600-bat capacity) at Strathcona and Ladore, and 2 singles near John Hart station. The rocket box design is a tall, multi-chambered style that has had very good success on Quadra Island (Kiser and Kiser 1999), where several hundred bats (presumed to be M. lucifugus) formed a maternity colony in a rocket box. One rocket box was installed at each generating station. Fig. 23. Pole-mounted maternity houses near John Hart station. Bat houses were located near the main bat access points to Strathcona and Ladore stations, and at John Hart station near the shed used as a night roost. They were installed in sunny locations at least 3 m above ground level, with easy access for bats and out of the way of BC Hydro operations. This resulted in one of the houses being mounted against a building (the Strathcona maternity house) and the remainder mounted on poles. Data from temperature data loggers (see Section 4.6.2) indicate that mounting the houses on buildings may improve the thermal regimes inside the houses.

37 30 Figure 24. Pole mounted rocket box at Strathcona station; close-up of a rocket box. 4.6 MONITORING Use of bat houses In 2002, we used visual searches for bats and/or guano and exit counts to document use of the 6 bat houses installed in February 2002 (Table 10). On 2 occasions, we observed bats in the houses. On several other occasions, guano accumulation was found below both the Strathcona and Ladore maternity houses. This indicates that bats are beginning to find the houses, and the presence of bats in the houses in spite of the lack of exclusion from the stations is encouraging. Table 10. Bat house inspection dates and number of bats observed. X = no bats observed. Date Ladore maternity Strathcona rocket Strathcona maternity John Hart rocket John Hart maternity Jun 10 X Jun 12 X X Jun 16 X Jul 16 X Jul 20 2 bats X 1 bat Aug 1 X X X X

38 Temperature monitoring To determine if the bat houses at the stations could provide similar environments to other available roost sites, we recorded temperatures inside roosts where we know bats successfully raise their young. For comparison, we also recorded the ambient temperature outside in the shade. This data provided a target temperature range for bat houses constructed for habitat enhancement and mitigation. Three of the houses (John Hart rocket box, Ladore nursery house, and Strathcona rocket box) had temperature monitors in them. One of the rocket box temperature loggers was stolen from the site during the summer. Data from the generating stations illustrates why they are such attractive roost sites, with stable high temperatures (Table 11). The generating stations were much warmer than the 2 other human-made roost structures that we sampled (bridge maternity roost and bridge night roost) (Fig 25). Table 11. Average daily temperature and daily variation in temperature of 4 roost sites and of the ambient temperature outside in the shade. Site Average temperature Daily (24 hour) variation in temperature mean ± s.d. 1 ( C) mean ( C) Generating station: Strathcona 25.5 ± Generating station: Ladore 22.9 ± Bat house: maternity style 19.5 ± Bat house: rocket box style 17.5 ± Concrete bridge night roost 17.5 ± Wooden bridge maternity roost 14.8 ± Ambient 14.0 ± s.d. = standard deviation Bat houses were warmer than ambient temperatures, and in fact warmer than a known maternity site (the wooden bridge colony). Bat houses can therefore achieve temperatures warmer than ambient, and thus can be beneficial to reproductive females. However, temperatures in the houses fluctuate widely, reflecting the houses small size and sensitivity to the elements, such as wind and direct sunlight. The fluctuating

39 32 temperatures mean that the houses likely do not currently provide ideal conditions for maternity colonies. One way to make the houses potentially more attractive to maternity colonies is to place them high up against a sunny, protected wall, to provide thermal mass and reduce temperature fluctuations. We recommend ongoing temperature monitoring, to investigate whether the pole-mounted houses should be moved to sunny sites on buildings, where the concrete walls will provide a more stable thermal environment. Fig. 25 represents the temperature regime in sites selected by bats. We made every attempt to place the data loggers where they could be retrieved without unduly disturbing the bats. Consequently, the recorded temperatures in Table 11 and Fig. 25 are conservative, because clustered bats can themselves cause an increase in temperature due to their body heat. This heat source can raise a roost temperature by up to 10 C (Altringham 1996) maximum daily average minimum Temperature (*C) Generating station: Strathcona Generating station: Ladore Bat house : Ladore maternity style Bat house: Strathcona rocket box style Wooden bridge maternity roost Concrete bridge night roost Ambient Figure 25. Average, minimum, and maximum temperatures recorded for colony sites, bat houses, and the ambient air, summer 2001 and Monitoring bat activity levels across the watershed We implemented a monitoring plan for bat activity levels around the watershed. We used bat detectors to record activity levels during 3 sampling sessions at each of 32 sites around the watershed (Appendices 2 and 5). Samples from each site were extremely variable, a common feature with bat activity data (Hayes 1997). Nevertheless, the results provide baseline data for any future long-term population monitoring efforts (Table 12).

