Wolf census results in the Lake Revelstoke area February, 2010

Wolf census results in the Lake Revelstoke area February, 2010 Harry van Oort, MSc RPBio Corey Bird, BSc Submitted to Ministry of Environment, Nelson, BC, and to The Columbia Basin Fish and Wildlife Compensation Program March 31, 2010

Acknowledgments Funding for this project was provided the BC Ministry of Environment, and from the Columbia Basin Fish and Wildlife Compensation Program. BC Ministry of Environment and the BC Ministry of Forests provided in-kind support by providing snow-machines and personnel. Leo DeGroot, Garth Mowat, Harry van Oort and Corey Bird coordinated this census program. Helicopter service was provided by Glacier Helicopters. Harry van Oort, Corey Bird, Rob Serrouya, Leo DeGroot, Ryan Gill, Cary Gaynor, Tom Roos, and John Townley contributed to field sampling. Garth Mowat provided constructive comments to an earlier draft of this report. Cover photo of Lake Revelstoke, looking west to the Monashees, February 18 th, 2010. ii

Abstract Populations of wolves and cougars have played an important role in the decline of mountain caribou. As part of the caribou recovery plan, the Ministry of Environment has been managing and monitoring populations of predators and their primary prey within caribou recovery areas since the winter of 2006/7. As part of the monitoring program, we conducted a wolf census in the Lake Revelstoke Survey area in February, 2010. The late winter was uncharacteristically warm with much of the precipitation falling as rain and thus, unconventionally, the census was conducted following several days of heavy rain during February 18 th 20 th. Nonetheless, survey conditions were conducive to determining wolf abundance. Old tracks were evident, but degraded substantially by the rain. Conditions following the rain event were clear, sunny and somewhat cold. Fresh tracks were visible on the surface hoar that developed on the rain crust, giving observers the ability to separate fresh tracks from those that were older. These conditions were ideal for determining pack size. The census was conducted using a combination of aerial and ground transects and covered a total transect length of 1552 km in 45.9 hours over 3 days. A minimum of 15 wolves in 5 packs were found during the census. Additional survey effort in wolf habitats bordering the study area detected a lone wolf in the Red Rock area, and no wolves in the Illecillewaet drainage. Wolf numbers were similar to the results from the previous winter. iii

Table of Contents 1 Introduction...1 2 Lake Revelstoke Survey Area...1 3 Methods...2 4 Results...3 5 Discussion...7 6 References...9 List of Figures Figure 1. Fresh wolf tracks registered in the surface hoar-frost, typical of the fresh tracks observed in the 2010 census, were excellent for determining pack size. Fresh tracks rarely penetrated the rain crust, making them challenging to detect...4 Figure 2. Transect lines and wolf pack locations from the February 2010 Lake Revelstoke wolf survey. Solid black dots represent wolf pack size and are in 4 sizes, from 1 to 7 wolves/pack. Encircled dots are minimum pack size estimates; naked dots are total counts. The coloured polygons the approximate areas used by the labeled caribou populations...5 List of Tables Table 1. Wolf census results for the 2010 Revelstoke wolf survey. Pack size is qualified by count type as being either a total count or a minimum pack size estimate. Minimum counts are qualified by an estimate, if possible, showing how many wolves might be missing from the Minimum Pack Size provided...4 iv

