Vancouver Island Marmot - Buttle Lake Supplementation Project Project # 14.W.CBR.01

(Photo by Mike Lester)

Prepared for the Marmot Recovery Foundation by: Cheyney Jackson, MSc, Marmot Recovery Foundation Sigurd Johnson, BNRP, BC Conservation Foundation Chris White, RB Tech, Fish and Wildlife Compensation Program Don Doyle, RPBio, Marmot Recovery Foundation

Prepared with the financial support of the Fish and Wildlife Compensation Program

EXECUTIVE SUMMARY

The Vancouver Island marmot (Marmota vancouverensis, Swarth 1911) is one of only five endemic land mammals in Canada (Nagorsen 2004), and was first listed as an endangered species in 1979 (Committee on the Status of Endangered Wildlife in Canada; Munro 1979). The National Recovery Strategies for the Vancouver Island marmot (Janz et al. 1994, 2000; Vancouver Island Marmot Recovery Team 2008) recommended captive breeding and reintroduction as the best hope for recovering the species, and described an ultimate goal of three geographically distinct metapopulations, each comprised of 150-200 marmots. Since 2003, the captive-breeding program has produced 445 marmots that were released to the wild. Initially, releases were conducted only to support the Nanaimo Lakes metapopulation, located at the southern extent of historic marmot habitat. Recovery efforts have achieved some success at re-establishing a metapopulation in this region. Starting in 2007, in order to expand the distribution into a greater portion of the marmot’s historic range, marmots were released to several mountains to the east and west sides of Buttle Lake; however, re-establishment of metapopulations outside the southern core were more challenging, likely due to the lack of extant colonies. In 2011, the Recovery Team approved testing of new release techniques in hopes of improving overwinter survival of released marmots. This project addressed that problem through use of multiple strategies aimed at accelerating species’ recovery. Project trials included (1) the introduction of spring supplemental feeding as a possible stimulant for successful breeding, (2) evaluation of wild-born marmots as a superior source of marmots for translocation to the Buttle Lakes area, and (3) evaluation of efficiency gained by pre-conditioning captive-bred marmots in a wild hibernation at Mt. Washington. Feeder trials were initiated in 2011, and translocation trials began in 2012. In 2014, we installed 10 feeders on Mt. Washington and 4 feeders in the Buttle Lake area. We released 15 captive-bred facility marmots, 8 captive-bred pre-conditioned marmots, and 12 wild-born marmots to seven sites near Buttle Lake. We counted 11 pups born on Mt. Washington, as well as documenting the first known litters at Flower Ridge, Marble Meadows, and Sunrise Lake, and the fourth known litter born at Tibetan Plateau.

-ii-

Table of Contents

EXECUTIVE SUMMARY ...... ii List of Tables ...... iv List of Figures ...... iv List of Appendices ...... v 1. INTRODUCTION ...... 1 2. GOALS AND OBJECTIVES ...... 2 3. STUDY AREA ...... 2 4. METHODS ...... 3 4.1 New reintroduction strategies ...... 3 4.1.1. Supplemental spring feeding...... 3 4.1.2 Translocation trials...... 4 4.2 Release Site Preparation ...... 4 4.3 Release Protocols ...... 5 4.4 Monitoring ...... 5 4.4.1 Radiotelemetry ...... 5 4.4.2 Measures of Success ...... 6 4.5.1. Translocation trials...... 7 4.5.2. Mortality inferences ...... 8 5. RESULTS ...... 8 5.1 Feeder trials and remote camera footage ...... 8 5.2 Translocation trials...... 9 5.2.1 – Detectability ...... 9 5.2.2 – Pre-conditioning at Mt. Washington ...... 9 5.2.3 – Marmots translocated in 2012 ...... 9 5.2.4 – Marmots translocated in 2013 ...... 9 5.2.5 – Marmots translocated in 2014 ...... 10 5.3 Mortalities ...... 10 5.4 Dispersal ...... 10 5.5 Reproduction and Population Size ...... 10 5.6 Statistical Analyses ...... 11 5.6.1. Translocation trials...... 11 5.6.2. Mortality inferences ...... 11 6. DISCUSSION ...... 12 6.1 Supplemental spring feeding...... 12 6.2 Survival by experience level ...... 13 6.3 Reproduction ...... 13 6.4 Detection ...... 14 6.5 Distribution and Dispersal ...... 14 7. RECOMMENDATIONS ...... 15 Acknowledgements ...... 17 REFERENCES ...... 18 FIGURES ...... 23 APPENDICES ...... 34

-iii-

List of Tables Table 1. In 2014, we installed 10 feeders at Mt. Washington and 4 feeders in the Buttle Lake area...... 21 Table 2. There were four different experience treatments for marmots released in the Buttle Lake region translocation trials (2012-2014)...... 21 Table 3. Thirty-two marmots were translocated to seven sites in the Buttle Lake area in 2014. ... 21 Table 4.Population estimates for the Buttle Lake area at the end of 2014...... 22

List of Figures Figure 1. Study area for feeder (2011-14) and translocation (2012-14) trials for the Vancouver Island marmot in the Buttle Lake area...... 23 Figure 2. Pup production by year at the wild colony on Mt. Washington...... 24 Figure 3. Prior to feeder trials, overwinter survival by captive-bred marmots released at Mount Washington averaged 64.3% (2003-2010, black line). We estimated 62.5%, 72.7%, and 84.6% overwinter survival by marmots being pre-conditioned in 2011-12, 2012-13, and 2013-14, respectively...... 24 Figure 4. Snow water accumulation at the Wolf River snow station in Strathcona Provincial Park, on Vancouver Island, BC. The winter of 2013-14 (dark blue) was characterized by much lower than average snow levels, limited snow accumulation, and rapid spring melt...... 25 Figure 5. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2012...... 26 Figure 6. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2013...... 27 Figure 7. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2014...... 28 Figure 8. A map showing some of the more notable dispersals by marmots released in the Buttle Lake area (2007-2014)...... 29 Figure 9. In 2014, we documented reproduction at five colonies in the Buttle Lake area...... 30 Figure 10. Regional population estimates for the Nanaimo Lakes, Forbidden Plateau and Western Strathcona regions...... 31 Figure 11. Estimated survival over a marmot’s first active season in the Buttle Lake area, grouped by release year within experience level...... 32 Figure 12. Estimated survival over a marmot’s first winter in the Buttle Lake area, grouped by release year within experience level...... 32 Figure 13. Estimated survival over a marmot’s first 14 months in the Buttle Lake area, grouped by release year within experience level...... 33 Figure 14. Estimated survival over a marmot’s first 26 months in the Buttle Lake area, grouped by release year within experience level...... 33

-iv-

List of Appendices Appendix I. Performance Measures – Actual Outcomes ...... 34 Appendix II. Confirmation of FWCP Recognition ...... 36 Appendix III. Timeline of activities at Mt. Washington...... 37 Appendix IV. Remote camera footage from the Buttle Lake area...... 38 Appendix V. Guidelines used to estimate population size for the Buttle Lake area in 2014...... 43 Appendix VI. Marmot summaries for release sites used in 2012, 2013 and 2014 translocation trials...... 44 Appendix VII. Photos from the 2014 field season...... 46

-v-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

1. INTRODUCTION The Vancouver Island marmot (Marmota vancouverensis) is a large ground squirrel endemic to Vancouver Island, in British Columbia, Canada. In 1979, the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) designated the Vancouver Island marmot as a nationally endangered species. This was confirmed by the BC Wildlife Act in 1980 (Munro et al. 1985), the International Union for the Conservation of Nature in 1982 (Thornback and Jenkins 1982), and the United States Endangered Species Act in 1984 (Vancouver Island Marmot Recovery Team 2008). In 1994, the National Recovery Strategy (Janz et al. 1994) called for increased inventory and research on the species. By 1997, the southern population of Vancouver Island marmots was still in decline and the Buttle Lake metapopulations were represented by a single, isolated colony on Mount Washington (Bryant 1998). Paleoecological evidence suggests that a shift in historic habitat from open heather and meadow to forest may have reduced availability of suitable habitat (Hebda et al. 2004) and increased predator success (Milko 1984). Recent population declines have been attributed primarily to predation (Aaltonen et al. 2009; Bryant and Page 2005, Bryant 2000). The reasons for population decline in Strathcona Provincial Park remain unclear, but were probably related to a combination of pressures from predation, weather, disease, landscape-level change, and colony isolation (Janz et al. 1994). The impoundment of Buttle Lake is hypothesized to have impeded marmot dispersal between colonies because of a lack of landscape connectivity (Janz et al. 2000). Prior to 2007, the last marmots reported in Strathcona Park were solitary individuals seen on Mount Albert Edward in 1992, and at the Westmin mine site in 1995 (A. Bryant, unpublished data). Following the recommendations laid out in the initial National Recovery Strategy for the Vancouver Island Marmot (Janz et al. 1994), a captive-breeding program was established in 1997. The Recovery Strategies describe an ultimate goal of two to three geographically distinct metapopulations, each comprised of 150-200 marmots (Janz et al. 1994, 2000; Vancouver Island Marmot Recovery Team 2008). Releases have been carried out since 2003 to increase the wild population and expand the geographic distribution of extant colonies. Initial reintroduction efforts focused on the southern core of historic habitat near the Nanaimo Lakes region, north of Lake Cowichan and south of Port Alberni. After several years of releases and shepherding efforts by field crew, a metapopulation of 150-200 marmots has been re-established in this area, and we are presently evaluating its viability. The distribution of Vancouver Island marmots in just a single metapopulation leaves the species at significant risk of demographic stochasticity and catastrophic events that could erode the recovery achieved to date. Multiple self-sustaining metapopulations are necessary to ensure the species’ persistence (VI Marmot Recovery Team 2008). Therefore, in 2007, the recovery project began releasing captive-bred marmots in historic habitat to the east and west sides of Buttle Lake. Reintroductions have proven more challenging in the Buttle Lake area than in the south, which had extant colonies to provide additional support for metapopulation recovery. An evaluation of reintroduction success for captive-bred Vancouver Island marmots determined that overwinter survival was the most limiting factor for marmots surviving their first post-release year (Jackson 2012). Once captive-bred marmots had survived their first wild hibernation, their predicted overwinter survival rate was equivalent to that of wild-born marmots (Jackson 2012). In 2012, with Recovery Team approval, we adapted our reintroduction strategy to introduce and evaluate several new techniques aimed at increasing overwinter survival by marmots released to the Buttle Lake area. These techniques included (1) the introduction of spring supplemental feeding as a possible stimulant for successful breeding, (2) evaluation of wild-born marmots as a superior source of marmots for translocation to the Buttle Lake area, and (3) evaluation of efficiency

-1-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

gained by pre-conditioning captive-bred marmots in a wild hibernation at Mt. Washington. This report describes our methods and results for 2014, the final year of this three-year project.