40 33 Table 12. Summary of baseline data from acoustic monitoring at 32 sites in the Campbell watershed. Sites were stratified as 1) at generating stations, 2) within 5 km of a station, or 3) greater than 5 km from a station. Stratum n Calls (passes and buzzes) / minute mean standard error min max The original concept of the detector monitoring program was to record 2 samples both before (2001) and after (2002) exclusion operations at Strathcona and Ladore stations. Neither station was successfully sealed during winter 2001, and the 2002 session became a third sample of baseline data. The situation was further complicated by maintenance work occurring at Strathcona station during July This work displaced much of the maternity colony from the station. There is, however, no evidence of the work impacting bat activity at the nearest detector sites. 4.7 EXTENSION/COMMUNICATION We have been involved, through various avenues, in raising awareness about this bat project and in making the results available to a range of potential end users. These efforts include: Extensive work with volunteers, including Campbell River and Comox/ Courtenay locals, Ministry of Water, Land, and Air Protection and Ministry of Forests researchers and staff, North Island College students, and Jason McNair from BC Hydro, Burnaby, Communication about the project with TimberWest, BC Parks head office, and Strathcona Park staff at their headquarters on Buttle Lake, Evening presentation and bat viewing walk for BC Parks, Miracle Beach Interpretive Program, Article in the Campbell River Mirror, 2001, Completion and distribution of monthly updates during summer season, 2001,

41 34 Guest lecture about the Campbell River bat project for a 3 rd year Ecology class at North Island College, Courtenay, BC, Poster presentation (Hydroelectric development and roost sites the Campbell River story) at the 31 st annual North American Bat Research Symposium in Victoria BC, October 2001, Completion and distribution of Year 1 progress report (Kellner, M. and S. Rasheed Campbell River watershed bat habitat enhancement and inventory project year 1 progress report. B.C. Hydro Bridge-Coastal Fish and Wildlife Restoration Program, Burnaby, B.C. 35pp.) to BCRP, John Hart staff, Ministry of Water, Land and Air Protection, Ministry of Forests, TimberWest, and BC Parks, Review of project with John Hart (JH) staff at JH morning meeting June 6, 2002, Interpretive signage developed and installed at Strathcona Dam campground (Appendix 6), Communication and sharing of data relating to bat houses with the North American Bat House Research Project, and Draft papers prepared for submission to scientific journals. Working titles: Relative activity of bats in natural and reservoir-affected foraging sites in coastal British Columbia, and Roost selection by bats in a highly modified forested landscape.