1 Introduction The decline of mountain caribou in the Columbia Ranges of British Columbia is likely the result of the numerical increase in the populations of other large ungulates and their predators following land-scape level habitat modifications (Wittmer et al. 2005; Wittmer et al. 2005; Apps and McLellan 2006; Wittmer et al. 2007)). In 2004, the Province of British Columbia established the Species at Risk Coordination Office (SaRCO) to direct recovery initiatives for mountain caribou. As part of the recovery program, the Ministry of Environment has been managing and monitoring ungulate and predator populations within and adjacent to caribou habitat. Predator monitoring focused on wolves (Canis lupis) and, to a lesser degree, cougar (Puma concolor). The predator-monitoring program began with the collection of preliminary data in the winter of 2006/7 and in the following two winters (2007/8 and 2008/9), data were collected in the North Columbia Mountains, the Central Selkirk mountains, the South Selkirk mountains, and the South Purcell mountains (Gaynor et al. 2007; van Oort et al. 2009; van Oort et al. 2010). In February of 2010, we conducted a wolf census in the Lake Revelstoke Survey Area, situated in the North Columbia Mountains. In this report, we provide the details of this survey, including the survey circumstances, conditions and results. For a more detailed account of the background of this project, please refer to the 2008/9 overview report (van Oort et al. 2010). 2 Lake Revelstoke Survey Area The area for this census, referred to as the Lake Revelstoke Survey Area, is located north of Revelstoke, between Revelstoke Dam and Mica Dam, and includes all appropriate habitats below 1200 m elevation, including all drainages that flow into Lake Revelstoke. This study area is contained by two large and heavily glaciated mountain ranges: the Northern Monashee Mountains to the west and the Northern Selkirk Mountains to the east - both ranges have high-elevation peaks that are extremely rugged, and extensively covered by glaciers (max elevation 3520 m). This area receives approximately 1490 mm precipitation per year (calculated using the application ClimateBC v3.21; [Wang, Hamann, et al. 2006 #610]. The valley bottom is dominated by very wet cold (vk) and wet cool (wk) subzones of the Interior Cedar Hemlock (ICH) biogeoclimatic zone (Meidinger and Pojar 1991). The study area is largely devoid of human settlements; exceptions include helicopter skiing operations (three large lodges run by Canadian Mountain Holidays), and the small townsite of Mica Creek at the north end of the survey area, in addition to two small guide-outfitting lodges. Very few houses or recreational properties exist in the area. Active logging operations occur throughout the survey area. Several trap lines are active in the study area. There were 3 additional areas outside the study area that were surveyed including: a northern extension to Red Rock harbor north of Mica, an eastern extension up the Illecillewaet watershed east of Revelstoke, and an extension up the Jordan River watershed northwest of Revelstoke. As in previous years, results are provided, but any detections of wolves outside of the survey area are excluded from the Lake Revelstoke Census count. 1

Moose and wolves are found throughout the study area. Wolf habitat can be found on the slopes above Lake Revelstoke along transmission lines and in cut blocks. There are many side valleys in this study area. The west side of Lake Revelstoke (east slope of the Monashees) is incised by many narrow and rugged side valleys, and most of the forestry at the southern end occurred on the east-facing slopes above the lake. In general the more northerly side-valleys in the Monashees are larger, and contain more wetlands and cut-blocks, therefore providing more suitable moose-wolf habitat. On the east shore of Lake Revelstoke, the side valleys are fewer, but much larger, and contain considerably more wetlands and forestry, and are known to support a large proportion of the wolf population in the survey area. 3 Methods As with previous years (van Oort et al. 2009; van Oort et al. 2010), we conducted a blitz type census where the entire survey area was covered in a 3-day period in an attempt to count all animals on one occasion using a combination of ground and aerial transects. The timing of the blitz was carefully chosen so that fresh tracks 2 to 3 days old can be readily differentiated from old tracks. Only fresh tracks were counted. Ideally, old tracks have degraded sufficiently or are buried under enough snow, that they do not detract from the observer s ability to spot a fresh track from the air. The ground component of the survey was done by arranging several observers systematically on preestablished transects, each following transects loaded into hand-held GPS units. The ground survey covered the major drainages on the east side of the Lake Revelstoke reservoir (Laforme Crk, Carnes Creek, Downie Creek, Goldstream Creek, and the Bigmouth Creek drainages) as well as logging roads south of Downie Creek. A ground transect was also conducted up the Westside Road below Frisby Ridge north of Revelstoke. The aerial component of the survey covered all major drainages on the west-side of the Lake Revelstoke reservoir to the Mica Dam as well as areas on the east side of Lake Revelstoke not covered by ground surveys. During ground transects, we recorded the location of each predator track we crossed during the survey, grouping any waypoints that were thought to be made by the same animal (or group of animals). If it was unclear how large the group of animals was that had made a set of tracks, we tracked wolves until a better estimate could be made. Estimates of group size were qualified as being either minimum counts the largest group size that we were certain of when the total number could not be determined (e.g., where wolves were following each other s tracks) or, total counts, where we were certain of the group size (e.g., when all tracks could be separated with certainty). Visual observations of wolves were noted as such unless the animals were partially hidden in the forest. Aerial transects were conducted using a Bell 206B helicopter with three experienced observers in addition to the pilot. Aerial transects were not linear; flight paths were chosen to survey appropriate habitats in each area and were highly influenced by track observations. Meandering flight paths or multiple passes were made when extensive habitats were encountered; otherwise, the transect would contour a slope, preferentially following unplowed roads, transmission lines, or cut block boundaries rather than flying over heavily timbered habitats where detection of tracks is unlikely. During an aerial 2

