Columbia River Project Water Use Plan. Kinbasket and Arrow Reservoirs Revegetation Management Plan

Transcription

1 Columbia River Project Water Use Plan Kinbasket and Arrow Reservoirs Revegetation Management Plan Implementation Year 1 Reference: CLBMON-11B -3 Revelstoke Reach Western Painted Turtle (Chrysemys picta bellii) Monitoring Program Study Period: Nicole Schiller MSc. Candidate Thompson Rivers University 900 McGill Rd, Kamloops, BC V2C 5N3 Office: (250) nrschiller@gmail.com Karl W. Larsen Associate Professor, Wildlife Ecology & Management Research Centre 204 Thompson Rivers University P.O. Box 3010, Kamloops, BC V2C 5N3 Phone: Fax: klarsen@tru.ca March 2011

2 2010-Annual Report - Revelstoke Reach Western Painted Turtle Monitoring Program Reference: CLBMON 11B3 Columbia River Project Water Use Plan, Arrow Lakes Reservoir Prepared for: BC Hydro Generation Water Licence Requirements 6911 Southpoint Drive Burnaby, BC Attn: Doug Adama Prepared by: Nicole Schiller and Karl W. Larsen Thompson Rivers University, 900 McGill Road Kamloops, British Columbia, V2C 5N3 2

3 Executive Summary Western painted turtles (WPT) (Chrysemys picta bellii) are the only native freshwater turtle in British Columbia. A northern population of WPT resides in the Upper Arrow Lakes Reservoir, British Columbia and was identified by BC Hydro as a component of the monitoring program CLBMON11B: Wildlife Effectiveness Monitoring of Revegetation and Wildlife Physical Works. CLBMON11B is a longterm monitoring program designed to identify opportunities and monitor the effectiveness of revegetation and wildlife physical works within the reservoir drawdown zone for various wildlife species including the WPT. Initial work on the WPT population residing in the Revelstoke Reach of the Arrow Lakes Reservoir began in the spring of The overarching objective of this study is to collect baseline information on the WPT population, and in doing so, provide a preliminary assessment of how and if BC Hydro s operations for the Arrow Lakes Reservoir influence the animals (particularly drawdown effects) and provide mitigation through physical works should this be necessary. The specific information collected focuses on the distribution, abundance, and demographics of the WPTs in the Arrow Lakes Reservoir. It also focuses on how the turtle s productivity varies across years, which type of habitats they use, and if any parts of these habitats are located within the drawdown zone. Information on nesting and overwintering locations are a priority, as these aspects of turtle ecology may be impacted directly or indirectly through winter mortality, nest inundation, predation risks, reservoir operations (e.g., water levels) and habitat alterations. A number of techniques were employed to collect the information, including mark-recapture and radio telemetry. We will soon be able to start addressing the central management questions, namely (1) Can minor adjustments be made to reservoir operations to minimize any impact to WPTs (2) Can physical works be designed to mitigate the impacts of reservoir operations on WPTs and (3) Would revegetation of the drawdown zone likely affect the availability and use of habitat by WPTs? The 2010 field season began in mid-april and went until the end of December. A total of 541 person hours and 1288 turtle sightings (excluding telemetry locations) over six months resulted in 2.38 sightings/hour. April was the month where the majority of turtles were detected in the reservoir. A total of 133 unique turtles were captured: 57 adults, 19 juveniles and 57 neonates. Of these animals, 3 were adult males, 54 were adult females, and 3 were juvenile females. A strong female-bias in the sex ratio of captured turtles was evident after the first field season. The trapping method and timing of most captures during the breeding season may have created the bias in the population. It may be possible that males are utilizing the habitat at different times or utilizing different habitat within the reservoir. Further research is required to determine the population status and abundance of the turtles residing in Revelstoke Reach. Six nesting locations were identified, Airport-Firebase, along the road at Montana Slough and Nichol Road, Williamson Lake and Turtle Pond. All nesting locations identified to date are located above the drawdown zone of the reservoir and all are anthropogenic in origin. On average 11 locations per animal were identified through telemetry on 17 individual turtles (16 females and 1 male). The longest movement detected between telemetry locations was 770 m, the average movement of turtles within the reservoir was 63 m and the greatest distance travelled by one individual was 10.4 km (turtle T65). Comparisons of the average distances moved by turtles outfitted with radio transmitters showed a significant difference between sites and between the comparison of changes in water levels defined as after flooding and before flooding. This difference may be due to variations in habitat types between the sites and the changes in water levels. 3

4 Data collected from the 2010 field season and the upcoming 2011 field season will provide baseline data crucial for future work and possible adaptive-management projects. It will also test the efficiency and feasibility of some of the techniques used to monitor and capture the turtles while providing the framework to a long-term monitoring strategy. 4

5 Acknowledgements We would like thank BC Hydro in providing the financial support for this research on the western painted turtle (Chrysemys picta belli) through Thompson Rivers University (TRU). Special thanks to Doug Adama (BC Hydro) who established the project and continues to assist in its development as well as being an active member of Nicole Schiller s Graduate Committee, Guy Martel who has provided support during project audits and on additional occasions and Purnima Govindarajulu who has provided support both as a member of N. Schiller`s Graduate Committee and the Western Painted Turtle Working Group. Field work was completed by Nicole Schiller and Tory Anchikoski of Thompson Rivers University. A special thanks to the volunteers who helped throughout the season, namely Darcy Schiller, Stephen Symes, and Philip Oreste. Cover Photo: Airport Marsh, Upper Arrow Lakes Reservoir, male western painted turtle. Photos Schiller, N. 5

6 Table of Contents Executive Summary... 3 Acknowledgements... 5 Table of Figures... 8 List of Tables... 9 Appendices Introduction Management Questions Scope of the Report Study Area Methods Surveys Public Reports and Surveys Visual Encounter Surveys Nesting and Emergence Surveys Trapping and Capture Active Trapping Passive Trapping Morphometric and Behavioural Data Mark-Recapture Photo ID Telemetry Transmitter Attachment Habitat Data Collection Temperature Data Loggers Nesting sites Environmental Stations Data Analyses Results General Observations Survey Effort Trapping Morphometrics Abundance and Numbers Sex Ratio Mortality & Observations Of Natural Predators Nesting

7 5.9 Human Development Telemetry Airport Marsh Montana Slough Turtle Pond Secondary Sites Movement Habitat Data and Turtle Activity Discussion During what portion of their life history (e.g. nesting foraging, and overwintering) do painted turtles utilize the drawdown zone in Revelstoke Reach? Which habitats do painted turtles use in the drawdown zone and what are their characteristics (e.g., pond size, water depth, water quality, vegetation, elevations band)? What is the abundance and productivity of painted turtles in the Revelstoke Reach and how do these vary across year? Sex Ratio Can physical works be designed to mitigate the impacts of reservoir operations on painted turtles? Basking Logs Nesting Site Conditions Management Questions Four, Five, Seven and Eight Upcoming work References Personal Communications

8 Table of Figures Figure 1. Arrow Lakes Reservoir, Revelstoke Reach British Columbia Canada, Western Painted Turtle Study Location. Study areas are sites that were surveyed using boats and had turtles sighted without the use of telemetry while study sites were areas survey by land which did not have sighting of turtles other than detection by telemetry. (Galdames 2011) Figure 2. Turtle nest emergent hole as shown by the red arrow Figure 3. Floating basking trap (Schiller 2010) Figure 4. Basking trap locations within Revelstoke Reach (Galdamez 2011) Figure 5. Turtle notching scheme. The marginal plates are numbered 1 to 12 on each side on the turtle (adapted from the Resource Inventory Committee of British Columbia 1998) Figure 6. Survey of western painted turtles by month and site, Revelstoke Reach, Figure 7. Density map of turtle sightings and captures in Montana Slough (Galdamez 2011) Figure 8. Density Map of turtle sightings and captures in Airport Marsh (Galdamez 2011) Figure 9. Density Map of turtle sightings and captures in Turtle Pond. Density bubbles indicates location of the observer. Turtle Pond s area is approximately 2.5 ha with the majority of sightings occurring in the middle of the pond (Galdamez 2011) Figure 10. Baited hoop net trap (Heinsohn s 2010) Figure 11. Plastron length (cm) of the different age classes of turtle captured. * indicate outliers within the data Figure 12. Plastron length of adult male (n=3) and female 9n=51) turtles captured. Females were significantly larger than males turtles (t = -6.71; df = 52; P < 0.00). * indicate outliers within the data Figure 13. Size frequency histogram of the straight plastron length (cm) of all turtles captured residing in the Revelstoke, British Columbia area separated by site and age class during the 2010 field season Figure 14. Nesting sites identified during the 2010 field season, Revelstoke Reach, Upper Arrow Lakes British Columbia, Canada (Galdamez 2011) Figure 15. Telemetry locations for turtle T65 Revelstoke Reach, Upper Arrow Lakes British Columbia, Canada (Galdamez 2011) Figure 16. Average distances moved per month by radio tagged turtles in Revelstoke Reach. * Indicates outliers in the data Figure 17. Average daily elevation levels per month.. * Indicates outliers in the data Figure 18. Comparison of average turtle movements before flooding and after flooding in Airport Marsh showing that greater movements were made after the water levels rose Figure 19. Comparison of average turtle movements after flooding between sites Airport Marsh and Montana Slough indicated that turtle movements were greater in distance at Montana Slough then Airport Marsh. However, smaller distance movements were more frequent in Airport Marsh then Montana Slough