42 35 5. DISCUSSION 5.1 BAT SPECIES AND DISTRIBUTION We captured 4 species of bats (or 5, if both M. keenii and M. evotis were captured) and potentially recorded 3 other species, of the 10 species that could occur in the Campbell watershed (Nagorsen and Brigham 1993). The lower watershed, which is highly disturbed, is heavily populated by very adaptable species (M. lucifugus and M. yumanensis). Of the 4 species we caught, 1 of them (M. californicus) was only captured in the upper watershed. The absence of this bat from our captures in the lower watershed is unusual because we had a higher sampling effort in the lower watershed. Although there is greater habitat disturbance in the lower watershed, M. californicus is reportedly adaptable, occurs across a range of habitats, and will use buildings and bridges as roosts (Nagorsen and Brigham 1993). M. californicus have been captured in other low elevation pristine and disturbed sites on Vancouver Island, including in Clayoquot sound (van den Driessche et al 1999), the Nimpkish valley (Grindal 1998), and the Davie valley (Kellner 1999). The 2 species for which we obtained no information were Corynorhinus townsendii (Towsend s Big-eared bat) and M. volans (Long-legged Myotis). C townsendii has been identified on the east coast of Vancouver Island, both to the north (Sayward) and south (Courtenay) of Campbell River (V. Craig, pers. comm.). M. volans has also been recorded at locations near Campbell River, including to the north (Grindal 1998, Kellner 1999), and to the east on Quadra Island (S. Holroyd, pers. comm.). It is probable that both of these species also occur in the Campbell watershed. We had a much higher capture rate in the upper watershed than in the lower watershed. Although it is clear that bats were easier to catch in the upper watershed, interpreting this result is difficult. It is possible that there is a higher density of bats in the upper watershed. However, there are numerous other explanations, including the fact that habitat differences may influence capture success. In the upper watershed, there are steep, very rocky dry hillsides, and watercourses are confined to the relatively narrow valley bottom. This means that foraging sites are concentrated, as opposed to being distributed across the landscape as they are in the gentle terrain of the lower watershed.

43 36 Thus, the steeper terrain (and consequently limited foraging sites) in the upper watershed may force bats to use the few valley-bottom sites we netted at, and result in a higher capture rate regardless of bat density. Nevertheless, we did capture 1 more species (M. californicus) in the upper than in the lower watershed. If this increase in diversity is real and not a sampling artifact, it could be the result of higher habitat and structural diversity of the upper watershed, with its numerous cliffs and old-growth forest stands ROOSTING HABITAT Roost trees were larger-than-average snags of bark class 2-5, in open forest stands, characteristics that are thought to promote access and solar heating. Our results are in agreement with other findings for roost selection by bats in western North America (Betts 1996, Vonhof and Barclay 1996, Brigham et al. 1997, Ormsbee and McComb 1998, Kellner 1999, Waldien et al. 2000), and stress the importance of maintaining large snags in a managed landscape. 5.3 HYDROELECTRIC DEVELOPMENT AND BAT HABITAT Assessing the impact of hydroelectric development on bats can be divided into impacts on foraging and on roosting habitat. Forest bats forage in wetlands, riparian areas, and floodplain habitats (Grindal et al. 1999, Waldien and Hayes 2001). These habitat types were inundated to develop the present-day large reservoirs with fluctuating water levels. While the reservoirs provide a degree of foraging habitat, we found that the use of reservoirs was low, compared to the level of use of natural ponds. Given that the original impoundment in the watershed led to the loss of 4517 ha of natural lakes, 936 ha of wetland, and 417 ha of creekside riparian habitat, this translates to a significant loss of bat foraging habitat. The reduced use of reservoirs compared to natural ponds is probably due to a decreased prey base available at reservoirs, because bat activity is correlated with insect abundance (Hayes 1997, Grindal and Brigham 1999). Reservoir edges are generally steep and non-vegetated, due to the fluctuating water levels (Lower Campbell: fluctuations of 15.2 m, Upper Campbell: 8 m), and aquatic productivity is low, relative to