2. GOALS AND OBJECTIVES

 Assess overwinter survival by pre-conditioned marmots at Mt. Washington.  Assess overwinter survival by marmots moved to the Buttle Lake area in the 2012-13 translocation trials.  Continue spring supplemental feeding. Install feeders at all emergence holes on Mt. Washington and at select sites in the Buttle Lake area.  At Mt. Washington, capture and implant yearlings and replace expired transmitters in older marmots.  Continue translocation trials. Translocate marmots to support existing fledgling colonies at Castlecrag, Drinkwater, Flower Ridge, Greig Ridge, Henshaw, Marble Meadows, and Sunrise Lake.  Monitor active season survival by marmots released to the Buttle Lake area in 2012-2014.  Monitor movements of marmots. When possible, locate hibernacula to aid verification of emergence in 2015.  Verify reproduction and survival by previously released or translocated breeding-aged females and their offspring.  Ongoing collection of field data to monitor marmot demographics and to answer research questions.

3. STUDY AREA Each release site (Figure 1) reflects typical Vancouver Island marmot habitat, consisting of wet sedge meadows in combination with talus boulder complexes in the alpine tundra (ATc) and subalpine zones (MHmm2), between 900 metres and 1400 metres in elevation (Bryant 1998). Release sites are located within the historically occupied range of the Vancouver Island marmot (see Janz et al. 1994, 2000; Vancouver Island Marmot Recovery Team 2008), and include the following: (i) Castlecrag (UTM: 10U 327898mE 5509715mN) is located at the south end of Forbidden Plateau, near the headwaters of the Cruickshank River. It sits in close proximity to Sunrise Lake, which means that it may serve as a midpoint for marmots dispersing from Mt. Washington, Sunrise Lake, or Henshaw. It is characterized by a large expanse of lush meadows and rock slides on the southern face of the mountain, which makes it ideal habitat for marmots. We installed a spring feeder near a successful hibernaculum used in 2011-12, 2012-13, and 2013-14. We released captive-bred marmots to Castlecrag in 2010 and 2011 and translocated marmots to the hill in 2013. (ii) Drinkwater basin (UTM: 10U 316487mE 5483468mN) is located at the south end of Buttle Lake and the north end of Della Falls, southwest of Flower Ridge. Drinkwater features slide habitat, meadows, boulder complexes and is surrounded by habitat that should support dispersing marmots. We rated Drinkwater highly on our initial 2005 surveys of suitable marmot habitat. We released captive-bred marmots to the Drinkwater in 2009 and 2010, and translocated groups of marmots to the hill in 2012 and 2013. (iii) Flower Ridge (UTM: 10U 319832mE 5485480mN) is at the south end of Buttle Lake near the headwaters of Price Creek. It is located between Drinkwater and Henshaw. The south end of Flower Ridge was chosen as a release site because it features highly suitable habitat, including meadow, talus

-2-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

and cliffs. We released captive-bred marmots to Flower Ridge in 2009, and translocated a group of marmots here in 2013. (iv) Greig Ridge (UTM: 10U 305628mE 5502021mN) is located approximately 10.5 kilometers west of Buttle Lake, within the Phillips Creek drainage that flows into the west side of Buttle Lake. Tibetan and Morrison Spire release sites are to the northeast of Greig Ridge. Greig Ridge includes over six kilometres of alpine/subalpine habitat and features the largest subalpine complex with high marmot habitat suitability in the Buttle Lake area. This site was initially chosen for reintroductions because of existing burrows (c1980). We have released marmots to Greig Ridge every year since 2007, including the translocation of mixed groups of marmots to the hill in 2012 and 2013. (v) Henshaw (UTM: 10U 320795mE 5491691mN) is located at the south east end of Buttle Lake near the headwaters of Henshaw Creek. Henshaw sits to the north of Flower Ridge and southwest of Castlecrag release sites. A broad band of lush alpine meadows and talus slides wrap around the mountain. Surveys completed in 2009 showed large amounts of suitable marmot habitat at this site. We installed a spring feeder at Henshaw in 2012. We released marmots at Henshaw in 2010 and 2011 and translocated a group of marmots here in 2013. (vi) Marble Meadows (UTM 10U 307763mE 5508601mN) is located 7km west of central Buttle Lake, south of Mount McBride, and at the north end of the ridge connecting Greig Ridge, Tibetan, and Morrison Spire. This location was selected as a release site after a marmot released at Greig Ridge in 2012 was detected here in 2013. The habitat, which is east facing and gently sloped, consists of an alpine heather bowl with a mixture of large boulder piles and limestone rock formations. (vii) Sunrise Lake (UTM: 10U 315875mE 5483088mN) is part of Forbidden Plateau. It is located on the north side of Jutland Mountain near the headwaters of Oyster River and Piggott Creek. Sunrise Lake sits northwest of Castlecrag and southwest of Mount Washington. Initially, we chose Sunrise because of its historical relevance - it was one of the last known locations in the park where a marmot was sighted (1981, A. Gibson, pers. comm.). We installed a supplemental feeder at Sunrise Lake in 2012 and 2013. We released marmots at Sunrise in 2010 and 2011 and translocated marmots here in 2013. (viii) Tibetan (UTM: 10U 306568mE 5503614mN) is the first peak on the connecting ridgeline from Greig Ridge to Morrison Spire and Marble Meadows. This site features a mixed boulder and talus slope, with many options for refuge. This site was chosen for its proximity to Greig Ridge and its potential to function as a geographical stepping stone to Morrison Spire. We installed a spring supplemental feeder in both 2012 and 2013. We released marmots to Tibetan in 2009 and 2010, and translocated two marmots here in 2012.

4. METHODS 4.1 New reintroduction strategies 4.1.1. Supplemental spring feeding Literature on reintroduction of species strongly recommends the use of wild-born animals when possible (Fischer and Lindenmayer 2000; Wolf et al. 1996; Griffith et al. 1989; Beck et al. 1994). However, until very recently, there were no wild Vancouver Island marmot colonies that were large enough to provide ‘surplus’ marmots for translocation into the Buttle Lake area without risking adverse demographic effects at local and metapopulation levels (D. Doyle, pers. comm.). Supplemental feeding has been demonstrated to increase adult body condition, percentage lactating, weaning litters, and litter size in Columbian ground squirrels (Karels et al. 2001). Therefore, we installed feeders at Mt. Washington in

-3-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

the springs of 2011-2013, with the idea that providing early, high-quality supplemental food would promote pup production on the hill and that this would make available wild-born candidates for translocation to the Buttle Lake watershed. We constructed feeders from PVC pipe and filled them with Mazuri® leaf-eater biscuits. The feeders at Mt. Washington were re-filled twice a week from the date of installation (early May, 2014) until the snow had melted enough to reveal natural food sources (early June, 2014). We also installed feeders at emergence burrows for captive-bred marmots that were completing their first wild hibernation at Mt. Washington as part of a pre-conditioning experiment (see 4.1.2). We installed larger feeders at select sites around Buttle Lake to see if it would increase the likelihood of reproduction at fledgling colonies (Table 1). Feeders at these sites were accessible only by helicopter, and so were not typically refilled after the initial installation.

4.1.2 Translocation trials An evaluation of reintroduction success for captive-bred Vancouver Island marmots released 2003-2010 identified a “release effect” of poor overwinter survival in the first post-release hibernation (Jackson 2012). Captive-bred marmots averaged only 58.3% survival, whereas wild-born and previously released marmots achieved near-perfect survival each winter. A common assumption across reintroduction projects is that wild-born animals are superior to captive-bred animals in their survival and reproductive abilities at new locations. To evaluate whether or not this may be true for the Vancouver Island marmot, we initiated translocation trials that released mixed groups of wild-born and captive-bred marmots to the Buttle Lake area.

We included marmots with one of four levels of wild experience in our translocation. Captive-bred marmots released with no wild experience were either direct-released marmots that hibernated at the Tony Barrett Mount Washington Marmot Recovery Center (TBMWMRC) on Mt. Washington for the winter prior to their release, or facility marmots that arrived at the TBMWMRC from Calgary or Toronto Zoos or the Mountain View Conservation and Breeding Centre approximately 60 days prior to their release. Pre-conditioned marmots were born in captivity but hibernated successfully as part of the wild Mt. Washington colony for ≥1 winter prior to their release. Wild-born marmots were born to the wild Mt. Washington colony and translocated as yearlings or 2yos after ≥1 wild hibernation. As Table 2 illustrates, we released pre-conditioned and wild-born marmots in each year of the translocation trials, but included direct-released and facility marmots in just 2012 in 2014, respectively. For the sake of simplicity, we will call all sites “release sites” and describe all marmots as being “released” regardless of whether they were released from a facility or translocated from Mt. Washington.

The translocation trials were designed to help us determine whether wild-born marmots were more likely to survive and reproduce in the Buttle Lake area than captive-bred marmots, and whether pre- conditioned captive-bred marmots that had gained experience in the wild at Mt. Washington were more likely to survive and reproduce post-translocation than direct-released or facility marmots that lacked any wild experience. 4.2 Release Site Preparation We selected release sites after we had confirmed overwinter survival of marmots, so that we could release new groups near the colonies that most needed support. In June and July, we conducted helicopter-based surveys of releases sites to verify that snow had sufficiently melted and forage plants would be available for translocated marmots. All release sites except Marble Meadows had been used in previous years, and so ground searches for appropriate release burrows were mostly unnecessary.