9 List of Tables Table 1. Survey effort per month, Revelstoke Reach, All primary sites combined Table 2. Survey effort and sightings of western painted turtles by month and site, Revelstoke Reach, Primary sites (Airport Marsh (AP), Montana Slough (MS), Williamson Lake (WL), Turtle Pond (TP)) Table 3. Survey effort and sightings of western painted turtles by month and site. Secondary sites (Williamson Lake (WL), Cartier Bay (CB), 9 Mile, 12 Mile) Table 4. Summary of juvenile and adult turtles captured by site and age class Table 5. Habitat Categories of Turtles located using telemetry and visual encounter surveys in Revelstoke Reach. n= 311.number of observations Table 6. Turtle activities within the drawdown zone at Revelstoke Reach Table 7. Summary of radio-tagged western painted turtle (Chrysemys picta bellii) in Revelstoke Reach of the Upper Arrow Lakes, British Columbia Table 8. Average turtle movement made by each turtle outfitted with a radio transmitter

10 Appendices Appendix 1. Photos of Study Sites Appendix 2. Turtle Movements Appendix 3. Radio Telemetry Locations of Turtles by Month Appendix 4. A Table Outlining the Management Objectives, Questions, and Hypotheses of CLBMON 11B

11 1.0 Introduction The Columbia River Water Use Plan (WUP) was developed through a multi-stakeholder consultative process to inform how to best operate BC Hydro s Mica, Revelstoke, and Hugh Keenleyside facilities so that environmental values, recreation, power generation, cultural/heritage values, navigation, and flood control are balanced (BC Hydro 2005). The goal of the WUP is to accommodate these values through either incremental changes in water storage and release facilities, or to undertake physical works and wildlife mitigation in lieu of changes to reservoir operations. A population of western painted turtles (WPT) (Chrysemys picta belli) occupies the upper reach (Revelstoke Reach) of the Arrow Lakes Reservoir near Revelstoke, BC. These turtles are a provincially blue-listed species and the intermountain population is listed as Special Concern under Schedule 1 of the Species at Risk Act (SARA) (COSEWIC 2006). Due to the Special Concern status of the species, its regional importance (Maltby 2000; Larsen & Legebokow pers. comm. 2009), and the location of the population at the northern extent of its range, the species was identified as a species that may benefit from wildlife physical works (Golder Associates 2009; D. Adama pers. comm. 2009). The WPT study is a component of the program CLBMON11B Wildlife Effectiveness Monitoring of Revegetation and Wildlife Physical Works and includes a literature review (Reference CLBMON 11B3: Ecology and Conservation of Turtles in Canada with an Emphasis on painted turtles (Chrysemys picta)). This study on painted turtles (CLBMON11B3) is a long-term monitoring program designed to assess the population of turtles residing in Revelstoke Reach, the effectiveness of revegetation and wildlife physical works in the Arrow Lakes Reservoir drawdown zone and to increase habitat availability for various wildlife species. 1.1 Management Questions A number of management questions and hypotheses have been proposed by BC Hydro to be addressed over an eleven-year period, all specific to understanding the life history of the WPT. Through CLBMON11B3, the following management questions will be addressed over this eleven-year period (BC Hydro 2009): Life History 1) During what portion of their life history (e.g., nesting, foraging, and overwintering) do painted turtles utilize the drawdown zone in Revelstoke Reach? 2) Which habitats do painted turtles use in the drawdown zone and what are their characteristics (e.g., pond size, water depth, water quality, vegetation, elevation band)? 3) What is the abundance and productivity of painted turtles in Revelstoke Reach and how do these vary across years? 4) Does the operation of the Arrow Lakes Reservoir negatively impact painted turtles directly or indirectly (e.g., mortality, nest inundation, predation, and habitat change)? Mitigation 5) Can minor adjustments be made to reservoir operations to minimize the impact on painted turtles? 6) Can physical works be designed to mitigate the impacts of reservoir operations on painted turtles? 11

12 Effectiveness monitoring 7) Does revegetating the drawdown zone affect the availability and use of habitat by painted turtles? 8) Do wildlife physical works (e.g. habitat enhancement) affect the availability and use of habitat in the drawdown zone by painted turtles? Management hypotheses that will be tested throughout the program include: HO 1 : Painted turtles are not dependent on habitats in the drawdown zone of Arrow Lakes Reservoir. HO 2 : The operations of the Arrow Lakes Reservoir do not affect painted turtle survival or productivity. HO 3: Habitat enhancement through revegetation or physical works does not mitigate the effects of reservoir operations on painted turtles. More specifically, wildlife physical works and revegetation projects do not change the utilization of the drawdown zone habitats by painted turtles in Revelstoke Reach. To initiate a long-term monitoring program, BC Hydro has funded a two-year pilot study with Thompson Rivers University. This initial study will aid in developing a long-term monitoring framework, and to understanding the relationship between WPTs and the reservoir, their productivity, abundance, and habitat use. The study may also generate tentative recommendations for enhancing habitat for the turtles residing in the reservoir. 1.2 Scope of the Report This report provides a preliminary assessment of the turtles residing in Revelstoke Reach using data collected during the 2010 field season (April to October 2010) An assessment of the status of the Arrow Lakes Turtle population will require a long-term data set to determine population trends and changes over time. This reports aim is to summarize the methods and results from the first field season, in Further analyses and recommendations are to be provided at a later date when adequate data have been collected. 2.0 Study Area The Arrow Lakes Reservoir is the portion of the Columbia River that was created in 1968 with the construction of the Hugh Keenleyside Dam near Castlegar, BC. The reservoir is also influenced from outflows at Revelstoke Dam, constructed in The area is associated with the southern interior mountains forest region, and the biogeoclimatic zone within the study area is the Interior Cedar Hemlock moist warm (ICHmw3) (Braumandl and Curran 1992). The reservoir is divided into the Upper and Lower Arrows Lake spanning 230 km, between the Monashee Mountains to the west and the Selkirk Mountains to the east. The mountains rise to an elevation of 2600 meters and are heavily forested. Revelstoke Reach is the area of importance and is approximately 40 km long. The current water licence allows for a 20 meter (420 m m) fluctuation in water levels within the so-called drawdown zone, and annual reservoir levels vary both in time and in magnitude (BC Hydro 2005). These changes in water levels cause seasonal flooding of riparian, wetland, and grassland habitats used by aquatic and terrestrial organisms and likely impact the WPT. Designated study sites were along the east side of the reservoir and include primary sites Airport Marsh (81.5 ha), Turtle Pond (2.5 ha), Montana Slough (28.3 ha), and secondary sites include Williamson Lake, Cartier Bay, 9 and 12 Mile (Figure 1). Primary sites were chosen based on initial 12