44 37 pre-impoundment conditions (BCFWRP 2000). These conditions are very different from the conditions at natural riparian and wetland habitats. Natural roosting habitat, such as the south-facing rock bluffs and large snags used by bats in our study, would also have been lost during impoundment. Low elevation and riparian habitats are typically nutrient-rich (Meidinger and Pojar 1991). These sites support the growth of large trees such as those selected by bats for maternity roost sites. Selection of scarce remnant old trees and stands for roosting suggests that these habitat elements are very important in this landscape, and offer superior roost sites relative to the surrounding second-growth matrix. Similar use of remnant old-growth patches has been reported in coastal California (Zielinski and Gellman 1999). Our findings on the use by bats of natural ponds and remnant forest patches stress the importance of maintaining and enhancing critical habitat and habitat elements in order to continue to support bat populations in the lower watershed. RECOMMENDATION 1: Enhancement of bat foraging habitat could be accomplished by developing wetlands and riparian habitats along reservoir edges, to mimic natural wetland / riparian conditions. This would require stabilizing water levels in artificial ponds and encouraging revegetation of shorelines. Such habitat enhancement projects are already defined as Wildlife Restoration Objectives 1 and 2 (BCFWRP 2000) for the Campbell watershed. We have confirmed that bats are among the many species that likely would benefit from these types of rehabilitation efforts. Possible target areas for enhancement are shallow sites near Buttle Campground and the lagoon located at the Strathcona Dam Campground. Protection of existing natural wetlands and lakes should also be promoted. RECOMMENDATION 2: In order to protect sparse natural roosting habitat in the lower watershed, efforts should be made to protect remnant forest patches and individual veteran trees and snags. The use of larger-than-average snags in second growth forests also suggests that there is potential to create roosting habitat in these forests. Creating snags and gaps in thinned, mature forests should provide potential roost sites. Any snag creation projects should aim to provide groups of snags rather than isolated structures, as research here and elsewhere (Lewis 1995, Kellner 1999, Waldien et al. 2000)

45 38 indicates that bats often switch between several roosts, and subsequent roosts are generally within 1 km. Human-made structures also provided roost sites for colonies of 2 species. We located M. lucifugus and M. yumanensis roosting in bat houses, loose cedar shakes on roofs and walls, cedar siding, under wooden and concrete bridges, and in generating stations. Further artificial habitat enhancement may benefit these adaptable species. However, we did not observe any other species in the human structures that we examined. We observed use by bats of 2 of the 6 bat houses we installed. Bats may use bat houses almost immediately, or it may take several years before the houses become occupied (Tuttle and Hensley 2000). Lack of occupancy immediately after installation cannot be used as a performance measure to assess the success of artificial roosts. Occupancy in bat houses often increases after the houses have been up for several years (Tuttle and Hensley 2000), and we expect the numbers of bats using the houses to increase over time. Whether the conditions in the houses will be suitable for maternity colonies remains to be seen, and should be determined by future monitoring. RECOMMENDATION 3: We recommend that the 6 houses be monitored by visual inspections each summer, to document use over several years. Repeated temperature monitoring (as in this study) would offer further insight into the possible need to move pole-mounted houses to building-mounted sites. Bat houses installed by TimberWest on private property in the lower watershed immediately had high occupancy by maternity colonies (Kellner pers. obs). This can likely be attributed to the complete removal of the Camp 8 buildings that had formerly housed large colonies of bats (Kellner and Rasheed 2001). The positive response of bats to these houses indicates that habitat enhancement using bat houses may be very effective in the lower watershed. Further habitat enhancement work with artificial roosts should be pursued. Bat houses properly positioned and located around the lower watershed will help provide roosting habitat for M. lucifugus and M. yumanensis, and potentially for other species as well.

46 BATS IN GENERATING STATIONS The initial impetus for bat work in the Campbell watershed was the large number of bats roosting in the generating stations. Successive attempts to exclude the bats from their specific roost sites and from the buildings have reduced the numbers of bats inside. However, bats were never excluded completely from the stations. Guano accumulation continues to be a problem to human workers, and a number of bat fatalities each year, caused through entrapment and desiccation or crushing, may certainly pose a problem for populations of these normally long-lived species. At present, the stations continue to be occupied by maternity colonies of up to 100 M. lucifugus and M. yumanensis bats. Continued attempts could be made to exclude all bats from the stations. If this course of action is taken by BC Hydro, we would again recommend the sealing of all access points outlined in our Year 1 progress report, completing all work before spring 2003, and monitoring the stations in summer 2003 to determine if/where more effort is required. However, given the current colony sizes and the amount of work required for a complete exclusion, we suggest a modified approach for managing the situation, including maintaining the exclusion work done to-date and educating workers. RECOMMENDATION 4: Given the shortage of resources and the fact that the age and condition of the buildings necessitate an ongoing effort to seal access points and maintain these seals, we recommend a strategy of accepting a low number of bats inside the stations, while trying to stabilize or decrease this number over time. If this route is taken, we recommend education for all people working in stations with bats. Specifically, we recommend attempting to reduce bat numbers, through: Maintaining sealing efforts to-date, particularly maintaining the seals around sliding roof covers and carefully closing these covers completely, When possible, proceeding with sealing efforts in winter 2003 and subsequent winters. This will further reduce access and bat numbers in the stations, and Changing conditions in the station during the reproductive period of May through July. This may be accomplished by increased ventilation, such as keeping doors and windows open, and scheduling maintenance work for early summer (beginning early