survey, the flight path was recorded using a GPS tracklog set to record position coordinates every 200m. Wolf tracks detected from the air were tracked so that the full extent of the track could be ascertained. When possible, tracks were verified on the ground, either by landing or by a separate field crew member on the ground if ground access was possible. In all cases, attempts to count the size of wolf packs were made, and qualified as being either minimum or total counts. 4 Results There was a lower-than average valley-bottom snowpack in 2010; however, the survey area generally had snowpack covering a 100% the ground. One minor exception was the deer winter range south of the Downie drainage on the east shore of the lake, which had bare ground in several areas. A 4 day rain event subsided on the night of Feb 16 th after degrading almost all of the wildlife tracks such old tracks were easily spotted and discounted during the survey. During the rain event, the snow line was near the upper elevation of the study area (1200 m). February 17 th was warm and sunny, and there was a short period in the afternoon when wolves left well-defined tracks in soft sun-affected snow, but such tracks were not commonly detected during the survey. The survey took place on February 18 th - 20 th. During the survey period, the weather was mild and the sky was clear; the sun generally did not soften the snow surface greatly during this period. In general, wildlife did not penetrate the snow after the 16 th February. The snowpack had a heavy rain crust that was sufficiently strong to support adult moose in some areas. In most areas however, moose were penetrating the surface, and wolf tracks were observed to penetrate the snow irregularly. Such occasional footprints provided important cues during the aerial survey component for locating wolf tracks. Surface hoar developed on top of the rain crust, and provided excellent conditions for counting wolves once the tracks were detected (Figure 1). The census covered a total transect length of 1552 km (409 km on the ground and 1143 km surveyed from the air) over 45.9 hours (34.9 hours on the ground and 11 hours from the air). A minimum of 15 wolves in 5 different groups were counted in the Lake Revelstoke Survey Area (Figure 2, Table 1). An additional lone wolf track was observed at Red Rock (pack 1; Table 1). No wolves were detected in the Illecillewaet drainage or the Jordan drainage. Two radio-collared wolves were located independently of the survey effort and visually observed (1 traveling alone near the mouth of Keystone Creek and the other traveling with a pack of 7 in the Goldstream); in both cases, the collared wolves had been detected independently by the survey, and the pack size was correctly identified (Table 1). 3

Figure 1. Fresh wolf tracks registered in the surface hoar-frost, typical of the fresh tracks observed in the 2010 census, were excellent for determining pack size. Fresh tracks rarely penetrated the rain crust, making them challenging to detect. Table 1. Wolf census results for the 2010 Revelstoke wolf survey. Pack size is qualified by count type as being either a total count or a minimum pack size estimate. Minimum counts are qualified by an estimate, if possible, showing how many wolves might be missing from the Minimum Pack Size provided. Pack # Detected in survey Detected by telemetry Pack Name Pack Size Count Type 1* Red Rock 1 Total 2 Goldstream 7 Minimum (+1) 3 Downie/Gregson 3 Minimum (+1) 4 Carnes 1 Total 5 Big Mouth 2 Minimum (+1) 6 Kirbyville 2 Total *not in the Lake Revelstoke survey area Lake Revelstoke Total: 15-18 4