13 surveys at the beginning of the 2010 field season, and on previous sightings as the majority of turtle sighting within the reservoir were made at Airport Marsh and Montana Slough (Adama pers. comm. 2010). Both primary and secondary sites were based on habitat quality as the secondary sites have less cover and more exposure to the main flow of the reservoir. In the case of Williamson Lake this site had good habitat however, high human disturbance as it is the local campground and recreational spot. Telemetry locations were also considered as turtles were only detected at 9 Mile and 12 Mile with the aid of telemetry. 3.0 Methods We used three main approaches to collect data in the field: visual encounters surveys, markrecapture, and radio telemetry. Mark-recapture and radio telemetry entailed capturing WPTs, which allowed for the collection of morphometric data, aging, and sex determination. An animal handling permit was obtained from the Ministry of Environment (Permit CB ) and an animal welfare permit was obtained through the Animal Care Committee at Thompson Rivers University (AUP R). For a more detailed protocol of the methods please see the companion document Arrow Lakes, Revelstoke Reach Western Painted Turtle Monitoring Program: Sampling Protocol (2011). 3.1 Surveys Public Reports and Surveys Communications with local naturalist clubs and public presentations helped promote awareness of the turtle research being conducted in the Arrow Lakes Reservoir. A specific account (turtlehotline@gmail.com) was established to promote reporting of public sightings and inquiries. During the 2010 field season, various news articles in the Revelstoke Times were published (including the address) and created positive feedback such as turtle locations and potential physical works in the area. One such article was a news article in The Province newspaper discussing the Revelstoke Reach monitoring program (February 11 th 2011). Stakeholder communication was facilitated by Jennifer Walker-Larsen and Greg Alexis of BC Hydro Visual Encounter Surveys Initial visual surveys, which are informal opportunistic searches, were conducted at the start of the field season to obtain preliminary data on where the turtles were located and where initial trap placement would be best. Once initial visual surveys were conducted, extensive weekly searches examining the entire area or the shorelines and ponds began and were continued until the end of September. Visual surveys followed the standard methodology described in Resource Inventory Standards (1998). Surveys were conducted on warm sunny days beginning in the early morning or late afternoon (Lefevre and Brooks 1995). Using binoculars, surveys began along known nesting areas, basking areas, then progressed along the shoreline and then into open water for a minimum of two person-hours at each site in the reservoir. In addition, aquatic vegetative islands such as bulrush (Scirpus lacustris) islands and inundated grassy areas were surveyed using a watercraft or wading through the area. It was generally difficult to approach a basking turtle within 20 meters within and outside the reservoir. This is likely an anti-predator strategy as turtles are the most agile in the water and camouflage well into their surroundings (Boyer 1965). 13

14 Figure 1. Arrow Lakes Reservoir, Revelstoke Reach British Columbia Canada, Western Painted Turtle Study Location. Study areas are sites that were surveyed using boats and had turtles sighted without the use of telemetry while study sites were areas survey by land which did not have sighting of turtles other than detection by telemetry. (Galdames 2011). 14

15 3.1.3 Nesting and Emergence Surveys Nesting surveys (May 10 th to July 26 th 2010) and emergent surveys (April 18 th to May 25 th 2010) were conducted by walking known nesting sites and areas that contained suitable nesting substrates. Nesting surveys were surveys looking for nesting turtles while emergent surveys were surveys looking for emergent holes (small holes in the soil with the approximate diameter of 5 cm) (Figure 2) where neonates leave their natal nest. In northern environments oviposting usually occurs when the temperatures are the warmest, usually in the late afternoon and evening (COSEWIC 2006). Therefore, nesting surveys were conducted at these times to optimize detection while emergent surveys were often led in the morning and late afternoon. Figure 2. Turtle nest emergent hole as shown by the red arrow. Known nests were monitored weekly through the active turtle breeding season (late May to early July; Schwarzkopf and Brooks 1987; COSEWIC 2006), ( and periodically throughout the summer and into the winter, to determine their status (intact, depredated or inundated). 3.2 Trapping and Capture Live trapping of turtles was required to conduct mark-recapture studies, affix radio-transmitters, and collect morphometric data. Trapping was conducted by searching and capturing animals and passively using basking traps Active Trapping Active trapping was performed using a dip net while walking along the shoreline, by wading or canoeing. The efficiency of this method depended largely on the physical properties of the study site and on time constraints Passive Trapping Basking traps are a type of floating trap consisting of a square frame often constructed from PVC, ABS pipe or wood with varying dimensions of less than a square meter (Figure 3) (Gamble 2006; Tran and Moorhead 2006). Initially, basking traps were moored into the study site without any mesh netting. This allowed a habituation period for the turtles to start using the traps for basking. When capture sessions were conducted, mesh netting was attached (usually in early morning) and the traps checked within a 12- hour period (Resource Inventory Standards 1998). Turtles bask on the central board of the trap and 15

16 plunge into the mesh bag upon leaving the basking site. The centre of the trap is open, which allows movement and room for turtle to surface and breathe. Baiting of basking traps with cat food or sardines has shown to increase capture rate. Basking traps are supposedly more effective early in the spring when the basking behaviour of the turtles is more pronounced (Resource Inventory Standards 1998). Basking traps were placed in areas where turtles were seen basking regularly at the beginning of the season and assessed use by initial visual surveys. Two traps were positioned in Turtle Pond, three in Montana Slough and three in Airport Marsh (Figure 4). Traps were moored into sites in late April to early May and left until mid-october. Later in the season, some traps were moved because water levels and habitat utilization changed. Figure 3. Floating basking trap (Schiller 2010). 3.3 Morphometric and Behavioural Data Morphometric data were collected on each turtle captured. Animals were weighed by placing them in a pillowslip and attaching it to a hand-held Pesola spring scale (to the nearest 0.1 g). The length and width of the plastron (bottom of shell) and carapace (upper shell) were measured as a straight line (Grayson and Dorcas 2004). The width of the carapace was measured from scute seven on either side of the turtle (P. Govindarajulu, pers. comm. 2010). Measurements were taken using appropriate sized calliper (to the nearest 0.1 mm) (Buhlmann and Vaughan 1991; St. Clair et al. 1994). The size averages of the turtles were reported as the plastron length rather than the mass to account for changes in mass through the retention or loss of eggs, and or water. In addition, plastron length was reported as a size indicator (Gibbons 1967). The activities of turtles were recorded as basking, walking, mating, nesting, stationary, swimming or unknown. Turtles (other than hatchlings and juveniles) were sexed by using secondary sexual characteristics and the relative positioning of the cloaca (McCartney and Gregory 1985; Matsuda et al. 1996). 16

17 Figure 4. Basking trap locations within Revelstoke Reach (Galdamez 2011). 17

18 3.4 Mark-Recapture Each turtle captured was permanently marked by notching the marginal scutes of the shell with a triangle file, allowing for the individual identification of turtles in the mark-recapture study (Cagle 1939). Neonates and juveniles were not marked, as their shells are soft and not fully ossified, so notching could cause deformities (Resource Inventory Committee 1998). Usually, there are 12 marginal scutes on each side of the carapace, labelled from one to twelve, ignoring the central top scute (Figure 5). The carapace is divided into the right side and the left side when looking down at an upright turtle. Notching schemes for this project were recorded with the first number indicating notches on the left side of the carapace with commas separating the individual scutes. A dash indicates the separation between the sides of the carapace and the following number is the notches located on the right side of the carapace. Again, a comma separates the individual scutes notched. For example, 0-1 indicates there is no notching on the left side of the carapace and a notch on the right side of the carapace on scute 1 (Figure 5). The notching scheme 2,3-11 reads that there are notches located on scutes 2 and 3 of the left side of the turtle as well as on scute 11 along the right side of the turtle s carapace. Figure 5. Turtle notching scheme. The marginal plates are numbered 1 to 12 on each side on the turtle (adapted from the Resource Inventory Committee of British Columbia 1998) Each turtle captured was given a Turtle ID, alongside its individual notching scheme, which consisted of the letter T followed by a number (i.e.,t1, T2...T82, T83). Each location after the initial capture was labelled with a location number such as, T1-01, T1-02, T1-03. If the turtle was released, the transmitter fell off, or mortality occurred, the last location was labelled as END (e.g., T1-END) Photo ID The colouration of the plastron is unique among individuals and provides another method to verify the identity of recaptured animals. With each turtle capture, a standard plastron photo was taken and added to a digital database. This database will be used to determine if individual turtles can be identified by their plastron and whether this pattern changes over time. The use of pattern-recognition software may enable long-term cataloguing and identification of individual turtles. 3.5 Telemetry Turtles were located weekly in the spring and summer using the triangulation and homing techniques using a Lotek Biotrack wide-band radio receiver ( MHz) (Lotek Wireless Fish and Wildlife Monitoring) and three prong yagi antenna to identify locations, habitat use, nesting locations and movements. 18