47 40 June), as having the sliding covers open obviously makes the generating station unsuitable. Workers should regularly and routinely receive information regarding bats and worker health. Workers should be informed that: There is no health hazard from the bats as long as they are not handled. The infection rate for rabies in bats is low, at 0.1 % (Nagorsen and Brigham 1993). Even if a bat carries rabies, this viral disease is usually transmitted through a bite, which can easily be avoided. If a live bat must be handled, wear thick leather gloves. If handling dead bats, pick them up in a plastic bag. If contact with a bat does occur, the person should seek medical attention. There are no health problems associated with regular guano clean up in British Columbia. Histoplasmosis, a fungal disease associated with guano deposits in moist, hot environments (such as the tropics or the southeastern U.S.), is extremely rare in coastal British Columbia (Nagorsen and Brigham 1993). In summer months, guano cleanups should be scheduled before work is to be done in the stations, to provide a clean workspace for maintenance crews, and Bats, while they may be startling, are harmless creatures who usually are trying to stay away from humans, either by hiding in crevices or by trying to find an exit from an enclosed space (such as a generating station). Success in managing bats in generating stations is directly related to the effort put in to solving the problem. We have attempted to provide the information necessary to either continue to work on exclusion or to accommodate the presence of colonies in the stations. Hopefully, BC Hydro staff can achieve a satisfactory and workable solution. 6. SUMMARY We met most of the objectives proposed for the Campbell River Bat Habitat Enhancement project. The Campbell River watershed contains several bat species, in spite of significant roosting and foraging habitat alteration as a result of hydroelectric

48 41 development. Given that we do not have pre-impoundment population or habitat inventory, we cannot accurately estimate the extent of the effect on bat populations. However, we found bats foraging and roosting in habitat that was similar to habitat that was certainly flooded when the reservoirs were created. To mitigate this habitat loss we make several recommendations including: RECOMMENDATION 1 - FORAGING HABITAT ENHANCEMENT: Develop wetlands and riparian habitat around reservoirs, by stabilizing water levels in artificial ponds and encouraging the revegetation of shorelines with native plant species, in order to mimic natural conditions. RECOMMENDATION 2 ROOSTING HABITAT ENHANCEMENT: Protect remnant forest patches and individual veteran trees and snags, both in old-growth and second-growth forest stands. Create canopy gaps and groups of snags, rather than isolated structures, in thinned, mature forests. RECOMMENDATION 3 - MONITORING: Monitor the 6 installed bat houses by visual inspections each summer, to document use over several years. Regarding bats in the stations, we make the following recommendation: RECOMMENDATION 4 - EXCLUSION: After two years of study of both the bat populations and BCH operations, we felt that complete exclusion of bats from the generating stations was not viable. Instead, accept a low number of bats inside the stations, continue to maintain and expand on exclusion work accomplished to-date, and provide sound worker training and education for working around the bats.

