Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

First year site fidelity and survival in reintroduced captive-bred Vancouver Island marmots (Marmota… Jackson, Cheyney L. 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2012_fall_jackson_cheyney.pdf [ 1.36MB ]
Metadata
JSON: 24-1.0072983.json
JSON-LD: 24-1.0072983-ld.json
RDF/XML (Pretty): 24-1.0072983-rdf.xml
RDF/JSON: 24-1.0072983-rdf.json
Turtle: 24-1.0072983-turtle.txt
N-Triples: 24-1.0072983-rdf-ntriples.txt
Original Record: 24-1.0072983-source.json
Full Text
24-1.0072983-fulltext.txt
Citation
24-1.0072983.ris

Full Text

FIRST YEAR SITE FIDELITY AND SURVIVAL IN REINTRODUCED CAPTIVEBRED VANCOUVER ISLAND MARMOTS (MARMOTA VANCOUVERENSIS)  by Cheyney L. Jackson  B.Sc., The University of Victoria, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2012  © Cheyney L. Jackson, 2012  Abstract The Vancouver Island marmot is a critically endangered sciurid endemic to British Columbia, Canada. By 1997, the species had collapsed to fewer than 150 individuals in total. Between 2003 and 2010, 301 captive-bred marmots were implanted with radiotelemetry transmitters and released to new sites and extinct colonies to supplement the wild population. I evaluated reintroduction success based on three short-term measures: first season fidelity to release site, active season survival from release to hibernation, and overwinter survival through early spring. I used generalized linear mixed models to evaluate the influence of sex and age, release practices and procedures, and the local and landscape-level attributes of release habitats on site fidelity and survival. Results suggested that poor overwinter survival has been the limiting factor in first-year reintroduction success. In all years, overwinter survival was lower for newly released captive-bred marmots than for their wild or previously-released counterparts. Release date was the variable most predictive of success, but was positively associated to site fidelity and active season survival and negatively associated to overwinter survival. Release date was the only predictor for overwinter survival, whereas site fidelity was also negatively impacted by the presence of resident females at the release site. Active season survival was highest for females, 2-year olds, and marmots released to sites with talus. I discuss these results in the context of current release practices for Vancouver Island marmots, and make recommendations for future release candidates, site characteristics, and release procedures.  ii  Preface The marmot data analyzed in this thesis was collected by many individuals over an eight-year period. The Marmot Recovery Foundation provided data on marmots released from 2003 to 2006, and I participated in collecting data on marmots released from 2007 to 2010. I was assisted in collecting data on vegetation and visibility characteristics at release sites by Rick Merriman, Chris White, Sean Pendergast, and Jerry MacDermott. The aerial imagery used to estimate habitat type cover was provided by the Province of British Columbia through the GeoBC Base Mapping and Cadastre Branch. Cutblock cover was estimated from spatial datasets that are the copyright property of Island Timberlands Limited Partnership, TimberWest Forest Corporation, and Western Forest Products Incorporated. Data were analyzed with their permission. I conducted my statistical analyses with the assistance of Richard Schuster (MSc), who designed the R code and Excel macros used to facilitate all-subsets analysis. My supervisor, Dr. Peter Arcese, provided guidance and input on my thesis. A version of Chapter 2 will be submitted for publication, with Dr. Arcese and Richard Schuster listed as co-authors.  iii  Table of Contents Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents ................................................................................................................... iv List of Tables .......................................................................................................................... vi List of Figures ....................................................................................................................... viii Acknowledgements ................................................................................................................ xi Chapter 1: Introduction ........................................................................................................ 1 1.1  Context ...................................................................................................................... 1  1.2  Research objectives ................................................................................................... 2  1.3  Study species ............................................................................................................. 2  1.4  Population history ..................................................................................................... 5  1.5  Recent literature ........................................................................................................ 5  1.6  Captive-breeding and recovery program .................................................................. 7  1.7  Releases and monitoring ........................................................................................... 8  1.8  Chapter descriptions.................................................................................................. 9  Chapter 2: Determinants of reintroduction success in captive-bred Vancouver Island marmots ................................................................................................................................. 10 2.1  Introduction ............................................................................................................. 10  2.2  Methods................................................................................................................... 