19 3.5.1 Transmitter Attachment In 2010, we used stainless steel wire to secure transmitters (SI-2F or AI-2F Holohil Systems Ltd. Transmitter, Ontario Canada) to the posterior of the carapace. Small holes were drilled along the marginal scutes (usually scutes nine and eleven on the left side of the carapace to minimize interference with breeding) with a cordless power drill (Grayson and Dorcas 2004). Due to the variability in turtle weight, size, age and the unknown demographics of the population within the reservoir, three sizes of transmitters were used to accommodate a range of animal size. As per guidelines set forth by the Canadian Council on Animal Care (CCAC) (2003), the entire transmitter package did not exceed 5% of body weight. Epoxy putty was used to streamline the edges of the attachment to prevent snagging on vegetation (Edge et al. 2009; J. Litzgus, pers. comm. 2009). 3.6 Habitat Data Collection A site assessment was conducted for each turtle identified through telemetry, systematic searches, or incidental observations. This assessment consisted of a circular vegetation (aquatic and or terrestrial) plot (5.64 m radius) to evaluate the percentage of emergent, submergent, floating-leaved vegetation, shrubs, forest, grass and herb, coarse woody debris (CWD) >5 cm, and CWD <5 cm. Other variables measured included: location (marked with GPS), time, date, water depth, water and air temperature at turtle location, elevation, precipitation, wind, cloud cover, distance to water/shore, activity of the turtle, landscape context (Reservoir, Road, Forest, Turtle Pond, Other), and habitat type (Marsh, Open Water, Shoreline, Pond, Floating Fen, Bulrush, Creek, Gravel) (Resource Inventory Standards 1998; Marchand and Litvaitis 2004). 3.7 Temperature Data Loggers Nesting sites Temperature data loggers were used to augment the information collected on the nesting of the turtles. Using ibutton temperature data loggers (Maxim Innovation Inc. Company) nine were buried in the ground at Red Devil Hill, three in a shaded area, three within the main nesting area and three on the grassy slope below the nesting site. At the Airport-Firebase nesting site six ibuttons were installed, three within the nesting area and three just outside the nesting area in the compact soil. The data loggers were installed in September and will record the temperature ( C) every four hours. (Hughes, 2009; pers. comm. P. Govindarajulu 2010). Each data logger was placed in a plastic bottle to avoid water damage and buried at a depth of ten cm, the approximate depth of a turtle s nest. A small wooden stake was placed beside each ibutton and UTM coordinates taken at each nest to facilitate site location and collection of data loggers the following season Environmental Stations Environmental stations were established from April to September to collect ambient air and water temperatures to compare to turtle activity on basking traps in Montana Slough (three stations), Airport Marsh (three stations) and Turtle Pond (two stations). At each location, an anchor such as a brick was tethered to a buoy and placed in the water. One ibutton sat on top of the brick approximately 7 cm above the bottom substrate, the approximate height of a turtle (Edge et al. 2009), and another ibutton was above the surface of the water. 4.0 Data Analyses Survey effort was calculated as the numbers of sighting or captures (not including telemetry locations) per person hour by each site and month. Due to the limited number of turtle captures and recaptures, the population of turtles is expressed at this time as the Minimum Number Known Alive (MNKA) and broken down into site by age class and sex. However, continued data collection may enable a longterm monitoring program to use the Jolly-Seber method, which estimates abundance, survival, and recruitment rates (Krebs 1999). However, the feasibility of this is yet to be determined. 19

20 To describe the demographics of the turtle population, the capture and morphometric data were used to report the average size of the turtles using plastron length and categorized each turtle into an age and class. Sex ratios were compared with a Chi Squared Goodness of Fit test and size-frequency histograms were constructed to assess the current population structure (McCartney and Gregory 1985; Larsen and Gregory 1989). Average distances moved by turtles were compared using non-parametric Kruskal Wallis or Mann Whitney U tests as the data did not follow normal distribution. Comparisons of turtle movements were made before water levels rose and during within the reservoir to test the effects of water levels on turtle movements. June 10 th, 2010 was designated the date where the comparisons were made separating the categories before flooding and after flooding. This date was chosen as the reservoir water levels were observed beginning to rise on June 3 rd, 2010 and reached an elevation of m on June 10 th Percent predation was determined by dividing the number of nests predated by the estimated number of reproductive females captured, assuming that each female laid one nest. A nest was determined to be predated on if it was dug up and the eggs destroyed. Nesting sites were not monitored to determine number of nests as the probability of disturbance was high and the scope of the project was to identify nesting sites not monitor their productivity. Data from the temperature data loggers were not included because a large number of them failed or were programmed incorrectly. These problems will be corrected for the field season in Results 5.1. General Observations The three general areas where turtles were observed (without the use of telemetry) were Airport Marsh, Montana Slough and Turtle Pond (Figure 1; Appendix 1. Site Photos). Turtles in Turtle Pond were abundant and easily observed, making this site the main contributor to the data set. Turtle Pond is approximately 2.5 ha in size, is located upslope from the reservoir proper and is nestled in amongst houses and other development. The riparian area surrounding the pond is relatively free of vegetation, allowing for a clear line of sight. Neighbouring residents have positioned four large floating logs in the pond that are used regularly by turtles for basking. Airport Marsh is an area in the reservoir dominated by bulrushes and pondweed (Potamogeton spp), especially during the summer months when the latter species has had a chance to grow (Appendix 1: Site Photos). This area is relatively large (approx ha) and visibility is poor due to emergent vegetation such as bulrushes and pondweed: for much of the active season, turtles most often were only heard as they left basking sites and dived into the water. By late summer, turtles in Airport Marsh more frequently were seen basking in the water under the cover of the pondweed with only the top of their head showing. Montana Slough is a large wetland complex adjacent to Airport Way Road. This area exists as a functional wetland that floods as reservoir levels rise. A portion of the wetland is a massive floating island of vegetation that provides unique turtle habitat. The wetland s dominant vegetation is Sphagnum spp., Salix spp., and Phalaris spp. (red canary grass). When water levels are high, small openings within the vegetation of the floating island are exposed and turtles access these locations by presumably swimming under the island. Not only does this provide habitat during high water but it also appears to provide cover. As with Airport Marsh, turtles here were often heard leaving their basking 20

21 sites rather than seen, and a poor line of sight and a large area to survey (28.3 ha) contributed to poor detection. 5.2 Survey Effort A total of 541 person hours and 1330 turtle sighting (excluding telemetry locations) over six months provided 2.46 sightings/hour across all sites. Turtles were sighted from April to September (Table 1) and only detected by telemetry thereafter. The detection of turtles differed significantly among months (x 2 =43.5; df=5; P<0.01; Figure 6).The majority of turtles sighted and/or captured within the reservoir were during the month of April (Tables 2 and 3) with the bulk of the sightings occurring in May at Turtle Pond. The highest number of turtles detected at one time (n=54) occurred on May 18, 2010 at Turtle Pond. Sightings began to decline in the reservoir in June and no sightings were made at the secondary sites (without telemetry) until late fall when two juvenile turtles were spotted basking in Williamson Lake (Table 3). Table 1. Survey effort per month, Revelstoke Reach, All primary sites combined. Primary Sites Combined (Airport Marsh (AP), Montana Slough (MS), Williamson Lake (WL), Turtle Pond (TP)) Month sightings hours sighting/hour April May June July August September Monthly average 7.55 Table 2. Survey effort and sightings of western painted turtles by month and site, Revelstoke Reach, Primary sites (Airport Marsh (AP), Montana Slough (MS), Williamson Lake (WL), Turtle Pond (TP)). Number of Surveys Sightings by Site Survey Hours by Site Month AP MS TP AP MS TP AP MS TP April May June July August September