-4-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

4.3 Release Protocols Captive-bred marmots had been born at one of three Canadian facilities: the Calgary Zoo (CZ), the Toronto Zoo (TZ), and the Mountain View Breeding and Conservation Center (MVF) in Langley, BC. These marmots were transported to the TBMWMRC on Mt. Washington in early May, where they were examined by the Project Veterinarian and kept in quarantine for a minimum of 30 days. In July, we also used the TBMWMRC as a temporary holding facility for wild-born and pre-conditioned marmots that had been re-trapped prior to their translocation to the Buttle Lake area (see Appendix IV for timeline of activities at Mt. Washington). The Project Veterinarian examined, ear tagged (National Band and Tag Co. Newport, KY USA) and implanted both facility and wild-born marmots with VHF radio transmitters (Holohil Systems Ltd. Ontario, Canada). Pre-conditioned marmots did not require surgery because they had been implanted prior to their release to Mt. Washington the summer before. In order to avoid additional stress following the invasive nature of the surgeries, newly implanted wild-born marmots were released back to the burrows from which they had been trapped. Implanted marmots were given a minimum of two weeks convalescence to allow for adequate healing from the implants before we re-trapped them in preparation for translocation. After being re-trapped, most wild-born and pre-conditioned marmots were kept in captivity for a short period of time, usually 1-3 days. This was necessary to enable a final examination by the veterinarian and to allow time to partition marmots into release groups that were balanced by sex and age. Marmots bred in captivity are not given access to traditional burrows; instead, they hide, hibernate, and raise their young in plywood “nest boxes” that have been filled with straw bedding and have exit holes at either end. Therefore, in order to facilitate the release of captive-bred marmots, field crew attempted to install nest boxes and bedding at sites that were due to receive facility marmots this year. This was successful at Sunrise, Flower Ridge, Greig Ridge, Henshaw, and Marble Meadows. It was not possible to install nest boxes at Castlecrag and Drinkwater, so crew placed bedding directly inside a natural burrow, and released marmots through the hole in just the front panel of a nest box. Wild-born marmots are not accustomed to using nest boxes, and pre-conditioned marmots had adjusted to using traditional burrows in their year on Mt. Washington, so these marmots were released into nearby burrows and not into the same nest boxes as facility marmots. 4.4 Monitoring 4.4.1 Radiotelemetry We tracked each marmot by radio telemetry throughout the field season. Telemetry receivers included the R1000 (Communications Specialist, Inc. CA, USA), Lotek SRX 400® and SRX 1000® (Lotek Wireless Inc. ONT. CA.). For antennas, we used the 3-element Yagi from Wildlife Materials (Wildlife Materials Inc. IL, USA). We conducted helicopter-based telemetry to collect most survival data for marmots in the Buttle Lake area, because the terrain is too steep and rugged for easy or efficient ground- based data collection. We flew multiple flights and at variable altitudes in an attempt to detect missing marmots. We did not try to determine the specific location of each marmot; instead, we focused on detecting the greatest number of signals possible to better inform our estimates of marmot survival. Implanted transmitters emit different pulse rates based on the transmitter temperature, which is ultimately determined by the body temperature of the marmot itself. We defined a marmot as being on “active” pulse if the pulse rate averaged 28 beats per minute. Marmots with an active pulse were considered to be alive. Because bare transmitters exposed to the sun have been known to heat to an

-5-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

active pulse rate when they should really indicate mortality, we conducted ground- and helicopter-based telemetry in the morning before temperatures became warm enough to create this effect. When there was an average pulse rate of <28 beats per minute, we considered the marmot to be on “mortality” pulse. Hibernating marmots cycle through periods of cooling and re-warming, and in a cool or “torpor” phase, their transmitter signal sounds like a mortality. Therefore, a marmot with a mortality pulse rate was evaluated as either dead or hibernating depending on the time of year (see 4.5.2). Whenever possible, we attempted to recover transmitters of marmots detected on mortality pulse to infer the cause of death. At kill sites, predators sometimes leave clues as to which species was responsible for the mortality, and we swabbed recovered transmitters from recent mortalities to help us verify the predator species by DNA.

4.4.2 Measures of Success In a study comparing reproduction by wild and captive marmots, Bryant (2005) found that the mean age of first breeding for wild females was 3.6 years. He estimated that the average number of pups in wild litters was 3.4 (SD=1.1, n=58), and determined that females usually produced pups either biennially (39.3%, n=11) or with a two-year interval (14.3%, n=4). Although females could wean pups in consecutive years (46.4%, n=13), Bryant estimated that the average interval between litters was 1.9 years (SE=0.7, n=17). In an ideal world, we would have defined feeder trial success as a (i) lower average age of first breeding, (ii) smaller average interval between litters, or a (iii) higher average litter size than that noted in the wild litters studied by Bryant (2005). However, we have found a high degree of individual variation in the breeding success of Vancouver Island marmots, and dams cannot always be identified with confidence, nor pups counted. There are also several positive indirect effects that may result from spring supplemental feeding, including (iv) the better body condition of pups and/or dams at hibernation time, (v) a greater likelihood of overwinter survival as a consequence of this better body condition, (vi) reproduction by females the following spring because of this better body condition, or (vii) the exposure of fed marmots to lower levels of predation risk a summer because they are in better body condition and can afford to be more cautious when foraging. The achievement of any direct or indirect effect would be a sign of success to us, and yet it is unlikely that we will ever be able to confirm these results statistically. We identified two short-term and three intermediate measures of success for our translocation trials. Short-term: (i) Survival through the first active season: the survival of a marmot from its date of release or translocation to the final inventory flights or ground surveys in September. Ideally, we would have waited until each marmot was on mortality signal and tracked them to a plugged burrow to confirm that they were underground and hibernating, but ground surveys are not practical for most sites in the Buttle Lake area. (ii) Survival through the first winter: the survival of a marmot, presumed to have survived to hibernation the previous year, through the winter until it is detected alive in the first two weeks of June the following year. Intermediate-term (iii) Survival through the first 14 months: the survival of a marmot from its date of release to surveys in late July or August the following year. (iv) Survival through the first 26 months: the survival of a marmot from its date of release to surveys in late July or August, two years in the future.

-6-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

(v) Successful breeding: the raising of a litter of pups to weaning so that at least one pup was seen or recorded by remote camera above ground on or after the first week of July. Because Vancouver Island marmots stay at their natal site until at least two years of age (Bryant 2005), pup litters could also be confirmed retroactively by documenting yearlings at a site. We predicted that wild-born marmots would achieve higher rates of intermediate-term survival and would record faster successful breeding in the Buttle Lake area when compared to pre-conditioned, direct-released or facility marmots. We also predicted that pre-conditioned marmots would achieve higher rates of first winter survival than marmots released directly from captivity with no experience in the wild, and would therefore record higher rates of survival and breeding than direct-released and facility marmots. Before we continue, we should explain an important distinction between this year and previous years in the detection of active season survival. This year, we conducted fall telemetry flights two weeks earlier than usual, in mid-September, in order to minimize the risk that the “mortalities” we detected would actually be marmots in the “torpor” phase of hibernation. Typically, we do not detect marmots in hibernation until October (Marmot Recovery Foundation, unpublished data). However, in the Nanaimo Lakes region this year, field crew conducted on-the-ground burrow searches for marmots on mortality signal, and confirmed that some marmots were in torpor by September 16, 2014. This is the earliest start to hibernation documented in that region in 22 years (1992-2014), and makes us question whether or not the marmots on mortality signal in the Buttle Lake area were actually hibernating,. Consequently, there are several instances where we were unable to determine whether a marmot was living or dead at the end of 2014. We included these marmots in our detectability results, but did not include them in our survival estimates. We will not know their survival status until June, 2015.

4.5 Statistical analysis 4.5.1. Translocation trials The translocation trials were designed to help us evaluate whether wild-born marmots survive and reproduce at higher rates than do captive-bred marmots in the Buttle Lake area, and whether there is a difference in survival between captive-bred marmots that are pre-conditioned and those that are released with no wild experience. Unfortunately, it is not possible to compare the levels of success achieved by different experience treatments using a classic statistical approach. To do so would commit sacrificial pseudo-replication (Hurlbert 1984), which ignores the variance introduced by releasing marmots in different years and on several different mountains. A more appropriate method for analyzing the data would be to model survival and reproduction using generalized linear mixed effects models (GLMM; glmer function, lme4 package, Bartoń 2012; glmmadmb function, glmmADMB package, Skaug et al. 2013), which can include random effects like year and mountain within the model structure itself. We attempted to run these models this year, but it was clear from model output that we did not yet have sufficient data to use this approach. Because there were no appropriate statistical methods of analysis to help us examine the data at this stage, we created simple charts to help us visualize the proportion of marmots from each experience treatment to survive the first active season, winter, 14 months, and 26 months in the Buttle Lake area. We grouped survival by experience level and plotted 2012 and 2013 separately. When viewing these charts, it is important to remember that the survival estimates were unlikely to have been uniform across all mountains.

-7-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

4.5.2. Mortality inferences When mortalities occur in marmots with working transmitters, this creates the potential for us to detect the mortality, track the signal to a transmitter, and recover evidence from the mortality site that may indicate the cause of mortality. This is often possible in the Nanaimo Lakes region, where many colonies are accessible by day hike. Decades of research at the Nanaimo Lakes colonies have shown us that although marmots can die from disease, transmitter issues, or accidents, marmot mortalities are generally attributable to either hibernation or predation (Jackson and Doyle 2013). These causes are seasonally specific. For Vancouver Island marmots, predation does not happen from November to March, because all marmots are in hibernation and there are no large predators able to access them when they are underground. Hibernation-related mortality does not happen during the summer or early fall, because all marmots are active in July and August, and no marmots are likely to die of hibernation-related causes in September or October. There are only three months of the year, April to June, when both hibernation and predation are possibilities, because some marmots will have emerged above ground and are therefore available to be predated upon. Mortalities that happen in April, May or October may not be trusted as such without recovery of a transmitter, because mortality and hibernation signals sound the same, and marmots could be alive but in torpor. In the Buttle Lake region, we rarely retrieve transmitters because of access issues and time limitations for field crew. Instead, we have always relied on a “timing classification system” to help us attribute mortalities to a cause. Mortalities that occur between mid-September and mid-June are classified as hibernation-related. Mortalities that occur between mid-June and mid-September are attributed to predation. This year, we evaluated our timing classification system in order to get a sense of its validity. We used the Buttle Lake timing classification system to categorize 125 known fate mortalities in the Nanaimo Lakes region. We created a confusion matrix to evaluate the overall accuracy of this classification system, and used the caret package (Kuhn 2012) to determine Cohen’s kappa coefficient (Cohen 1960) and the True Skill Statistic (TSS; Allouche et al. 2006) to assess its predictive ability. The kappa coefficient and TSS evaluate the improvement of a predictive model by estimating the proportion of correct predictions beyond those that could be expected by chance alone.