12  2.2.1  Study species ....................................................................................................... 12  2.2.2  Reintroduction program ...................................................................................... 13  2.2.3  Release procedures.............................................................................................. 14  2.2.4  Definitions of success ......................................................................................... 15 iv  2.2.5  Data collection .................................................................................................... 16  2.2.5.1  Re-sighting .................................................................................................. 16  2.2.5.2  Habitat characteristics ................................................................................. 17  2.2.5.3  Statistical analyses ...................................................................................... 18  2.3  Results ..................................................................................................................... 20  2.3.1  Site fidelity .......................................................................................................... 20  2.3.2  Active season survival ........................................................................................ 22  2.3.3  Overwinter survival ............................................................................................ 23  2.3.4  Model assessment ............................................................................................... 23  2.4  Discussion ............................................................................................................... 24  2.4.1  Site fidelity .......................................................................................................... 26  2.4.2  Active season survival ........................................................................................ 28  2.4.3  Overwinter survival ............................................................................................ 30  2.5  Implications and recommendations ........................................................................ 34  Chapter 3: Conclusion ......................................................................................................... 50 References .............................................................................................................................. 57 Appendices ............................................................................................................................. 71 Appendix A - Representation of newly released captive-bred Vancouver Island marmots across categorical predictors for site fidelity, active season and overwinter survival. ....... 71 Appendix B - Relative importance, parameter estimates, unconditional standard errors, and 95% unconditional confidence intervals for all variables used to predict site fidelity, active season survival, and overwinter survival. Predictors with large standard errors (> parameter estimates) have little reliability. ......................................................................... 72  v  List of Tables Table 1. Summary of the number of captive-bred marmots released in each year of this study that met the criteria for analysis (see 2.2.4), and that were used to predict site fidelity, active season and overwinter survival (in bold). Overwinter survival had the smallest sample size because marmots that died during the active season were excluded, and additional marmots went missing between fall and the following spring. ............................................................. 37  Table 2. Descriptions, range of values, and predicted direction of influence for variables used to predict site fidelity, active season survival, and overwinter survival in newly-released, captive-bred Vancouver Island marmots (Marmota vancouverensis). Covariates were supported by literature on this or similar species, or were potentially amenable to management. ........................................................................................................................... 38  Table 3. Rates of site fidelity, active season and overwinter survival in newly released marmots. Estimated success, parameter estimates, unconditional standard errors, and 95% unconditional confidence intervals are provided for variables with >0.5 relative importance to predictive models of site fidelity, active season and overwinter survival. Models for each success metric were constructed using ‘all-subsets analysis’, and parameter estimates were attained by averaging models with AICc Δi <7 (see Statistical analyses, 2.2.5.3). ................ 39  Table 4. Discriminatory performance of averaged models when predicting site fidelity, active season and overwinter survival. Prevalence measures the number of occurrences of success in the data, and unbalanced prevalence is known to bias sensitivity (the proportion of correctly predicted successes), specificity (the proportion of correctly predicted failures), and vi  Cohen’s kappa coefficient (the proportion of correct predictions after accounting for chance). The True Skill Statistic (TSS) operates independently of prevalence (see Discussion, 2.4), and suggested that overwinter survival was the most predictable measure of success, but models for site fidelity and active season survival were still more predictive of success than chance alone. ........................................................................................................................... 40  vii  List of Figures Figure 1. Location of 30 release sites (2003-2010) for captive-bred Vancouver Island marmots, British Columbia, Canada. ...................................................................................... 