22 Figure 6. Survey of western painted turtles by month and site, Revelstoke Reach, Table 3. Survey effort and sightings of western painted turtles by month and site. Secondary sites (Williamson Lake (WL), Cartier Bay (CB), 9 Mile, 12 Mile). Number of Surveys Sightings Survey Hours Survey Effort - Sightings/Hour Month WL CB 9 Mile 12 Mile WL CB 9 Mile 12 Mile WL CB 9 Mile 12 Mile WL CB 9 Mile 12 Mile April May June July August September During the 2010 field season, 130 unique turtles were captured within the reservoir and 3 outside of the reservoir. Of these, 57 were adults, 19 were juveniles and 57 were neonates (from nesting sites) (Table 4). The total amount of time allocated to this capture effort (not including neonate captures) was 541 person-hours, for a capture rate of 0.14 turtles for every hour of searching (or on average seven hours per turtle detection). A greater density of turtles were captured (excluding neonate captures at nesting sites) in Airport Marsh (0.785 turtles/ha) then in Montana Slough (0.247 turtles/ha); however, Turtle Pond had the greatest density at 21.6 turtles/ha based on sightings (Figures 7, 8 and 9). 22

23 Table 4. Summary of juvenile and adult turtles captured by site and age class. Site Juvenile Female Juvenile Male Juvenile Unknown Age Class Adult Female Adult Male Adult Unknown Neonates Airport Marsh (including Red Devil Hill) Montana Slough Turtle Pond Williamson Lake Total

24 Figure 7. Density map of turtle sightings and captures in Montana Slough (Galdamez 2011). 24

25 Figure 8. Density Map of turtle sightings and captures in Airport Marsh (Galdamez 2011). 25

26 Figure 9. Density Map of turtle sightings and captures in Turtle Pond. Density bubbles indicates location of the observer. Turtle Pond s area is approximately 2.5 ha with the majority of sightings occurring in the middle of the pond (Galdamez 2011). 26

27 5.3 Trapping Traps were positioned in the water an average of 15 days prior to being initially armed for trapping. Thereafter the traps were left in position unarmed to increase habituation. In total, the traps were only set to capture turtles for approximately 40 hours because turtles were not seen using the trap platforms. Over the four month period, only three turtles in total were observed utilizing the basking trap located outside of the reservoir (Turtle Pond) and only two turtles were captured (3% of the total number of turtles captured) using the traps within the reservoir (Montana Slough). Hence, the methods that proved most effective in capturing turtles during the 2010 field season were simple hand captures and the use of a dip net. Further testing of the efficiency of basking traps and the feasibility of using them in a mark-recapture study will continue throughout the 2011 field season with the addition of another type of trap, the hoop trap (Figure 10). These will be used in conjunction with last season s basking traps, which were left in the reservoir over winter to increase the habituation time of the basking platforms. Figure 10. Baited hoop net trap (Heinsohn s 2010) 5.4 Morphometrics The average straight-line plastron length was used as a standard for reporting turtle size between age classes (Figure 11) and sex (Figure 12) (Gibbons 1967) Plastron Length (cm) Neonate Juvenile Adult Age Class Figure 11. Plastron length (cm) of the different age classes of turtle captured. * indicate outliers within the data. 27

28 22 20 Plastron Length (cm) Female Sex Male Figure 12. Plastron length of adult male (n=3) and female 9n=51) turtles captured. Females were significantly larger than males turtles (t = -6.71; df = 52; P < 0.00). * indicate outliers within the data. 5.5 Abundance and Numbers The Revelstoke Reach population is composed of a MNKA of 130 turtles (of all age and sex classes). The data collected from the 2010 field season illustrate that the population has a larger number of adult and neonate turtles than juveniles, and more specifically the reservoir population has a larger number of mature females than males (Figure 13). Outside of the Revelstoke Reach Population in Turtle Pond, the maximum number of turtles sighted during visual surveys was 54 individuals. Thus, the MNKA estimate at this time is 54 in Turtle Pond (no neonates were sighted, so this number is quite likely an underestimate). 28

29 50 40 Age Class Adult Adult Juvenile Juvenile Neonate Sex Female Male Female unknown unknown Frequency Plastron Length (cm) Figure 13. Size frequency histogram of the straight plastron length (cm) of all turtles captured residing in the Revelstoke, British Columbia area separated by site and age class during the 2010 field season. 5.6 Sex Ratio Fifty-four adult turtles were captured within the reservoir, consisting of 51 females and 3 males (the latter all at Airport Marsh), a female biased sex ratio (x 2 =42.7; df=54; P<0.01), which also has been detected in southern BC (Macartney and Gregory 1985). 5.7 Mortality & Observations Of Natural Predators No dead marked or radio tagged turtles were found. However, three turtles were found dead; one adult, age unknown, was found intact along the east side of the airport runway and two additional adults were found, their shells in pieces on the northwest side of the airport runway. All mortalities seemed longstanding as the colours in the plastrons were faded and bleached by the sun. 5.8 Nesting Nesting surveys identified six nesting locations: Red Devil Hill, Airport-Firebase (inbetween the aviation club property and the wildfire firebase), Turtle Pond, Nichol Road, Williamson Lake and Montana Slough (Figure 14). The majority of the nests were discovered by finding emergence holes on Red Devil Hill and Airport-Firebase nesting sites, with 29 nests located on Red-Devil Hill and 19 located at the Airport-Firebase location. The other nest sites were discovered by walking the shoreline or from sightings made by the public. The total number of nests laid during the 2010 field season is unknown as only nesting sites were identified. However, there were 16 nests predated. This could amont to 30 % considering only the number of reproductive females captured within Airport Marsh and assuming that each one had laid a nest either at Red Devil Hill or Airport-Firebase as the numnber of nest laid during the 2010 season was not determined. Turtle Pond is suspected of hosting a large breeding population, as there were as many as 54 turtles located at this site. Confirmation of the nesting activity and productivity will occur during the 2011 season. 29

30 Figure 14. Nesting sites identified during the 2010 field season, Revelstoke Reach, Upper Arrow Lakes British Columbia, Canada (Galdamez 2011). Two study animals were dicovered nesting along roadways. Turtle (T21) was discovered nesting alongside a residental road (Nichol Road). Participation from the public helped confirm turtle activity in this area. Turtle T32 was discovered nesting along the shoulder of Airport Way at Montana Slough. 30

31 5.9 Human Development The Airport-Firebase nesting site is subject to high human disturbance as one of the Wildfire Bases is stationed next to the site. The firebase is aware of the sensitive habitat located on the property; however, there has been talk about leasing additional land, including the nesting site, to the firebase. Measures have been taken to inform the appropriate stakeholders but open communication should take priority between the stakeholders (Columbia Shuswap Regional District, the Base manager, Ministry of Environment, BC Hydro, and the public) Telemetry Seventeen turtles were outfitted with VHF radio transmitters throughout the course of six months. On average each turtle was located 11 times (Appendix 2) and the data points were considered independent from one another if separated by a 24 hour period (Compton et al. 2002) Airport Marsh Throughout the season, 11 turtles were outfitted with radio transmitters from Airport Marsh, ten adult females and one adult male (T79). Of these, two turtles (T74 and T79) carried transmitters in to the winter. Turtle T74 is an adult female and is the only turtle that moved from Airport Marsh to a water body outside of the reservoir, Williamson Lake. She has remained in the same location since, which may indicate either mortality or a winter hibernation site. Turtle T79 is the only male turtle outfitted with a radio transmitter. He was captured on September 22 nd 2010 and has remained in the area close to the aviation club property in Airport Marsh, even into the winter months Montana Slough Five turtles were outfitted with transmitters in Montana Slough. Turtles T32, T61 and T64 stayed close to the shoreline along the Airport Way road. Turtles T32 and T64 positioned themselves in a deep water (>2 m) pond with a thick organic mud bottom. They have been located in this pond from late fall and into the winter, indicating the pond may be a hibernation site. T65 is an adult female tagged at the north end of Montana Slough in an inundated grassy area and was last located at 12- Mile on October 14 th, 2010 (Figure 15). This location is the furthest south a WPT has been documented in Revelstoke Reach. 31

32 Figure 15. Telemetry locations for turtle T65 Revelstoke Reach, Upper Arrow Lakes British Columbia, Canada (Galdamez 2011). 32