Figure 2. Transect lines and wolf pack locations from the February 2010 Lake Revelstoke wolf survey. Solid black dots represent wolf pack size and are in 4 sizes, from 1 to 7 wolves/pack. Encircled dots are minimum pack size estimates; naked dots are total counts. The coloured polygons the approximate areas used by the labeled caribou populations. 5

Tracks from pack 2 were located in 2 places in the Goldstream during the ground portion of the survey done on February 19 th. The first set of tracks dated to the afternoon/evening of the 17 th and were found at the back of the French Creek/Sweeperbill FSR (between French Creek and the Norman Wood), the second set were fresh tracks (on surface hoar) dating to the 19th, and were found at the front end of the Goldstream on the transmission line off the French Creek FSR. Separate tracking for the first set of tracks by two observers on the ground determined that there were 7 or 8 wolves in the pack, an estimate later confirmed on the second set of tracks by both ground observers. This pack was also later located from the air on February 19 th using telemetry, and 6 or 7 were visually observed (5 in the open, 1 or 2 that stayed in the trees) near the transmission line off the French Creek FSR. A download of wolf locations from the GPS collar on a wolf from this pack confirmed that this pack had been at both track sites. There were many older tracks observed in the Downie Creek drainage (Pack 3) on February 18 th. In one location, the observers were able to count a total of three 3 wolves. A single track was also observed, but it was unclear if this was made by an additional animal or not. On February 20 th, fresh tracks on surface hoar made by 3 wolves (2 large and one small) were found leading out of the Downie and led south to the where they were picked up again on Gregson FSR. Pack 4 was a single wolf. On February 18 th, observers found tracks from a lone wolf on the side of Highway 23 north of Carnes Creek with tracks pre-dating the rain event. No other tracks were found for this wolf during the ground portion of the survey however, because this track was sufficiently isolated from other wolf groups, and tracking options are limited in this area due to snowmobile activity and poor snow coverage, this wolf was counted in the census as a single animal. Later, this conclusion was supported, as the lone wolf turned out to be collared with a downloadable GPS collar, and was located via telemetry on February 20 th. The animal was visually observed as travelling alone near the mouth of Keystone Creek at that time. GPS data showed that this wolf had been at the location of Carnes Creek prior to the census. Additional un-related field tracking (post-census investigation of wolf GPS clusters) also confirmed that this wolf was in fact travelling alone. This wolf was collared on the west side of Lake Revelstoke in 2008/9 and has spent most of the past year on the Monashee side of the reservoir. Pack 5 in the Big Mouth drainage (Table 1) were successfully detected and tracked from the air to their current location. The tracks were first identified by occasional foot penetrations, and were tracked on surface hoar over many kilometers up the Louis Lee tributary. In several places, the tracks clearly showed three animals, but there were only two bed sites seen at two locations. When the pack was visually observed at the edge of timber, only two animals were counted. On the ground, we obtained evidence that a lone animal had entered the drainage after traveling down from Mica during the rain event. During a similar time period there were two or three wolves handling a moose kill near the front end of the Big Mouth. These observations suggested to the observers that there were three animals in total two residents and a loner, but that the lone wolf was socializing at a distance from the resident pair. All independent tracks from the lone wolf were older, and therefore counted this group as a minimum count of 2 with potential for one more (Table 1). 6