5. RESULTS The translocation trials were designed to achieve several goals related to survival and reproduction; however, for some measures of success we have not yet collected enough data to furnish a full statistical analysis. Therefore, we will first provide simple descriptive summaries (5.1 to 5.5) and then present the statistical results in section 5.6. 5.1 Feeder trials and remote camera footage (i) Castlecrag – We sighted two emergence holes during spring flights, and installed feeder tubes beside them. We set up remote cameras at each feeder, but only recorded footage of marmot use at the main hibernaculum. This footage revealed two tagged adults, but there was no footage of yearlings to confirm survival of last year’s pup(s), and no footage to suggest a litter was born in 2014. (ii) Marble Meadows – We found two emergence holes approximately 500m apart, and installed a feeder and remote camera at the upper site. Camera footage revealed regular use by two adults, and the appearance of two pups in July. Several videos showed pups hiding in the feeder tubes, and also consuming leftover biscuits.

-8-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

(ii) Sunrise – We observed one emergence hole, and installed a single feeder at the site. Camera footage showed two resident adults and, in July, at least one pup. Two marmots hibernated at a separate sublocation, but we did not find the emergence hole to identify exactly where they hibernated. (iv) Tibetan – We observed one emergence hole in the spring, and installed a single feeder on-site. Remote camera footage showed two tagged adults and two untagged yearlings or 2yo. Field crew observed three pups, an untagged yearling and untagged adult in the fall. (vi) Mt. Washington - We identified 10 emergence holes and installed 10 feeders and remote cameras at each location. Field crew documented 3 litters born to females aged 10yo (Claire), 8yo (Nicola) and 3yo (Hollis) and counted 11 pups in total. This was the second and third consecutive years of breeding for Hollis and Nicola, respectively. Based on the number of yearlings observed in 2013 and 2014, we must have underestimated the number of pups born in 2012 and 2013 by ~50%. Assuming a similar effort, range of error, and standard litter sizes of 3-4 pups, we believe there were probably 15-18 pups born in 2014, and we expect to see 11-16 yearlings in 2015 (Figure 2).

Please see Appendix V for images of Vancouver Island marmots recorded by remote cameras at Castlecrag, Flower Ridge, Marble Meadows, Sunrise, and Tibetan.

5.2 Translocation trials We translocated 9 wild-born marmots, 8 pre-conditioned marmots, and released 15 facility marmots in the Buttle Lake area in 2014 (Table 3).

5.2.1 – Detectability Similar to previous years, site surveys were limited by time, weather, rugged terrain, and the large number of marmots in the region. We detected 66% (29 of 44) and 71% (47 of 68) of all marmots released in translocation trials in the Buttle Lake area that were available to be detected at the start and end of the 2014 active season, respectively.

5.2.2 – Pre-conditioning at Mt. Washington Of the 16 captive-bred marmots released to Mt. Washington in 2013, there were 13 that survived the active season and hibernated on the ski hill (2013-14). This spring, we confirmed 77% overwinter survival by pre-conditioning marmots (10 alive, 2 dead, 1 missing; Figure 3). We later detected the missing marmot alive, which increased the overwinter survival rate to 85%. The 2014 spring was particularly mild when compared to previous years, with low snow levels and rapid melt (Figure 4).

5.2.3 – Marmots translocated in 2012 At the end of 2013, 22 of 30 marmots that had been translocated in 2012 that could still have been alive in the Buttle Lake area. Three marmots died overwinter, and two marmots died during the 2014 active season. At the end of 2014, any marmots still alive that had been translocated in 2012 had survived for 26 months in the Buttle Lake area. Over this time period, we estimate a maximum possible survival of 6-35% by captive-bred direct-released marmots, 0-40% by captive-bred pre-conditioned marmots, and 50-75% by wild-born marmots (Figure 5). For reproduction results, please see Section 5.5.

5.2.4 – Marmots translocated in 2013 Of 27 marmots translocated in 2013, there were 25 that could still have been alive in the Buttle Lake area at the end of the 2013 active season. During the 2013-2014 overwinter, one wild-born and three

-9-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

pre-conditioned marmots died, and during the active season, there were two additional mortalities. By the end of the 2014 active season, there were a maximum 19 of 27 marmots that had been translocated in 2013 that could still be alive. We estimated a maximum possible 14-month survival of 36-73% of captive-bred pre-conditioned marmots and 44-69% of wild-born translocated marmots (Figure 6). We detected one marmot on mortality signal in September that could have been either a true mortality or an early hibernation. For reproduction results, please see Section 5.6.

5.2.5 – Marmots translocated in 2014 We translocated 32 marmots in 2014, and confirmed survival by 63-100% of wild-born marmots, 56- 100% of pre-conditioned marmots, and 67-100% of facility marmots (Figure 7). We detected 7 marmots on mortality signal in September, and these could have been either true mortalities or early hibernations.

5.3 Mortalities We confirmed 11 mortalities in the Buttle Lake area in 2014, although there could have been as many as 24 if all the marmots on mortality signal in September were deceased. Only 1 transmitter was recovered during the active season, and that marmot had been predated. There was no predator sign at the mortality site to indicate which species was responsible. In the Buttle Lake region, it is rare to recover a transmitter, and so for all other cases, we used the timing of the mortality to infer the cause. We attributed 7 mortalities to hibernation and 3 to suspected predation. Since releases began in the Buttle Lake area, there have been a total of 174 marmot mortalities. We recovered transmitters for 47 of those mortalities, and attributed 24 mortalities to predation, 9 to hibernation, 4 to accidents, and 10 to unknown causes. Evidence implicated all three major predators: cougars (n=10), eagles (n=6), and wolves (n=4).

5.4 Dispersal Between 2007 and 2014, we documented ~15-20 dispersals by marmots in the Buttle Lake area (Figure 8). This includes two recent natal dispersals by wild-born Mt. Washington marmots to the Sunrise Lake colony (2012, 2014). The remaining dispersals were movements by captive-bred marmots away from their release sites. In 2014, there were 4 notable dispersals recorded in the Buttle Lake area: a wild-born 2yo female from Mt. Washington to Sunrise Lake (~11km), a pre-conditioned female from Mt. Washington to Castlecrag (~12km), a wild-born male from his release site on Flower Ridge back to his natal colony on Mt. Washington (~31km), and a direct-released female who went missing after her translocation in 2012 and was re-discovered on Mt. Phillips at the end of 2014 (~10km). Hikers this year photographed two unidentified marmots at a new location on Mt. Frink, and a freshly-used burrow on the Golden Hinde ridge system.

5.5 Reproduction and Population Size This year was the 4th consecutive year of successful breeding at Mt. Washington, and the 12th year of breeding since 2000. As mentioned in 5.1, field crew observed litters on Mt. Washington (3 litters, 11-12 pups) and Tibetan (1 litter, 3 pups), and remote cameras identified litters at the Flower Ridge (1 litter, ≥1 pup), Marble Meadows (1 litter, ≥2 pups), and Sunrise Lake (1 litter, ≥1 pup) colonies (Figure 9). At the end of the active season, we estimated that there were 64-85 marmots in the Buttle Lake area, plus an additional 43-46 marmots on Mt. Washington (Table 4, Figure 10).

Appendix VI describes the guidelines used for estimating population size. For greater detail on the status of colonies at each release site, please see Appendix VII.

-10-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

5.6 Statistical Analyses 5.6.1. Translocation trials (i) First active season We included marmots from 2012 and 2013 in this plot, but not from 2014, because there was some uncertainty about which marmots had truly survived. As Figure 11 illustrates, direct-released and pre- conditioned marmots achieved only moderate survival in 2012, and there was a great deal of uncertainty in the survival by wild-born marmots released that year. Low detection rates meant that only 13% were confirmed alive that fall, although survival could have been as high as 88%. This speaks to the challenge in estimating short-term survival for marmots in such a challenging landscape, especially because 5 of the 6 missing wild-born marmots were confirmed alive the following active season. Pre-conditioned and wild-born marmots both achieved high rates of survival in 2013. There were no marmots direct-released in 2013. (ii) First winter Missing marmots created some uncertainty in survival rates for both direct-released and wild-born marmots in 2012, and this made it difficult to determine which experience level performed best (Figure 12). The minimum level of survival confirmed for those groups was on the lower end of what we have recorded in the Buttle Lake region in recent years (~40%), but these estimates were not precise enough to be of concern. Pre-conditioned marmots achieved comparable winter survival to the 58.3% average survival reported for newly-released captive-bred marmots (2003-2010; Jackson 2012). The winter survival estimates for marmots released in 2013 were much more precise, and showed near-perfect survival by wild-born marmots. (iii) First 14 months Over 14 months, wild-born marmots released in 2012 survived better than direct-released or pre- conditioned marmots released that year (Figure 13). The ranges of survival leave some uncertainty; however, survival between direct-released and wild-born marmots would be comparable only if all missing direct-released marmots had survived and all missing wild-born marmots had died. We believe this is unlikely. Pre-conditioned marmots released in 2012 appeared to have achieved the lowest rates of survival, although they may not have survived much differently than direct-released marmots. For marmots released in 2013, survival was comparable between pre-conditioned and wild-born marmots, but it appears likely that wild-born marmots may have survived at a slightly higher rate. (iv) First 26 months Wild-born marmots released in 2012 survived for their first 26 months at much higher rates than direct- released or pre-conditioned marmots (Figure 14). Although there is still some imprecision in estimates, there is no situation in which direct-released or pre-conditioned marmots could have survived better than wild-born marmots. Still, we remain cautious with our confidence in these visual data patterns. This dataset included only 30 marmots released in 2012, and we often see a great deal of variation in survival from year to year.

5.6.2. Mortality inferences Of the 125 known-fate mortalities in the Nanaimo Lakes region, the timing classification system correctly identified 82 predation-related and 26 hibernation-related mortalities, but falsely classified 15 predations as being hibernation-related and 2 hibernation-related mortalities as being caused by predation. This suggests that we have consistently underestimated the effects of predation in the Buttle

-11-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Lake area. We estimated Cohen’s kappa coefficient and the True Skill Statistic (TSS) as 0.664 and 0.774, respectively. According to the magnitude guidelines developed by Landis and Koch (1977), this system achieved “substantial” agreement with the data. A classification system with no predictive ability would have produced kappa coefficient and TSS values of 0, and a system with perfect predictive ability would have scored 1.