33 Turtle Pond Three turtles were captured in Turtle Pond using a dip net: T8, T9, and T10. T9 was the only turtle outfitted with a transmitter. T9 was an adult female tracked from May 18 th to July 12 th. She did not leave Turtle Pond during that time. After her transmitter fell off due to rusting, only visual surveys were continued. Turtles T8 and T10 were not detected after their intial capture. This site is outside of the project focus area, but functions as a reference site Secondary Sites Williamson Lake, Cartier Bay, 9 Mile, and 12 Mile were considered secondary sites and were surveyed if time and weather permitted. No turtles were discovered at these sites during visual and systematic searches, but turtle T65 was recorded on the west side of the reservoir, 9 mile and 12-mile through telemetry Movement In general, turtle movements were observed by month and are summarized as follows: Through May and June turtles in Airport Marsh moved towards the shoreline and the nesting sites at Red Devil Hill and the airport firebase (Appendix 3, Map 1 and 2). From June and into July, turtle movement in Airport Marsh was still close to shorelines but locations were beginning to be discovered north (Appendix 4: Map 2). In Montana Slough the turtles moved closer to shore (Appendix 3: Figure 3). Movements from July into August were not as detailed due to the loss of some transmitters, but movements within Airport Marsh seemed simlar to movement from June and July (Appendix 3: Map 3). In Montana Slough, the turtles moved eastward where there was more influence of the reservoir ( Appendix 3: Map 4). In August to September, movements were small and typically towards the shoreline. The majority of the turtles then became stationary suggesting that they have reached their hibernation sites ( Appendix 3: Map 4 and 5). Movements from October to December were detected using the triangulation technique and considered stationary. More specifically, the first year of telemetry identified turtle movement between Airport Marsh and Williamson Lake, as shown by turtle T74, which moved into Williamson Lake in the fall between the days of September 14 th 2010 to September 23 rd In addition, the movements of turtle T65 shows that turtles can move as far south as the 12 Mile location. Anecdotal evidence from residents of Sunnyside Road (the access road to Turtle Pond) supports the notion that there is movement Turtle Pond and Airport Marsh, as well as from Turtle Pond to Williamson Lake (J. Maitre, H. Tarzwell, and S. Vandermeer pers. comm. 2010). One sighting occured on September 14 th 2010, where an adult turtle was seen crossing the road near 3 rd Street and Garden Avenue between 12:30 and 13:00 in the town of Revelstoke (A. Guidinger pers. comm. 2010). The average daily distance moved by turtles outfitted with transmitters was 63.3 m (n=164, SD=124) with the largest movements occurring in September (Figure 16). It appears that as the water levels rose in July to maximum pool the average daily distances moved by the turtles gradually deceased and increased in August (Figures 16 and 17).The longest daily movement observed was 770 m. The greatest total distance traveled by an individual turtle (T65) was 1040 m, observed between July 30 and Oct 14, Overall, there was a significant difference between the average daily distances traveled by radio tagged turtles (Kruskal Wallis= 50.3, df = 15, P < 0.001); and a significant difference 33

34 in daily movements between turtles in Airport Marsh ( = =33.93 m, n=103, SD=48.17) and Montana Slough ( = =86.3m, n=63, SD=157.1) (Mann Whitney U-Test = -8.81, upper and lower confidence limits , -1.35, P=0.002). Figure 16. Average distances moved per month by radio tagged turtles in Revelstoke Reach. * Indicates outliers in the data. Figure 17. Average daily elevation levels per month. * Indicates outliers in the data. 34

35 Comparisons between changes in water levels defined as before flooding (May 1 st to June 10 th 2010) and after flooding (June 11 th to October 14 th, 2010) found a significant difference in Airport Marsh (Figure 18: Mann Whitney U-Test = 1387, 95% CI: 6.18, 38.65, P=0.003) and demonstrated greater movements after a rise in water levels. A comparison was not made between water level changes in Montana Slough as not enough data were available. However, a comparison between Airport Marsh and Montana Slough during the after flooding event was significant (Figure 19: Mann Whitney U-Test = 5187, 95% CI: 0.99, 19.12, P=0.02) Frequency Before After Distance Intervals (m) Airport Marsh Figure 18. Comparison of average turtle movements before flooding and after flooding in Airport Marsh. 35

36 80 Frequency Airport Marsh >1000. Distance Intervals (m) After Flooding >1000 Montana Slough Figure 19. Comparison of average turtle movements after flooding between sites Airport Marsh and Montana Slough indicated that turtle movements were greater in distance at Montana Slough then Airport Marsh. However, smaller distance movements were more frequent in Airport Marsh then Montana Slough Habitat Data and Turtle Activity Each sighting location or turtle capture was characterized into a habitat type and reflected into nine categories (Table 5). The dominant habitat type was open water followed by shoreline. Often turtles were detected swimming or basking along the shorelines of open water or areas where there were sufficient amounts of cover (i.e., Cyperaceae spp., dead organic matter). The activity detected most often was basking, occurring on the shoreline, in the water, or on vegetation or in the mud (Table 6). However, the majority of turtles detected in Turtle Pond were basking on four large basking logs present in the pond. 36

37 Table 5. Habitat Categories of Turtles located using telemetry and visual encounter surveys in Revelstoke Reach. n= 311.number of observations. Habitat Type within the 5.64 meter radius plot N Percent Open Water Shoreline Gravel Marsh Inundated Land Bulrush Floating Fen Pond Creek Table 6. Turtle activities within the drawdown zone at Revelstoke Reach. Activity Within the Reservoir Outside the reservoir N Percent N Percent Basking Swimming Stationary Nesting Walking on Land (not related to nesting) Dead Mating Unknown Discussion This preliminary assessment of the painted turtle population began in April 2010 and will continue into 2011 by assessing habitat use and population metrics. Direct or indirect impacts from the reservoir operations on the painted turtle are unknown thus far. Nonetheless, this discussion and the literature conducted provides a preliminary assessment of the first year of data collection in regards to answering the management questions of CLBMON 11B3. Further exposure through the newspaper, presentations and posters will continue into the 2011 field season with the anticipation of positive feedback. Appendix 4 provides a summary of the objectives, management questions, hypotheses and measures that have been taken to address them. 37

38 6.1 During what portion of their life history (e.g. nesting foraging, and overwintering) do painted turtles utilize the drawdown zone in Revelstoke Reach? Visual and systematic surveys and telemetry were conducted from April 19 th to October 14 th The majority of turtles sighted and/or captured within the reservoir were during the month of April. Continued surveys determined that all age classes of turtles utilized areas within the drawdown zone throughout the season. However, information regarding male and juvenile turtles remains sparse as only three male turtles were detected. Based on the results form 2010, turtles were present within the drawdown zone throughout all seasons. Six nesting sites were idenfied all from Airport Marsh North. Red Devil Hill and Airport- Firebase sites have a considerable sized breeding population while the Turtle Pond nesting site is assumed to be productive but confirmation is pending the 2011 nesting season. All nesting sites currently identified have been outside of the drawdown zone but have been anthropogenic in origin. Nesting surveys south of Airport Marsh will be a priority for the 2011 nesting season, in an effort to identify additional sites that may be subject to the fluctuation in water levels. 6.2 Which habitats do painted turtles use in the drawdown zone and what are their characteristics (e.g., pond size, water depth, water quality, vegetation, elevations band)? Turtles were sighted and captured in many types of habitats within the drawdown zone. The majority were located in open water, followed by the shoreline habitats of floating mass of vegetation, bulrush islands and marshes. These areas provide cover as well as close access to the water. A small number of known turtles (Approx. 9-11) known to reside in Montana Slough have been sighted using pond habitat within a wetland complex and grassy areas to the east within the drawdown zone. However, when the area becomes inundated a portion of the vegetation rises with the water levels creating a mass of floating vegetation and changing the habitat availability. This area is unique as it provides habitat and cover for the turtles when the majority of the area is inundated. In Airport Marsh, turtles were sighted basking within or on islands of bulrush vegetation (Cyperaceae spp) or in areas where there were significant amounts of dead organic matter (DOM) in which they could hide while still basking. Seasonal movements detected were related to nesting and potential movements north or inward to the shoreline in response to the changing water levels (Appendix 3). Differences were detected between average turtle movements within Airport Marsh and between Airport Marsh and Montana Slough. In addition, differences in water levels defined as before flooding and after flooding found deviation in turtle movements. This variation may in part be explained by differences in habitats at the two sites and the exposure of these sites to reservoir levels: Montana Slough is located at an elevation level of 436 m and Airport Marsh at 438 m and is sheltered by the Columbia River by the Airstrip. Further research and survey effort are required to determine which habitats are utilized and whether there is a difference in space and time. 6.3 What is the abundance and productivity of painted turtles in the Revelstoke Reach and how do these vary across year? Given the lack of mark-recapture data during the 2010 season, estimates of population size or abundance are impractical. Age-class structure data therefore provide the most useful information at this time for commenting on the status of the population, following similar interpretations by Schreiber and Schreiber (1983) and Macartney and Gregory (1985). 38