Pack 6 was a pair of wolf tracks on the west side of the lake near Seymour Creek. The tracks were first detected as occasional foot penetrations, but the tracks left on surface hoar provided unequivocal evidence that these tracks were made by 2 animals. Near Kirbyville Creek there was a nucleus of older tracks in edition to the fresh surface hoar tracks. All of the 5 wolf pack detections in the survey area, except pack 4, included the observation of fresh tracks on surface hoar. In all 4 groups with fresh tracks, there were older tracks observed in the vicinity of the fresh tracks. Only one set of older tracks were detected that were not associated with fresh tracks. The latter tracks were detected during the aerial survey. These were made by a minimum count of 2 wolves at an old wolf kill site between Horne Creek and Scrip Creek. It is possible that we missed fresh tracks at this location; however, there were very few older tracks, which were situated by an old kill site, and the site was situated across the lake from Pack 5. 5 Discussion Ideal tracking conditions for wolf census require that old tracks can be differentiated from fresh tracks, and typically involve a fresh snow fall event of 20 cm or more about 2 days prior to the survey. The winter of 2009/2010 was unusually warm in the interior mountains of British Columbia, and the precipitation often occurred as rain instead of snowfall at valley bottom. Ideal conditions occurred over the Christmas holidays but did not occur again after the New Year. By February, the snowpack was starting do disappear in the dryer parts of the study area, and so we attempted to conduct the survey after a large rain event. The rain event lasted several days prior to the survey, ending on February 16 th. The day following the rain storm (February 17 th ) was warm and sunny with isothermic snow conditions which later froze that evening, and conditions remained somewhat cold for the survey with tracks registering on the surface hoar-frost that developed. While unconventional, these conditions did allow a way for observers to separate fresh tracks from those that were older as hoped, and several aspects of the results provided assurance that the conditions were adequate for our purposes. Fresh tracks were challenging to detect; in contrast, the older tracks, which were made during warm conditions, were easily detected. With one exception, only the fresh tracks were used in estimating the number of different packs. The one wolf which was detected and counted without fresh tracks, was later confirmed to be separate from other groups because it was wearing a GPS telemetry collar. This animal was located after the survey results were interpreted, in the dense forest habitats where deer occur along Highway 23, and the data confirmed that it was alone, and responsible for the older track we found. Not detecting packs would appear to be the largest potential for error in the 2010 survey, given that surface hoar tracks were challenging to detect. However, our aerial survey was conducted far more carefully than we would normally have done in the past by moving over the landscape more slowly, conducting ground checks regularly, and spending more time examining each habitat block to make sure that tracks were not missed. Our results suggested that our detection rate was actually above 7

average, compared with previous years. First, consider that very old tracks were erased from the snow, and that the older (mid-aged) tracks were generally easy to detect, and found in isolated patches. If we were missing the fresh tracks we would expect to find some mid-aged tracks that were not associated with fresh tracks. Only one such set of mid-aged tracks was observed that we could not clearly associate with fresh tracks, and it is possible that these were made by the Big Mouth pack located across the lake 1. Hence, the number of mid-aged tracks that were unaccounted for was minimal one pack at most. Second, we independently detected both of the collared wolf groups in the study area via the survey tracking methods, suggesting a high detection rate. This was confirmed by independent telemetry work. Hence, the results suggest that our estimate of the number of wolf groups was reasonably complete. The tracking conditions were ideal for determining group sizes, which was a strength of the tracking conditions - and this is reflected by a reasonably tight estimation of group size in our results. Half of the 6 pack size estimates were considered to be total counts. The 3 minimum counts were thought to be off by no more than one animal. Hence we feel that the census estimate is both accurate and precise. The total number of wolves detected did not differ greatly between this and the previous winter s census estimates. In 2010 we estimated a minimum of 15 wolves in the Lake Revelstoke Area; during the 2008/9 Lake Revelstoke Census a minimum of 13 wolves were counted (van Oort et al. 2010). A notable difference between these years is that in 2008/9, there were 3 wolves observed in the Goldstream (minimum estimate) compared with 7 in the current results. In the summer of 2009, the Goldstream pack was known to have successfully raised a litter in the French Creek area, which corroborates the expansion of this pack. There also seems to be a shift in the number of wolves on the east and west sides of Lake Revelstoke. In 2008/9 there were 6 wolves on the west side (in 3 groups Ratchford, Nagle, and Seymour), and 7 on the east side in 2 groups (Goldstream and Downie; van Oort et al. 2010) while this year we observed only 2 wolves on the west side in 1 group (Kirbyville) and 14 on the east side in 4 groups (Bigmouth, Goldstream, Downie, and Carnes). The collared wolf detected at Carne s Creek would have normally been detected on the west side. In the past it has been normal for wolves to use both sides of the reservoir at the north end 1 (Gaynor et al. 2007). Fewer wolves were detected at Red Rock and up the Illecillewaet drainages in 2010 compared with previous years. We place less importance on the fluctuations of wolf numbers at Red Rock, Illecillewaet and Jordan over the short term, as these sites that are external to the Lake Revelstoke Survey Area are likely used by wolves on a part-time basis. Nonetheless, it is surprising that we never detected wolves in the Illecillewaet drainage, given the high potential for scavenging road/rail wildlife mortalities in this area. The wolf population does not appear to have changed since 2009, and appears to be lower than in 2007 (Gaynor et al. 2007). Data from collared wolves in the study area showed relatively high levels of transience over the last year. A collared Goldstream wolf made a foray into Downie Creek in January/February 2010 which is uncommon. Two wolves were found starving in the Downie Creek area within 1 week of one another in September 2009 (1 was shot near the Downie RV resort and the other was a radio collared wolf found dead through VHF telemetry). Investigations into the GPS locations of 1 These were at an old moose kill across the lake from the Big Mouth. Note that the former Big Mouth pack was known to use both sides of Lake Revelstoke. 8