49 42 7. LITERATURE CITED Altringham, J.D Bats: biology and behaviour. Oxford University Press Inc., New York. 262 pp. Barclay, R.M.R Population structure of temperate zone insectivorous bats in relation to foraging behaviour and energy demand. J. Animal Ecol. 60: Betts, B.J Roosting behaviour of Silver-haired (Lasionycteris noctivagans) and Big Brown Bats (Eptesicus fuscus) in northeastern Oregon. Pp in R.M.R. Barclay and R.M. Brigham, eds. Bats and Forest Symposium, October 19-21, 1995, Victoria, British Columbia, Canada. Research Branch, B.C. Min. Forests. Victoria, B.C. work. Pap. 23/ pp. Betts, B.J Effects of interindividual variation in echolocation calls on identification of Big Brown and Silver-haired bats. J. Wildl. Manage. 62: Bridge-Coastal Fish and Wildlife Restoration Program Strategic Plan. Volume 2: Campbell River Watershed. 47 pp. Brigham, R.M., M.J. Vonhof, R.M.R. Barclay, and J.C. Gwilliam Roosting behaviour and roost-site preferences of forest-dwelling California bats (Myotis californicus). J. Mammal. 78: Firman, M.C., C. Godwin, and R.M.R. Barclay Bat fauna survey of West Shuswap and South Thompson River Region, British Columbia. Unpublished report prepared for B.C. Min. Environment, Lands and Parks. Victoria, B.C. Green, R.N. and K. Klinka Field guide to site identification and interpretation for the Vancouver Forest region. Land Management Handbook no. 28. Research Program, B.C. Min. Forests. Victoria, B.C. 285 pp. Grindal, S.D Bat survey of BC Hydro s Ladore and Strathcona Generating Stations. Report prepared for B.C. Hydro, Environmental Services. Burnaby, B.C. 26 pp. Grindal, S. D Habitat use by bats in second- and old-growth stands in the Nimpkish valley, Vancouver Island. Northwest Sci. 72: Grindal, S. D. and Brigham, R. M Impacts of forest harvesting on habitat use by foraging insectivorous bats at different spatial scales. Ecoscience 6: Grindal, S.D., J.L. Morissette, and R.M. Brigham Concentration of bat activity in riparian habitats over an elevational gradient. Can. J. Zool. 77:

50 43 Hamilton, I.M. and R.M.R. Barclay Patterns of daily torpor and day-roost selection by male and female big brown bats (Eptesicus fuscus). Can. J. Zool. 72: Hayes, J.P Temporal variation in activity of bats and the design of echolocationmonitoring studies. J. Mammal. 78: Holroyd, S.L Influences of some extrinsic and intrinsic factors on reproduction by big brown bats (Eptesicus fuscus) in southeastern Alberta. M.Sc. Thesis. University of Calgary. Holroyd, S.L., R.M.R. Barclay, L.M. Merk and R.M. Brigham A survey of the bat fauna of the dry interior of British Columbia. Ministry of Environment, Lands and Parks, Victoria, B.C. Wildlife Working Report No. WR pp. Kellner, A.M.E Activity and roost selection of bats in coastal forests on northern Vancouver Island. M.Sc. Thesis. Simon Fraser University, B.C. 95 pp. Kellner, M. and S. Rasheed Campbell River watershed bat habitat enhancement and inventory project year 1 progress report. Report prepared for B.C. Hydro Bridge-Coastal Fish and Wildlife Restoration Program, Burnaby, B.C. 35 pp. Kiser, M. and S. Kiser Bat houses and exclusion in British Columbia. Bat Res. News 7:3-4. Kunz, T.H Roosting ecology of bats. Pp in T.H. Kunz, ed. Ecology of Bats. Plenum Press, New York. 425 pp. Kurta, A. K.J. Williams, and R. Mies Ecological, behavioural, and thermal observations of a peripheral population of Indiana bats (Myotis sodalis). Pp in R.M.R. Barclay and R.M. Brigham, eds. Bats and Forest Symposium, October 19-21, 1995, Victoria, British Columbia, Canada. Research Branch, B.C. Min. Forests. Victoria, B.C. work. Pap. 23/ pp. Lewis, S. E Roost fidelity of bats: a review. J. Mammal. 76: Meidinger, D., and J. Pojar (compilers and editors) Ecosystems of British Columbia. B.C. Min. For. Special Report Series No pp. Nagorsen, D.W. and R.M. Brigham Bats of British Columbia. Royal British Columbia Museum Handbook. UBC Press, Vancouver, B.C. 174 pp. O'Farrell, M. J., Miller, B. W. and Gannon, W. L Qualitative identification of freeflying bats using the Anabat detector. J. Mammal. 80: Ormsbee, P.C Characteristics, use, and distribution of day roosts selected by female Myotis volans (Long-legged Myotis) in forested habitat of the central Oregon Cascades. pp in R.M.R. Barclay and R.M. Brigham, eds. Bats