39 Based on the 2010 data, the Revelstoke Reach population consists of a large number of adults and neonates and more specifically, more reproductive females than males. Turtle populations consist of more adults than juveniles as there are high rates of mortality in juveniles and neonates. The population has at least two productive nesting areas within the reservoir (Airport-Firebase and Red Devil Hill) and one outside the reservoir at Turtle Pond. The data are limited and display a female bias, and clearly more than one year of study is required to determine the population status and trend, especially in long-lived species where lag times are present (Resource Inventory Standards 1998) Sex Ratio The reason(s) for the strong female bias in the adult sex ratio is not clear at this time. The main method of capture, hand or dip net captures may be more effective in capturing females, particularly as females move and come to shore to nest. Another possible explanation may be that male turtles are utilizing habitats at different times or other habitats within the drawdown zone and not detected. A large number of reproductive females is a good indicator that the population is recruiting. This is particularly important in a species that has high rates of juvenile mortality such as is assumed for WPT populations (COSEWIC 2006). Macartney and Gregory (1985) working in Kikomun Creek Provincial Park (south-eastern BC) also reported a female skewed sex ratio, with most individuals being reproductively-sized females. That population s sex ratio was not as skewed as in Revelstoke Reach (62:38 vs. 94:06) and was based on a large sample size. Although having a large number of females is important to the long-term survival of the population, a low number of male turtles may contribute to a low reproductive rate or low genetic diversity within the population (Parker and Whiteman 1993). Further research is required to determine whether the bias is influenced by the trapping method or differential behaviour on the part of the males. If a female bias truly exists, future studies may wish to look into the genetic diversity of the population as well as the mechanisms responsible for the bias. 6.4 Can physical works be designed to mitigate the impacts of reservoir operations on painted turtles? Basking Logs Basking is an important activity as it increases the body temperature of the turtle promoting digestion, vitamin D production, and shedding of the skin, which inhibits the growth of algae and the attachment of external parasites (Macartney and Gregory 1985; Schwarzkopf and Brooks 1985).To date there are no data that suggest that basking sites are limited within the reservoir. Turtles respond to rising water levels by moving closer to shore and utilizing new shoreline, floating vegetation or logs. However, whether these sites afford different levels of protection or basking opportunities is not clear. The animals seemed unresponsive to the artificial basking sites (our traps provided within the 2010 field season. Nevertheless, the animals in Turtle Pond are readily utilizing basking logs originally provided by nearby landowners and the individual responsible for placing basking logs in Turtle Pond indicated that at least a year elapsed before the animals began using them as basking sites (Tarzell pers. comm. 2010). The placement of similar basking logs in the reservoir itself would in theory provide supplementary basking sites that would rise and fall with water levels, making them readily available over a greater time period. Whether the turtles would respond to this manipulation, and over what time frame, would be an interesting experiment. We will continue to monitor the response of turtles to artificial basking traps in the reservoir through 2011 to determine if usage rates increase Nesting Site Conditions Revelstoke Reach has two major nesting areas within the reservoir: Airport-Firebase and Red Devil Hill. Based on our knowledge to date, Red Devil Hill may be one of the more important habitats for 39

40 reservoir turtles, and should be maintained. This site is becoming encroached with vegetation and canopy cover, suggesting its suitability as a nesting site may be diminishing. Initial recommendations would be to periodically clear the herbaceous cover and clear trees to increase solar radiation in the area to optimize neonate development (Schwarzkopf and Brooks 1987: COSEWIC 2006) and increase nesting habitat. However, the long-term maintenance of this site, and its proximity to the road, suggest that an alternative nesting site should be considered for the long-term viability of this population. Annual rototilling of the Airport-Firebase nest site has been conducted by a resident living on the property (J. Van Dam, pers. comm. 2010). Rototilling has kept the soil loose with no establishment of herbaceous plants and minimal encroachment by roots of a nearby tree. All other areas surrounding the nesting site are compact and not suitable for nesting; additional rototilling of these areas may provide added nesting habitat. Continual rototilling of the area and possibly at Red Devil Hill would ensure that the soil does not become compact or that vegetation encroachment does not occur. A management activity would be to rotor till annually or biannually a portion of the areas by hand to decrease vegetation and soil compaction. Pulling vegetation by hand will ensure the survival of any undetected nests. Rototilling or hand pulling of vegetation would have to be done after nestlings have emerged and before adult females return to nest. However, intense monitoring and marking of the nests may be required to pinpoint this window in time ( roughly the first few weeks in May). Even so, in some years this window of time may be extremely short or even non-existent, as overlap between ovipositioning and emergence by the nestlings can occur. The resident who has been voluntarily rototilling this site anticipates moving out, so the long-term maintenance of this site is in question. 6.5 Management Questions Four, Five, Seven and Eight Results to date are very tentative, limiting our ability to address these questions. For example, nesting sites identified during the 2010 season were all located above the drawdown zone; however, additional nesting sites that have yet to be discovered may be subject to changes in water levels. To answer questions four, five, seven and eight requires knowing the population trend (something difficult to accomplish in two years) or failing that, whether there appears to be any indication that something is impacting the population like poor reproductive success (which could be compensated for by immigration) or winter mortality. After one field season, the data do not bring in evidence any aspect of the turtle population s ecology which may be affecting long-term viability (e.g., lack of successful nesting sites, constrained habitat use). However, there are some observations that may be potentially problematic, and that need further investigation: the long term viability of anthropogenic nesting sites a strong female bias in the sample size that may be real or an artefact of sampling feasibility of a mark-recapture study uncertainty about immigration and emigration rates from the reservoir and how they contribute to the long-term viability of the population, and uncertainty about the mortality rate in the population and what might be the causes of mortality During the 2011 field season, we will continue to address the female bias in the turtle population, the trap efficiency of the turtles (using basking traps and hoop traps) and the feasibility of using a markrecapture study to determine the population abundance and productivity. Identifying the long-term viability of the nesting sites, the immigration and emigration rates of the turtles, habitat enhancement 40

41 and possible physical works will require an extended monitoring program. The winter mortality rate will begin to be assessed this coming spring as the survival rate of the turtles with radio transmitters will be determined. 7.0 Upcoming work After the 2011 field season, data collected over the two-year pilot study will help assess how reservoir operations may or may not affect WPTs. Although the duration of this pilot project will not allow the identification of any long-term effects of the reservoir operations (e.g., long-term persistence of the animals, physical works and revegetation), it will provide baseline data crucial for future work and possible adaptive-management projects. The 2011 field season will incorporate the use of hoop traps to test the capture rate and feasibility of using them in conjunction with the mark-recapture study and estimating of the population. Nesting surveys will be concentrated in areas south of Airport Marsh and within the drawdown zone. The use of temperature data loggers will be a priority as well as attaching radio transmitters to male and juvenile turtles to collect habitat data and locations within the reservoir. 41