6. DISCUSSION 6.1 Supplemental spring feeding The data do not yet suggest that spring supplemental feeding helps to increase the overwinter survival of pre-conditioning marmots, but we have come to realize that the feeders could provide benefits that are not immediately apparent. Reproductive success in marmots can be a function of female body condition (M. marmota, King and Allainé 2002; Hackländer and Arnold 1999), food resources in the spring (M. flaviventris, Andersen et al. 1976; but see Woods and Armitage 2003), food resources the previous year (M. caudata aurea, Blumstein and Foggin 1997), and the length of the previous growing season (M. flaviventris, Van Vuren and Armitage 1991). Supplemental feeding of Columbian ground squirrels (Spermophilus columbianus) was found to increase spring body weight and decrease the age at which females achieved reproductive maturity (Dobson and Kjelgaard 1985). Captive-bred Vancouver Island marmots released to the wild rarely produce offspring until they have survived at least two hibernation periods. Therefore, if spring feeders artificially extend the growing season and contribute to a more rapid recovery of female body condition from their first wild hibernation, they might also stimulate more rapid reproduction following translocation to the Buttle Lake area. Of the 4 litters produced near Buttle Lake in 2014, 3 were at sites with feeders. The only site that received a feeder but did not appear to produce pups was Castlecrag. Incidentally, we had some questions about the functionality of the remote camera at that site, which recorded only two images during a week-long 100-person search for a missing individual, despite several search-and-rescue volunteers noting its presence. If we did miss a litter at Castlecrag and any pups survive to be yearlings, we may be able to identify it retroactively in 2015. Marmot pups have a lower survival likelihood of survival than other age classes (M. flaviventris, Cohas et al. 2009, Lenihan and Van Vuren 1996; M. marmota, Allainé et al. 1998; M. vancouverensis, Aaltonen et al. 2009). The primary causes of mortality for Vancouver Island marmot pups appear to be predation and hibernation (MRF, unpub. data). Hibernation-related mortality of M. flaviventris pups is associated to poor mass gain (Armitage et al. 1976; Lenihan and Van Vuren 1996) or mass gain below a critical threshold (Woods and Armitage 2003). Prior to weaning, the growth of M. flaviventris pups is influenced by the body condition of the dam and the food available to her (Lenihan and Van Vuren 1996), and M. marmota dams have been found, in times of few resources, to protect their own body condition at the expense of their offspring (Allainé et al. 1998). Therefore, pups may not directly consume biscuits, which have often been finished by adults prior to July when pups first come above ground, but they may still benefit from feeders having been onsite, through the nutritional advantages given to their dams. In 2014, the early snow melt may have translated to a lesser need for and slower consumption of supplementary food. Remote camera footage showed that the pups at Marble Meadows eating biscuits when they first came above ground in early July. It is not practical to provide supplemental food for marmots throughout the summer, although this would almost certainly contribute to higher rates of overwinter survival, particularly by pups. Remote cameras have documented black bears using feeders in the late spring and early summer months, and we would not want to risk attracting black bears to marmot sites. And although small sample sizes and the dilute

-12-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

nature of potential benefits mean that we have no direct evidence to prove that feeders are beneficial to Vancouver Island marmots, we believe that the potential benefits from the provision of spring supplemental food vastly outweigh the minimal cost of installing and maintaining feeders. This is especially true because we typically install and maintain feeders while flying in the Buttle Lake area to achieve other priorities.

6.2 Survival by experience level It was a useful exercise to plot survival by experience level and year of release. Plots of marmot achievement at short-term measures of success (first active season, first winter) provided less direction in the experience level that performed best over those time frames, but were still important to consider. No marmot released in July could contribute to metapopulation growth via reproduction without having first survived one year in their new locations, and we expect that most will not contribute until they have survived at least two years. However, if we had noticed a clear trend that one experience level was surviving very poorly relative to the others during the first active season or winter, we could have tailored future releases to help us identify the underlying reasons for this difference. Therefore, we feel that it is still a worthwhile exercise to continue evaluations of short-term performance. When we reviewed the plots of 14-month survival, there was still some uncertainty between experience levels. Wild-born marmots were predicted to have the greatest minimum survival estimates, and the greatest maximum potential survival estimates. Yet, as a result of imprecision in estimates because of missing marmots at each experience level, the plot showed that direct-released and pre-conditioned marmots could have survived similarly. At 26 months, it was apparent that wild-born marmots achieved greater survival than pre-conditioned or direct-released marmots, regardless of any imprecision in estimates. The plot also revealed that pre-conditioned marmots may not have performed differently than direct-released marmots over this time period. We will need to observe these trends for multiple years of releases and, ideally, substantiate them statistically, in order to be confident that they will persist across a range of environmental conditions and predation activity.

6.3 Reproduction There has been consistent reproduction at Mt. Washington over the past four years. This is likely a result of several females reaching reproductive maturity at approximately the same time, since the average age of first breeding is 3.6 years (Bryant 2005). However, there is also some circumstantial evidence regarding first age of reproduction and frequency of breeding that suggests that spring feeders could have contributed to fecundity on the hill. In 2013, two wild-born females that were candidates for translocation produced their first litters at age two and therefore could not be moved, three females bred for their third consecutive year at Mt. Washington since feeders were installed, and a 9yo marmot weaned her first successful litter the year we provided a feeder at her hibernaculum. In 2014, there were 3 females that produced litters on Mt. Washington, including an 8yo that bred for her third consecutive year, a 10yo that was at the upper known threshold for successful breeding in the wild, and a 3yo who bred for the second year in a row after first breeding at age 2. Unfortunately, we have eroded much of our ability to detect the influence of feeders on Mt. Washington by translocating the same wild-born females that would have eventually contributed to the dataset on reproduction. Apart from Mt. Washington, there have been only two sites that achieved repeat reproduction in the Buttle Lake area since releases began in 2007, and only one of those sites (Tibetan) where pups were confirmed to have survived to become yearlings. Low levels of reproduction could be a consequence of the high proportion of newly-released marmots in the region, because breeding-aged captive-bred

-13-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

marmots in any region rarely reproduce before they have survived for two years in the wild. It was our hope that the inclusion of pre-conditioned and wild-born marmots in release groups would facilitate a shorter time interval between release and reproduction. In 2014, we documented pup litters at Flower Ridge (wild-born 3yo female, one year post-release), Marble Meadows (direct-released 4yo female, two years post-release), Sunrise Lake (captive-bred 4yo female, four years post-release) and Tibetan (wild 5yo female, born on site). The males at Flower Ridge, Sunrise and Tibetan were a 3yo pre-conditioned male one year post-release, a 4yo pre-conditioned marmot also one year post-release, and a 5yo direct- released male, three years post-release, respectively. We do not know the identity of the male at Marble Meadows. It should be noted that these breeders included one female and two males that bred after a single post-release hibernation. We hope to document additional examples of first-spring breeding in the next few years.

6.4 Detection In 2012, we achieved very low rates of detection of translocated marmots at the end of their first active season. Our ability to detect tagged marmots seemed to improve in 2013 when we flew at a higher altitude and planned longer, more expansive searches of the area around each release site. Although these intensive searches helped us to detect a greater number of marmots to confirm active season survival, they provided less information about the location of each animal. To confirm that a marmot is alive once it has been detected requires just 30-60 seconds of listening to a consistent transmitter pulse, whereas following a nearby transmitter signal to its source can take 5-10 minutes. We focused on collecting survival data rather than following marmots to their precise locations. This season, we continued to fly at high altitudes (>6,000 ft) in our searches for missing translocated marmots. We discovered that even though we could hear some signals from great distances, it was also possible to fail to detect marmots that were known to be present in an area while flying at high altitude; therefore, covering an area at both high and low altitudes would be the most effective approach. This is impractical because of the number of marmots for which to search (there are 17 currently missing) and impossible because of the vast terrain that would need to be searched for each of these marmots. The past two seasons, field crew camped on Greig Ridge in early September, and they were able to detect many missing marmots over 3-4 days of hiking. This type of detection success would not be possible at all sites, since Greig Ridge is a long, gentle ridge system that facilitates telemetry in several directions and can be hiked somewhat extensively in a few days. However, in August 2015, we intend to conduct trapping trips at several locations in the Buttle Lake area (see Recommendations), and this will give us the opportunity to conduct ground-based telemetry to augment our aerial searches. Hopefully this will increase our overall detection success.

6.5 Distribution and Dispersal The distribution of Vancouver Island marmot colonies in the Buttle Lake area has become much less fragmented since the start of the translocation trials. In 2011, there were only three mountains with settled, resident marmots. By the end of 2014, there were eight mountains with resident marmots, and five that had produced recent litters. There have been a few instances now where newly-released marmots have abandoned their release sites to colonize a new site nearby, such as Tibetan, Marble Meadows and Mt. Phillips. These dispersals increased the number and network of colonies near Buttle Lake and helped us to identify new sites that are suitable for marmot habitation. By releasing marmots at existing colonies as well as at sites that previously-released marmots have colonized themselves, we have strengthened the Forbidden Plateau and Western Strathcona metapopulations and given marmots

-14-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

that disperse in the future a much greater likelihood of encountering other marmots in the Buttle Lake area.