51 44 and Forest Symposium, October 19-21, 1995, Victoria, British Columbia, Canada. Research Branch, B.C. Min. Forests. Victoria, B.C. work. Pap. 23/ pp. Ormsbee, P. C. and W.C. McComb Selection of day roosts by female long-legged Myotis in the central Oregon Cascade range. J. Wildl. Mange. 62: Perlmeter, S.I Bats and bridges: patterns of night roost activity in the Willamette national forest. Pp in R.M.R. Barclay and R.M. Brigham, eds. Bats and Forest Symposium, October 19-21, 1995, Victoria, British Columbia, Canada. Research Branch, B.C. Min. Forests. Victoria, B.C. work. Pap. 23/ pp. Rasheed, S.A. and S.L. Holroyd Roosting habitat assessment and inventory of bats in the MICA Wildlife Compensation Area. Report prepared for B.C. Hydro, B.C. Min. Environment, Lands and Parks, and Parks, Canada. Pandion Ecological Research Ltd. 77 pp. Resources Inventory Branch, ELP Inventory methods for bats: Standards for components of British Columbia s Biodiversity No.20. Resources Inventory Committee. 47pp. Rydell, J., A. Entwistle, and P.A. Racey Timing of foraging flights of three species of bats in relation to insect activity and predation risk. Oikos 76: Thomas, D.W. and LaVal, R.K Survey and census methods. Pp in T.H. Kunz, ed. Ecological and behavioural methods for the study of bats. Washington, D.C. Smithsonian Institute Press. Tuttle, M.D. and D.L. Hensley The Bat House Builder s Handbook. Bat Conservation International. 35 pp. van den Driessche, R., M. Mather, and T. Chatwin Habitat selection by bats in temperate old-growth forests, Clayoquot Sound, British Columbia. Report prepared for B.C. Min. Environment. Nanaimo, B.C. 29pp. Vonhof, M.J Roost-site preferences of Big Brown Bats (Eptesicus fuscus) and Silver-haired Bats (Lasionycteris noctivagans) in the Pend d'oreille Valley in southern British Columbia. Pp in R.M.R. Barclay and R.M. Brigham, eds. Bats and Forest Symposium, October 19-21, 1995, Victoria, British Columbia, Canada. Research Branch, B.C. Min. Forests. Victoria, B.C. work. Pap. 23/ pp. Vonhof, M.J. and R.M.R. Barclay Roost-site selection and roosting ecology of forest-dwelling bats in southern British Columbia. Can. J. Zool. 74:

52 45 Waldien, D. L., Hayes, J. P. and Arnett, E. B Day-roosts of female long-eared Myotis in western Oregon. J. Wildl. Manage. 64: Waldien, D.L. and J.P. Hayes Activity areas of female long-eared Myotis in coniferous forests in western Oregon. Northwest Sci. 75: Wilkinson, L.C., P.F.J. Garcia & R.M.R. Barclay Bat survey of the Liard River watershed in northern British Columbia. Unpublished report submitted to B.C. Min. Environment, Lands and Parks. Victoria, B.C. 40 pp. Zielinski, W. J. and S.T. Gellman Bat use of remnant old-growth redwood stands. Cons. Biol. 13:

53 46 APPENDIX 1. ROOST SITES PHOTO GALLERY Figure A1.1. Examples of roost trees used by radio-tagged bats. Note the exfoliating bark, sun exposure, and open canopy. Figure A1.2. South-facing, rocky outcrops also provide roost sites. Figure A1.3. Cracks in this exposed limestone housed a maternity colony of M. lucifugus.

54 47 Figure A1.4. Inspecting bats at a night roost under a concrete bridge. Figure A1.5. Bats clustered on the warm concrete. Figure A1.6. A wooden bridge used as a maternity colony by M. lucifugus. Figure A1.7. Sunny cedar-shake roofs and walls also provide roosting habitat for bats.