42 8.0 References BC Hydro Consultative Committee report: Columbia River Water Use Plan, Volumes 1 and 2. BC Hydro Power Corporation, Burnaby, BC. BC Hydro Terms of Reference: Columbia River Project Water Use Plan: CLBMON11B: Kinbasket and Arrow Lakes Reservoirs Revegetation Management Plan. Boyer D.R Ecology of the basking habit in turtles. Ecology. 46(1.2): Braumandl, T. F., and M. P. Curran A field guide for site identification and interpretation for the Nelson forest region. British Columbia Ministry of Forests, Victoria. Buhlmann, K.A., and M.R. Vaughan Ecology of the turtle Pseudemys concinna in the New River, West Virginia. Journal of Herpetology 25(1): Cagle, F.R A system of marking turtles for future identification. Copeia 1939(3): Canadian Council of Animal Care CCAC guidelines on: choosing an appropriate endpoint in experiments using animals for research, teaching and testing. Ottawa Ontario. [accessed January 2010]. Compton, B.W., J.M. Rhymer, and M. McCollough Habitat selection by wood turtles (Clemmys insculpta): an application of paired logistic regression. Ecology 83(3): COSEWIC COSEWIC assessment and status report on the western painted turtle Chrysemys picta bellii (pacific coast population, intermountain-rocky mountain population and prairie/western boreal - canadian shield population) in Canada. Vol. vii + 40 pp. Ottawa: Committee on the Status of Endangered Wildlife in Canada ( Edge, C.B., B.D. Steinberg, R.J. Brooks, and J.D. Litzgus Temperature and site selection by blandings turtles (emydoidea blandingii) during hibernation near the species northern range limit Canadian Journal of Zoology 87(9): Gamble, T The relative efficiency of basking and hoop traps for painted turtles (Chrysemys picta). Herpetological Review 37(3): 308. Gibbons, J.W Variation in growth rates in three populations of the painted turtle, Chrysemys picta. Herpetologica 23(4): Golder Associates Columbia River Project Water Use Plan; Reference: CLBWORKS-29A. Volume I: Arrow Lakes Reservoir Wildlife Physical Work feasibility Study; Study Period: Prepared for BC Hydro. Kamloops. British Columbia. Grayson, K.L., and M.E. Dorcas Seasonal temperature variation in painted turtles (Chrysemys picta). Herpetologica 60(3):

43 Hughes, E Draft protocol for data loggers at painted turtle (Chrysemys picta) nest sites. Prepared for the British Columbia Ministry of Environment (2009): ( Wildlfe/painted%20turtle/Monitoring/Forms/AllItems.aspx) Accessed from the Western Painted Turtle Working Group SharePoint site accessed on January Inwood, D Taking care of the turtles; B.C. Hydro studies impacts of reservoir on reptiles. Prepared by The Province Newspaper. February 11 th Available from ern+painted+turtle/ /story.html [accessed February 2011]. Krebs, C.J Ecological methodology. 2 nd edition. Benjamin Cummings, Menlo Park, California. 620 pp. Larsen and Gregory Population size and survivorship of the common garter snake, Thamnophis sirtalis, near the northern limit of its distribution. Holarctic Ecology 12: Lefevre, K. and R.J. Brooks Effects of sex and body size on basking behaviour in a northern population of the painted turtle, Chrysemys picta. Herpetologica 51(2): MaCartney, M. and P.T. Gregory The western painted turtle in Kikomun creek provincial Park. M.Sc. thesis, Department of Biology, The University of Victoria, British Columbia, Victoria, B.C. Maltby, F.L Summary report: painted turtle (Chrysemys picta) nest site enhancement and monitoring: red devil hill nest site at Revelstoke BC. Report to B.C. Hydro`s Columbia Basin Fish and Wildlife Compensation Programs. Revelstoke, B.C. Matsuda, B.M., P.T. Gregory, and D. M. Green Amphibians and reptiles of British Columbia. Victoria, British Columbia, Canada: Royal BC Museum. Marchand, M. N., and J. A. Litvaitis Effects of habitat features and landscape composition on the population structure of a common aquatic turtle in a region undergoing rapid development. Conservation Biology 18(3): Parker, G., and H.H. Whiteman Genetic diversity in fragmented populations of Clemmys guttata and Chrysemys picta marginata as shown by DNA fingerprinting. Copeia. 1993(3): Resource Inventory Standards Inventory for pond-breeding amphibians and painted turtles standards for components of British Columbia s biodiversity No.37. Prepared by the Ministry of Environment, Lands and Parks Resources Inventory Branch for the Terrestrial Ecosystems Task Force Resources Inventory Committee. Version 2. Victoria, British Columbia Canada. Heinshon s. Heinshon s Country Store Turtle and snake traps. Available from [accessed January 2010]. Schiller, N. and L.W. Larsen CLBMON-11B3 Arrow Lakes, Revelstoke Reach western painted turtle monitoring program: Sampling Protocol. Unpublished report by Thompson Rivers University, Kamloops, BC, for BC Hydro Generation, Water Licence Requirements, Burnaby, BC. 35pp. + Apps. 43

44 Schwarzkopf, L. and R.J. Brooks Application of operative environmental temperatures to analysis of basking behaviour in Chrysemys picta. Herpetologica. 41(2): Schwarzkopf, L. and R.J. Brooks Nest-Site selection and offspring sex ratio in painted turtles, Chrysemys picta. Copeia. 1987(1): St. Clair, R.C., P.T. Gregory, and J. M. Macartney How do sexual differences in growth and maturation interact to determine size in northern and southern painted turtles? Canadian Journal of Zoology 72(8): Tran, S.L., and D.L. Moorhead A note on effective basking trap size. Herpetological Review 37(3):

45 9.0 Personal Communications Adama. D. Natural Resource Specialist. BC Hydro - Water License Requirements. Golden, British Columbia Greg Alexis. Senior Media Relations Advisor. BC Hydro Vancouver, British Columbia Larsen. K.W. Associate Professor, Wildlife Ecology & Management Thompson Rivers University. Kamloops, British Columbia Legebokow, C. Ecosystem Officer. Ministry of Environment. Revelstoke, British Columbia Canada Litzgus, J. Associate Professor, Department of Biology. Laurentian University. Sudbury, Ontario, Canada Galdamez, J. Courtesy of Lands and Resources Agency - GIS Service. Ktunaxa Nation Council Guidinger, A. BScPT, BScKin, RCAMT, CGIMS. Helios Rehabilitation, Revelstoke, British Columbia Canada Van Dam, J. Resident on Airport Club Property, Revelstoke, British Columbia Canada Vandermeer, S. Landowner, Sunnyside Road, Revelstoke, British Columbia Canada Maitre, J. Landowner, Sunnyside Road, Revelstoke, British Columbia Canada Tarzwell, H. Landowner, Sunnyside Road, Revelstoke, British Columbia Canada Walker-Larsen, J. Stakeholder Engagement Advisor. BC Hydro. Revelstoke British Columbia

46 Appendix 1. Photos of Study Sites Photo 1. Turtle Pond (Schiller 2010). Photo 2. Turtle Pond (Schiller 2010). 46

47 Photo 3. Airport Marsh (Schiller 2010). 47

48 Photo 4. Airport Marsh (Schiller 2010). Photo 5. Montana Slough (Schiller 2010). Photo 6. Montana Slough, shoreline of floating fen wetland (Schiller 2010). 48

49 Appendix 2. Turtle Movements Table 7. Summary of radio-tagged western painted turtle (Chrysemys picta bellii) in Revelstoke Reach of the Upper Arrow Lakes, British Columbia. Turtle ID Capture Location UTM Coordinates Zone 11U Easting Northing Number of Locations T2 Airport Marsh T3 Airport Marsh T4 Airport Marsh T5 Airport Marsh T6 Airport Marsh T7 Airport Marsh T9 Turtle Pond T11 Airport Marsh T12 Airport Marsh T13 Airport Marsh T32 T47 T61 T64 T65 Montana Slough Montana Slough Montana Slough Montana Slough Montana Slough T74 Airport Marsh T79 Airport Marsh

50 Table 8. Average turtle movement made by each turtle outfitted with a radio transmitter. Name Mean Movement St. Dev. T T T T T T T T T T T T T T T T

51 Appendix 3. Radio Telemetry Locations of Turtles by Month. Map 1. Radio telemetry locations of turtles during the month of May 2010 (Galdamez 2011). 51

52 Map 2. Radio telemetry locations of turtles during the month of June 2010 (Galdamez 2011). 52

53 Map 3. Radio telemetry locations of turtles during the month of July 2010 (Galdamez 2011). 53

54 Map 4. Radio telemetry locations of turtles during the month of August 2010 (Galdamez 2011). 54

55 Map 5.. Radio telemetry locations of turtles during the month of September 2010 (Galdamez 2011). 55