7. RECOMMENDATIONS After three years of translocations, we have made a strong start at building the dataset needed to answer our research questions. At present, however, we can model only simple, short-term success variables because most marmots released for the translocation trials have lived in the Buttle Lake area for less than two years. We must continue to monitor the long-term performance by marmots translocated in 2012-14, and to release and monitor new marmots. This will enable us to create the larger, more robust dataset that is necessary for appropriate statistical analyses of our trial results. In order to ensure that there are wild-born marmots available to be used in these translocation trials, we must make every effort to encourage successful reproduction on Mt. Washington. In the spring of 2015, we intend to continue feeder trials at Mt. Washington and the use of feeders at select sites in the Buttle Lake area. We identified six potential sites where there are pairs or groups of adult marmots that have demonstrated site fidelity has been consistent occupation might be reproduction next year. If conditions allow, we will install feeders and remote cameras at these sites. Because we found that remote cameras were a great help in identifying reproduction and confirming the survival of untagged marmots at sites, we will make their installation and maintenance a priority again next year. However, we will install them in early July, when pups spend much more time at their natal burrow entrance, and we may be more likely to capture footage of multiple pups in the same frame. Until 2014, there had only been a single site (Tibetan) in the Buttle Lake area that produced a litter of pups known to have survived to become yearlings. After confirming reproduction at four sites in 2014 and the presence of untagged or non-telemetered marmots at Castlecrag, Marble Meadows and Tibetan, we aim to trap and implant marmots at ~4 sites around Buttle Lake in 2015. We expect 2015 to be our last year of intensive translocation trials that include such large numbers of facility, pre-conditioned and wild-born marmots. We anticipate implanting and translocating approximately10-20 wild-born and 6-10 pre-conditioned marmots from Mt. Washington into the Buttle Lake area. We intend to release 10-15 facility marmots directly to the Buttle Lake area to continue our comparative evaluation of the survival of pre-conditioned versus direct-released captive-bred marmots. These marmots will be brought to the Mt. Washington facility in early May. We will release an additional 12-17 captive-bred marmots on Mt. Washington for pre-conditioning over the 2015-16 winter, in anticipation of their release into the Buttle Lake area in 2016. In 2014, we had hoped to fill a four-month First Nations wildlife technician position, but were disappointed not to hear from any candidates. We still believe that this position would provide unparalleled opportunities for an individual to gain experience in the field of endangered species conservation recovery, and so we intend to offer this position again in 2015. We will make a greater effort this winter to connect with the First Nations with traditional territories in the Buttle Lake area, and to identify and address some of the circumstances that could have prevented participation in our project this past year. We were encouraged by the success achieved by marmots in the Buttle Lake area during the 2014 field season. Reproduction rates in particular provided some evidence that we have gained traction in our recovery efforts in this region. If we examine the order of milestones achieved in the recovery of the Nanaimo Lakes metapopulation, we see that a critical turning point was the contribution to population

-15-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

growth through natural reproduction by wild-born and established marmots. In combination with our releases of captive-bred marmots, it dramatically increased the effective population growth each year. This past field season was the first year where fledgling colonies helped to increase population size at more than two sites. By continuing our efforts to support the existing fledgling colonies in the Forbidden Plateau and Western Strathcona regions, we hope to achieve even better success in 2015.

-16-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Acknowledgements We would like to thank the organizations and individuals that made this project possible. Special thanks to the Fish and Wildlife Compensation Program for their continued financial support. We gratefully acknowledge the support of Environment Canada and TimberWest Forest Corporation. The Marmot Recovery Foundation (MRF) provided significant financial support for the project, and we thank MRF Executive Director Viki Jackson and Donor Relations and Accounts Manager Kim Metz for their unwavering commitment to recovering the species. The Ministry of Forests, Lands, and Natural Resource Operations, West Coast Region, provided support through Jerry MacDermott and Sean Pendergast, as well as through their infrastructure and field gear. Katie Calon of the British Columbia Conservation Foundation (BCCF) provided administrative support. BC Parks’ Andy Smith shared his knowledge of the park as well as backcountry expertise, and BC Parks posted our Observers Program poster on their website for Strathcona Provincial Park. West Coast Helicopters provided all transport to and from the Park, putting in extra effort to pick up and drop off crews in poor weather. We would like to express our gratitude to Peter Gibson, Jesse Percival and the staff at Mount Washington Alpine Resort who shared information about marmot sightings and helped us to operate safely on the hill. PhD candidate Richard Schuster provided informal statistical guidance. Thanks to our Project Coordinator, Don Doyle, for providing guidance to the project and sharing many words of wisdom. Thanks also to project veterinarian Dr. Malcolm McAdie for his dedication and hard work in and out of the field, and to the field crew (Chris White, Sigurd Johnson, Mike Lester, Patrick Reid, Iona Kearns, Crystal Reid, and Cheyney Jackson) for their tireless commitment and creativity in working towards project goals.

-17-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

REFERENCES Aaltonen K, Bryant AA, Hostetler JA, Oli MK. 2009. Reintroducing endangered Vancouver Island marmots: survival and cause-specific mortality rates of captive-born versus wild-born individuals. Biological Conservation. 142 (10): 2181-2190. Allainé D, Graziani L, Coulon J. 1998. Postweaning mass gain in juvenile alpine marmots Marmota marmota. Oecologia, 113(3): 370-376. Allouche O, Tsoar A, Kadmon R. 2006. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). Journal of Applied Ecology, 43(6), 1223-1232. Armitage KB, Downhower JF, Svendsen GE. 1976. Seasonal changes in weights of marmots. American Midland Naturalist: 36-51. Andersen DC, Armitage KB, Hoffmann RS. 1976. Socioecology of marmots: female reproductive strategies. Ecology: 552-560. Bates D., Maechler M, Bolker B, Walker S. 2011. lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1-7. Retrieved October 15, 2014, from: http://CRAN.R- project.org/package=lme4. Beck BB, Rapaport LG, Price MRS, Wilson AC. 1994. Reintroduction of captive-born animals, in: Olney PJS, Mace GM, Feistner ATC, (Eds.). Creative conservation: interactive management of wild and captive animals. Chapman and Hall: London UK. pp. 265-286. Blumstein DT, Foggin JM. 1997. Effects of vegetative variation on weaning success, overwinter survival, and social group density in golden marmots (Marmota caudata aurea). Journal of Zoology, 243(1): 57-69. Bryant AA. 2000. Relative importance of episodic versus chronic mortality in the decline of Vancouver Island marmots (Marmota vancouverensis) in: Darling L. (Ed.), Proceedings of a conference on the biology and management of species and habitats at risk. Queen’s Printer. Victoria. pp. 189-195. Bryant AA. 2005. Reproductive rates of wild and captive Vancouver Island marmots (Marmota vancouverensis). Canadian Journal of Zoology. 83(5): 664-673. Bryant AA, Janz DW, deLaronde MC, Doyle DD. 2002. Recent Vancouver Island marmot (Marmota vancouverensis) population changes in: Armitage KB, Rumianstev VU (Eds.). Holarctic Marmots as a Factor of Biodiversity. ABF Publishing House, Moscow. pp. 411. Bryant AA, Page RE. 2005. Timing and causes of mortality in the endangered Vancouver Island marmot (Marmota vancouverensis). Canadian Journal of Zoology. 83(5): 674-682. Cohas A, Bonefant C, Kempenaers B, Allaine D. 2009. Age-specific effect of heterozygosity on survival in alpine marmots, Marmota marmota. Molecular ecology, 18(7): 1491-1503. Dobson FS, Kjelgaard JD. 1985. The influence of food resources on life history in Columbian ground squirrels. Canadian Journal of Zoology, 63(9): 2105-2109. Fielding AH, Bell JF, 1997. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation, 24(1), 38-49. Fischer J, Lindenmayer DB. 2000. An assessment of the published results of animal relocations. Biological Conservation. 96(1): 1-11.

-18-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Griffith B, Scott JM, Carpenter JW, Reed C. 1989. Translocation as a species conservation tool: status and strategy. Science (Washington). 245(4917): 477-480. Hacklander K, Arnold W. Male-caused failure of female reproduction and its adaptive value in alpine marmots (Marmota marmota). Behavioral Ecology, 10(5): 592-597. Hebda RJ, McDadi O, Mazzucchi D. 2004. Habitat history and the decline of the Vancouver Island marmot (Marmota vancouverensis) in: Hooper TD. (Ed). Proceedings of the Species at Risk 2004 Pathways to Recovery Conference. BC Species at Risk 2004 Pathways to Recovery Conference Organizing Committee. Victoria, BC. pp. 1-11. Hurlbert SH. 1984. Pseudoreplication and the Design of Ecological Field Experiments. Ecological Monographs, 54(2), 187-211. Jackson CL. 2012. First year site fidelity and survival in reintroduced captive-bred Vancouver Island marmots (Marmota vancouverensis). MSc thesis. University of British Columbia. pp. 1-72. Jackson CL, Doyle DD. 2013. Vancouver Island Marmot Project – Final Report 2013. Prepared for the Vancouver Island Marmot Recovery Team. Nanaimo, BC. pp. 1-42. Janz D, Blumensaat C, Dawe NK, Harper B, Leigh-Spencer S, Munro WT, Nagorsen D. 1994. National Recovery Plan for the Vancouver Island Marmot. Report No. 10. Recovery of Nationally Endangered Wildlife Committee. Ottawa, ON. Janz DW, Bryant AA, Dawe NK, Schwantje H, Harper B, Nagorsen D, Doyle D, deLaronde M, Fraser D, Lindsay D, Leigh-Spencer S, McLaughlin R, Simmons R. 2000. National Recovery Plan for the Vancouver Island Marmot (2000 Update). Report No. 19. Recovery of Nationally Endangered Wildlife. Ottawa, ON. Karels TJ, Byrom AE, Boonstra R, Krebs CJ. 2000. The interactive effects of food and predators on reproduction and overwinter survival of arctic ground squirrels. Journal of Animal Ecology, 69(2): 235-247. King WJ, Allainé D. 2002. Social, maternal, and environmental influences on reproductive success in female Alpine marmots (Marmota marmot). Canadian Journal of Zoology, 80(12): 2137-2143. Kuhn M., 2012. caret: Classification and Regression Training. R package version 6.0-37. Retrieved October 15, 2014, from http://cran.r-project.org/web/packages/caret/index.html. Landis JR, Koch GG. 1977. The measurement of observer agreement for categorical data. Biometrics, 33, 159-174. Lenihan C, Van Vuren D. 1996. Growth and survival of juvenile yellow-bellied marmots (Marmota flaviventris). Canadian Journal of Zoology, 74(2): 297-302. Meyer D, Dimitriadou E, Hornik K, Weingessel A, Leisch F, Chang CC, Lin CC. 2014. e1071: Misc functions of the department of statistics (e1071), TU Wien. R package version 1.6-4. Retrieved October 15, 2014, from http://cran.r-project.org/web/packages/e1071/index.html. Milko RJ. 1984. Vegetation and foraging ecology of the Vancouver Island marmot (Marmota vancouverensis). M.Sc. thesis. University of Victoria. Victoria, B.C. Munro WT, Janz DW, Heinsalu V, Smith GW. 1985. The Vancouver Island marmot: status and management plan. BC Ministry of Environment. Wildlife Branch. Wildlife Bulletin No. B-39. Victoria, BC.