the lone wolf (pack 4) showed very little success in foraging earlier in the winter, or during the previous summer (2009) 2. These observations, which are anecdotal to the current survey, suggest that wolf numbers are being successfully depressed by ungulate management. In future, it would be helpful to conduct several surveys in one winter to provide estimates of precision using these methods. 6 References Apps, C.D. and McLellan, B.N. 2006. Factors influencing the dispersion and fragmentation of endangered mountain caribou populations. Biological Conservation 130: 84-97. Gaynor, C., van Oort, H., and Mowat, G. 2007. Predator surveys within Kootenay region mountain caribou recovery areas: data summary report. Prepared for the Government of British Columbia, Ministry of Environment, Nelson, BC. Meidinger, D.V. and Pojar, J. 1991. Ecosystems of British Columbia. Special Report Series 4. Report prepared by British Columbia Ministry of Forests, Victoria, BC. van Oort, H., Bird, C., Mowat, G., and De Groot, L. 2010. Winter predator census results in the Kootenay-Columbia Caribou Recovery areas from 2006 to 2009. Report prepared for Ministry of Environment, Nelson, BC, and the Columbia Basin Fish and Wildlife Compensation Program, Nelson, BC. van Oort, H., Gaynor, C., Bird, C., Mowat, G., and DeGroot, L. 2009. Winter carnivore surveys in the Kootenay's mountain caribou recovery areas; winter 2007/2008. Report prepared for the Government of British Columbia, Ministry of Environment, Nelson, BC. Wittmer, H.U., McLellan, B.N., Seip, D.R., Young, J.A., Kinley, T.A., Watts, G.S., and Hamilton, D. 2005. Population dynamics of the endangered mountain ecotype of woodland caribou (Rangifer tarandus caribou) in British Columbia, Canada. Canadian Journal of Zoology 83: 407-418. Wittmer, H.U., McLellan, B.N., Serrouya, R., and Apps, C.D. 2007. Changes in landscape composition influence the decline of a threatened woodland caribou population. Journal of Animal Ecology 76: 568-579. Wittmer, H.U., Sinclair, A.R.E., and McLellan, B.N. 2005. The role of predation in the decline and extirpation of woodland caribou. Oecologia 144: 257-267. 2 In addition, there were 2 other known wolf mortalities in the survey area during 2009-10: one large male was trapped on the west side of the reservoir and another wolf was killed on the highway near Bigmouth Creek. 9