-19-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Munro WT. 1979. COSEWIC status report on the Vancouver Island marmot Marmota vancouverensis in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. pp 1- 12. Nagorsen DW. 2004. Canada’s endemic mammals at risk: recent taxonomic advances and priorities for conservation. In Proceedings of the “Species at Risk: Pathways to Recovery” Conference. Scaug H, Fourner D, Nielsen A, Magnusson A, Bolker B. 2013. glmmADMB: Generalized Linear Mixed Models using AD Model Builder. R package version 0.7.5. Retrieved November 20, 2014, from http://glmmadmb.r-forge.r-project.org. Schwarz, CJ. 2014. Comparing proportions – Chi-square (χ²) tests. In Course Notes for Beginning and Intermediate Statistics. Retrieved November 10, 2014, from http://www.stat.sfu.ca/~cschwarz/CourseNotes Swarth HA. 1911. Two new species of marmots from British Columbia. University of California Publications in Zoology. 7(6): 201-204. Teixeira CP, Schetini de Azevedo C, Mendl M, Cipreste CF, Young RJ. 2007. Revisiting translocation and reintroduction programmes: the importance of considering stress. Animal Behaviour, 73(1), 1-13. Thiele J, Markussen B. 2012. Potential of GLMM in modelling invasive spread. CAB Reviews, 7(016), 1-10. Retrieved November 20, 2014, from http://www.uni- muenster.de/imperia/md/content/landschaftsoekologie/oekologischeplanung/team/publikationenjan/p otential_of_glmm.pdf Thornback J, Jenkins M. 1982. The IUCN Mammal Red Data Book: Threatened mammalian taxa of the Americas and the Australasian zoogeographic region (excluding Cetacea). IUCN. Gland. Switzerland. pp. 225-227. Vancouver Island Marmot Recovery Team. 2008. Recovery strategy for the Vancouver Island marmot (Marmota vancouverensis) in British Columbia. Prepared for the B.C. Ministry of Environment. Victoria, BC. Wolf CM, Griffith B, Reed C,Temple SA. 1996. Avian and Mammalian Translocations: Update and Reanalysis of 1987 Survey Data. Conservation Biology. 10: 1142–1154. Woods BC, Armitage KB. 2003. Effect of food supplementation on juvenile growth and survival in Marmota flaviventris. Journal of Mammalogy, 84(3): 903-914.

-20-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

TABLES Table 1. In 2014, we installed 10 feeders at Mt. Washington and 4 feeders in the Buttle Lake area.

Table 2. There were four different experience treatments for marmots released in the Buttle Lake region translocation trials (2012-2014).

Table 3. Thirty-two marmots were translocated to seven sites in the Buttle Lake area in 2014.

-21-

VI Marmot - Buttle Lake Supplementation Project ______November 26, 2014

Table 4 . Population estimates for the Buttle Lake area at the end of 2014.

-22-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

FIGURES

Washington

Sunrise Lake Marble Meadows Morrison Spire Castlecrag Tibetan

Greig Ridge

Henshaw

Flower Ridge

Drinkwater N

Mt. Washington 2012 releases/translocations 0 10km 2013 releases/translocations 2014 releases/translocations

Figure 1. Study area for feeder (2011-14) and translocation (2012-14) trials for the Vancouver Island marmot in the Buttle Lake area.

-23-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 2 . Pup production by year at the wild colony on Mt. Washington. In 2012, field crew counted 9 pups but trapped 15 yearlings in 2013, increasing the pup count by 6. In

2014, field crew counted 11 pups in 3 litters. We assume that we missed observing a litter

or two, and expect to find 11-16 yearlings at Mt. Washington in 2015.

Figure 3. Prior to feeder trials, overwinter survival by captive-bred marmots released at Mount Washington averaged 64.3% (2003-2010, black line). We estimated 62.5%, 72.7%, and 84.6% overwinter survival by marmots being pre-conditioned in 2011-12, 2012 - 13, and 2013 - 14, respectively. The higher survival rates in the past two years could have been influenced by mild winter conditions with low snow levels and a rapid spring melt.

-24-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 4. Snow water accumulation at the Wolf River snow station in Strathcona Provincial Park, on Vancouver Island, BC. The winter of 2013-14 (dark blue) was characterized by much lower than average snow levels, limited snow accumulation, and rapid spring melt.

-25-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

(a) Wild-born (b) Pre-conditioned

(c) Direct-released (d) Minimum estimated survival (July 2012 – September 2014.

Figure 5. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2012. (a) Wild-born marmots were born at the wild colony at Mt. Washington. (b) Pre-conditioned marmots were captive-bred and released to the wild colony on Mt. Washington at least one summer prior to their translocation; therefore, they survived ≥1 hibernation in the wild before being released near Buttle Lake. (c) Direct-released marmots were captive-bred and transported to the Mt. Washington facility the fall before their release. They hibernated in the facility, and were released to the Buttle Lake area the following summer without any previous wild experience. (d) A greater proportion of wild-born marmots than pre- conditioned or direct-released marmots were confirmed to have survived to September 2014. However, nine marmots were missing, and the survival status of these marmots could change the magnitude of difference for which experience treatment performed best over 26 months in the Buttle Lake area. -26-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

(a) Wild-born (b) Pre-conditioned

(c) Direct-released N/A (d) Minimum estimated survival (July 2013 – Facility: N/A September 2014.

Figure 6. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2013. Marmots of “unknown” status were found on mortality signal in the second week of September, but could have been hibernating, so their true survival status will not be known until June 2015. (a) Wild- born marmots were born at the wild colony on Mt. Washington. (b) Pre-conditioned marmots were captive-bred and released to the wild colony on Mt. Washington at least one summer prior to their translocation; therefore, they survived ≥1 hibernation in the wild before being released near Buttle Lake. (c) There were no direct-released or facility marmots included in the translocation trials for 2013. (d) A greater proportion of wild-born marmots than pre-conditioned marmots were confirmed to have survived to September 2014. However, the status of the missing and “unknown status” marmots could change these results. -27-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

(a) Wild-born (b) Pre-conditioned

(c) Facility (d) Minimum estimated survival (July 2014 – September 2014.

Figure 7. Current survival status for Vancouver Island marmots released in the Buttle Lake area in 2014. Marmots of “unknown” status were found on mortality signal in the second week of September, but could have been hibernating, so their true survival status will not be known until June 2015. (a) Wild-born marmots were born at the wild colony on Mt. Washington. (b) Pre-conditioned marmots were captive-bred and released to the wild colony on Mt. Washington at least one summer prior to their translocation; therefore, they survived ≥1 hibernation in the wild before being released near Buttle Lake. (c) Facility marmots were captive-bred and transported to the Mt. Washington facility at least 30 days before their release. They were released directly to the Buttle Lake area without any previous wild experience. (d) A greater proportion of facility marmots than wild-born or pre-conditioned marmots were confirmed to have survived two months from their release to September 2014. However, there are several marmots of missing or unknown status, and no marmots have yet been confirmed as mortalities, so these short-term results are inconclusive. -28-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 8. A map showing some of the more notable dispersals by marmots released in the Buttle Lake area (2007 - 2014). It is encouraging to see that some marmots found and joined other colonies. However, the distribution of colonies is still limited enough that many dispersers died before they encountered other marmots. Most dispersals were <12km in straight line distance. In order to increase the likelihood of successful dispersal in the future, we need to increase the number of persistent colonies on all sides of Buttle Lake.

-29-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 9 . I n 2014, we documented reproduction at five colonies in the Buttle Lake area. Three sites had been occupied for at least four years (Sunrise, Tibetan, and Washington), and two sites had only been successfully colonized in 2013 (Flower Ridge, Marble Meadows). Vancouver Island marmots typically breed biennially, so those same females are unlikely to breed again next year. However, releases and translocations have always been structured to establish multiple breeding-aged females in each colony. Assuming that key individuals achieve overwinter and early summer survival, there could be as many as six colonies with breeding-aged pairs that could potentially produce litters in 2015.

-30-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 10. Regional population estimates for the Nanaimo Lakes, Forbidden Plateau and Western Strathcona regions. Although the Nanaimo Lakes metapopulation is the only one close to achieving the target metapopulation size described in the National Recovery Strategies, the distribution of colonies in the Forbidden Plateau and Western Strathcona regions has greatly expanded since 2012, and 2014 was the best year for reproduction to date.

-31-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 11. Estimated survival over a marmot’s first active season in the Buttle Lake area, grouped by release year within experience level. The vertical lines represent the range of possible survival for experience levels with marmots missing that could have been either living or dead. The upper bar represents the maximum possible survival for each experience level, if all the missing marmots were detected alive. Red circles draw attention to the minimum possible survival by each experience level, which is the percentage of marmots confirmed to have survived the first active season.

Figure 12. Estimated survival over a marmot’s first winter in the Buttle Lake area, grouped by release year within experience level. The vertical lines represent the range of possible survival for experience levels with marmots missing that could have been either living or dead. The upper bar represents the maximum possible survival for each experience level, if all the missing marmots were detected alive. Red circles draw attention to the minimum possible survival by each experience level, which is the percentage of marmots confirmed to have survived the first winter.

-32-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Figure 13. Estimated survival over a marmot’s first 14 months in the Buttle Lake area, grouped by release year within experience level. The vertical lines represent the range of possible survival for experience levels with marmots missing that could have been either living or dead. The upper bar represents the maximum possible survival for each experience level, if all the missing marmots were detected alive. Red circles draw attention to the minimum possible survival by each experience level, which is the percentage of marmots confirmed to have survived the first 14 months.

Figure 14. Estimated survival over a marmot’s first 26 months in the Buttle Lake area, grouped by release year within experience level. The vertical lines represent the range of possible survival for experience levels with marmots missing that could have been either living or dead. The upper bar represents the maximum possible survival for each experience level, if all the missing marmots were detected alive. Red circles draw attention to the minimum possible survival by each experience level, which is the percentage of marmots confirmed to have survived the first 26 months.

-33-

VI Marmot - Buttle Lake Supplementation Project ______November 26, 2014

APPENDICES Appendix I. Performance Measures – Actual Outcomes

-34-

VI Marmot - Buttle Lake Supplementation Project ______November 26, 2014

-35-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix II. Confirmation of FWCP Recognition

Our research, with recognition of the substantial support from the Fish and Wildlife Compensation Program, has been presented to the:

 Vancouver Island Marmot Recovery Team (February 14, 2014)  Vancouver Island Marmot Captive Management Group meeting (July 24-25, 2014)  Vancouver Island University third-year Resource Management students (September 26, 2014)  Cowichan Valley Naturalists Society (October 6, 2014)  Canadian Institute of Forestry, Vancouver Island Section (October 22, 2014)

BC Parks posted a link to our Observer poster on their Strathcona Provincial Park website. We acknowledged FWCP on this poster, at: www.env.gov.bc.ca/bcparks/explore/parkpgs/strath

Sample presentation slides that acknowledge FWCP:

-36-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix III. Timeline of activities at Mt. Washington.

-37-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix IV. Remote camera footage from the Buttle Lake area.

Remote camera footage from the Castlecrag colony revealed two adults with distinctive molt patterns. Both adults had their own territories; one lived on the south side of the mountain (top and middle photos), and the other in a talus basin on the north side of the ridge (bottom photo).

-38-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

A remote camera at the Flower Ridge colony captured footage of two adults (top and middle) and a pup (in background of bottom photo).

-39-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

A remote camera at the Marble Meadows colony recorded one tagged adult (top), two pups (middle), and two unknown adults (bottom).

-40-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

A remote camera at the Sunrise colony captured footage of two resident adults, two transient adults and a pup (bottom left) from this footage.

-41-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Remote cameras at the Tibetan colony captured footage of two tagged adults (top) and two untagged 2yos (bottom).

-42-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix V. Guidelines used to estimate population size for the Buttle Lake area in 2014.

-43-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix VI. Marmot summaries for release sites used in 2012, 2013 and 2014 translocation trials.

Castlecrag Aerial telemetry confirmed overwinter survival by four known marmots, Kim (2yoF), Mia (6yoF), Tuffy (2yoF) and Rosco (2yoM) on site. During spring emergence flights, we saw two emergence holes, one of which was at a new location. We installed feeders at both locations, although only Mia’s feeder was visited freque ntly. We installed remote cameras at Mia’s main hibernaculum and at a nearby talus basin used in the 2010 releases, and footage revealed an adult marmot at each site. Field crew translocated four marmots (Thunder , Chianti, Mariko and Flake) to the talus basin. Field crew conducting inventory in September found a pre-conditioned marmot named Shiraz (2yoF) plugged into the new emergence hole that had been discovered in the spring. Shiraz was a captive-born marmot released on Mt. Washington in 2013. She would have been translocated in 2014, but she had gone missing, and had dispersed a straight-line distance of ~12km from her release site to Castlecrag. At the end of the 2014 active season, we estimated that there were 8-10 marmots at this site.

Drinkwater During spring emergence flights, we detected six telemetered marmots : Atticus (4yoM), Carol (3yoF),

Horizon (4yoF), Kerry (2yo F), McGinty (2yoM), Wheatley (2yoM) from previous translocation trials. Due to the rugged terrain and limited time with a helicopter, we did not pin down their exact location. In July, field crew translocated four marmots (Beatrice, Iga, Ion, Podrick) to a new release site, slightly north of previous years’ release sites. At the end of the 2014 active season, we estimated that there were 8-12 marmots at this site.

Flower Ridge During spring emergence flights, we detected two marmots from the 2013 translocations trials close to the release site (Allira, a 3yoF, and Hawking, a 3yoM), and saw one of them sitting at the emergence hole. In July, field crew installed a remote camera at this hibernaculum and recorded footage of two adults and at least one pup. This was the first litter born at this site. The two other marmots released at this site in 2013 were missing at the start of the 2014 active season. Jonah, a 2yoM, had been born on Mt. Washington and then translocated to Flower Ridge as a yearling on July 20, 2013. We re-trapped him on Mt. Washington on July 21, 2014, which meant that he had traveled a straight-line distance of ~30km in approximately one year. This is the first instance that we have recorded of a translocated marmot returning to their natal colony, and the longest straight-line distance travelled in a year.

In July, field crew translocated four marmots (Josh, Stein, Sansa, Little Rascal) to a previously-used hibernaculum on the ridge, a few hundred meters north of Allira and Hawking. During fall inventory flights, field crew were on the ground to remove the nest box from this release site, and found that the box was nearly completely filled with dirt. Inside and also below the nest box, they discovered some marmot bones and the bare transmitter from a marmot released in 2009 who had died in hibernation. We believe the newly translocated marmots “pushed” his bones and transmitter into the release box while excavating the burrow. At the end of 2014, we estimated that there were 6-7 adults/yearlings and ≥1 pup on site.

Greig Ridge In spring, we detected only a few marmots that had been translocated in previous years. In July, field crew translocated five marmots (Tiffany, Soupbone, Garbo, Samwell, Riesling) to a release site on Greig Ridge that was last used in 2008. During a fall inventory trip, field crew located all marmots translocated in 2014, and a much higher percentage of previously translocated marmots. A hiker reported finding fresh burrows on the same ridge as the Golden Hinde, a peak 5km west of the Greig Ridge colony location; however, during fall flights we did not detect any telemetered marmots in the area. At the end of the 2014 active season, we estimated that there were 11-14 marmots at this site.

-44-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Henshaw In spring, we observed a very large emergence hole in the main meadow, and confirmed survival by five marmots from the 2013 translocations (2yoFs Emma and Malty, 3yoMs Smokey and Ian). We also recorded the overwinter mortality of 2yoM Lester2. In July, field crew translocated five marmots (Pinot, Rosee, Skeena, Eddard, Tyrion) to a boulder cluster in the main meadow, and detected an active signal for the missing 4yoF, Wascana. In the fall, we confirmed two mortalities at the site. At the end of the 2014 active season we estimated that there were 8-9 marmots on site.

Tibetan We did not translocate marmots to this site in 2014. During spring and fall trips to the area, a total of one untagged adult, two untagged pups and one untagged yearling were observed in the main talus slope. Field crew installed a remote camera in the area, and footage revealed two tagged adults and two untagged adults. The two telemetered adults (5yoM, Skipper, and 5yoF, Cherry) were confirmed alive on all visits and flights to this location. At the end of the 2014 active season, we estimated that there were 3-5 adults and 3 pups on site.

Marble Meadows This spring, we confirmed survival of an adult female (3yo Margaret) who had moved here from Greig Ridge and hibernated successfully 2013-14. Two staff from Mt. Washington reported finding a second emergence hole at this remote site, just a few hundred meters away, which confirmed that there was at least one untelemetered marmot that also hibernated at this site. In July, field crew translocated five marmots (Iza, Arya, Moses, Col Mustard, Bow) to support this new colony. A remote camera that was set up near the upper emergence hole showed two pups, one tagged adult and two unknown, tagged marmots that were likely newly-translocated marmots exploring their new habitat. These pups were the first born at this location. Mid-summer, we recovered the transmitter from Margaret, the female that we believed to have bred, but saw footage of at least one pup in the company of another adult. At the end of the 2014 active season, we estimated that there were 4-6 adults/yearlings and ≥1 pup at this location.

Morrison Spire We did not translocate marmots to this site in 2014.

Sunrise During spring emergence flights, we confirmed survival by the 4yo Margaret, 3yo Horatio, and 2yos Ora and Flash. Field crew installed a remote camera at this site, and footage showed two resident adults and at least one pup, the first litter born at this site. In July, field crew translocated five marmots (Thomas, Highwood, Iron Maiden, Dana, Soleil) to two nearby burrows. Camera footage at the pup burrow showed two additional marmots passing through the area; it is likely that these were marmots that had been translocated in mid-July and were exploring the area. At the end of the 2014 active season, we estimated that there were 8-9 adults/2yos/yearlings on site, as well as 1-3 pups.

-45-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Appendix VII. Photos from the 2014 field season.

Field crew practice searching for avalanche beacons as part of their Avalanche Safety Training (photos by Cheyney Jackson).

-46-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Crew member Chris White after installing a feeder at the Tibetan colony (June 2, 2014; Sig Johnson).

Remote camera footage shows several pre-conditioned marmots using a feeder on Mt. Washington, May 2014. Pre-conditioned marmots achieved high overwinter survival in 2013-14.

-47-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

After a mild winter, the snow melt on Mt. Washington was at least two weeks ahead of typical spring conditions (May 30, 2014; Mike Lester).

Rascal, a wild-born marmot, gathers some bedding material in early June (Mike Lester ).

-48-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Untagged yearlings from Mt. Washington. These two marmots were trapped and implanted with radio telemetry and were part of the wild-born demographic for translocation to the Buttle Lake area (top: Cheyney Jackson; bottom: Sigurd Johnson).

-49-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Crew member Patrick Reid carries a marmot to its release site (Greig Ridge, July 2014).

In 2014, we released 12 captive-bred marmots for pre- conditioning on Mt. Washington (Cheyney Jackson).

-50-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Remote camera installation frequently required creativity and innovation. In this photo, Sigurd

Johnson installs a remote camera on a piece of rebar in front of the Flower Ridge hibernaculum. This camera captured footage of the first known litter to Chris White programs a remote camera strapped to a be born at this site (Cheyney Jackson). tree at the Castlecrag colony (Patrick Reid).

At Marble Meadows, there were no boulders or trees The remote camera at Henshaw was strapped to a near the hibernaculum, so Sigurd Johnson strapped fallen snag that we leaned out into the air over a pile the camera to the feeder, and used a pile of rocks to of boulders. hold the feeder upright. This remote camera also captured footage of a pup litter.

-51-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Telemetry on Greig Ridge (Chris White).

Sigurd Johnson conducts fall telemetry in the alpine on Greig Ridge, Strathcona Park, September 2014 (Chris White).

-52-

VI Marmot - Buttle Lake Supplementation Project November 26, 2014

Project veterinarian Malcolm McAdie performs an The field crew debriefs at the end of a trapping session implant surgery on Mt. Washington while crew member (Mike Lester). Chris White takes notes (Mike Lester).

In the spring, this new emergence hole on Castlecrag was a mystery to us because none of the known resident marmots had hibernated here. In September, 2014, field crew found this burrow plugged again, and detected a missing pre-conditioned female from Mt. Washington, Shiraz, hibernating inside. She traveled ~12km, and saved us the cost of a flight. (Cheyney Jackson)

-53-

The marmot that created this emergence hole on Castlecrag this spring was a mystery to us until the end of the 2014 field