41  Figure 2. Relative importance of 13 predictors of site fidelity for newly released, captivebred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date and resident female received >0.5 relative support as predictors and were included in 100% and 75% of models in the AICc Δi <7 subset, respectively. ........................................ 42  Figure 3. Predicted probability of site fidelity in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would hibernate within 1km of the release site, and were calculated using model-averaged parameter estimates and varying the values of release date and resident female presence while maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Site fidelity is predicted to increase with release date, but was lower overall for marmots released to sites with resident females within 500m of the release burrow. ........................... 43  Figure 4. Relative importance of 10 predictors of active season survival for newly released, captive-bred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date, age class, sex and talus cover received > 0.5 relative support as predictors and viii  were included in 100%, 86%, 74% and 73% of models in the AICc Δi <7 subset, respectively. ............................................................................................................................ 44  Figure 5. Predicted probabilities of active season survival in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would survive to hibernation, and were calculated using model-averaged parameter estimates and varying (a) age class, (b) sex, and (c) talus cover, while maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Active season survival is predicted to increase with release date and was higher for females, 2-year-olds, and marmots released to sites with talus cover within 1km of the release burrow. ..................................... 45  Figure 6. Estimated annual overwinter survival and 95% confidence intervals for wild-born marmots (never in captivity), established marmots (captive-bred and with >1 prior wild hibernation), and newly released captive-bred marmots (2006-2010). Captive-bred marmots that were ‘newly released’ in one year were then considered to be ‘established’ in subsequent years. Estimates of survival were much higher for wild-born and established marmots than for captive-bred marmots undergoing their first wild hibernation, particularly in 2009 and 2010......................................................................................................................................... 46  Figure 7. Relative importance of 10 predictors of overwinter survival for newly released, captive-bred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date was the only variable with >0.5 relative support, and was included in 88% of ix  models in the AICc Δi <7 subset. 'Pre-rel. hibernation days' refers to the number of days a marmot spent in hibernation prior to release, used as a proxy for quality of past hibernation experience. .............................................................................................................................. 47  Figure 8. Predicted probabilities of site fidelity, active season and overwinter survival in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would achieve each success metric, and were calculated using model-averaged parameter estimates and maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Site fidelity and active season survival are predicted to increase with release date, but overwinter survival was predicted to be highest for marmots released early in the active season. ..................................................................... 48  Figure 9. Predicted probabilities of active season and overwinter survival, and weighted average survival to the spring following release. Average survival was calculated by weighting the probabilities of active season and overwinter survival on each potential release date by the relative fraction of time required to achieve those individual success metrics (73 and 243 days, respectively, assuming a non-leap year). ......................................................... 49  x  Acknowledgements I am profoundly grateful to my supervisor, Dr. Peter Arcese, whose knowledge of ecological systems and enthusiasm for conservation biology were paramount in the completion of this thesis. I am also indebted to my supervisory committee, Drs. Kathy Martin and Brian Klinkenberg, for their constructive feedback. I offer my sincere appreciation to Don Doyle and Viki Jackson of the Marmot Recovery Foundation for their endless support and facilitation of my research. I would also like to thank the Vancouver Island Marmot Recovery Team for their ideas and encouragement at all stages of this work, and the many individuals that helped collect the eight years of field data used in this study. To the field crew that helped with my vegetation plots, I apologize for the thistles, cliffs, wasp stings, and my role in your newfound antipathy towards all things green and growing. My colleagues at the Arcese and Martin labs were invaluable as resources, soundingboards and friends, and I greatly appreciate the feedback they provided and the generosity of spirit with which they provided it. Special thanks go to Richard Schuster for his patience in helping me with R, discussing multi-model inference, and for sharing his knowledge of all things computer-related, and to Mark Drever for the benefit of his statistical knowledge as well as the use of his library. I offer my enduring gratitude to my parents, my sister and my friends for their support, encouragement and willingness to discuss marmots at any hour of the day. This research would not have been possible without the financial support of the University of British Columbia, NSERC, and the Mohammed bin Zayed Species Conservation Fund.  xi  Chapter 1: Introduction 1.1  Context  The use of captive-breeding and reintroduction programs in endangered species recovery is one of the most debated approaches to biodiversity conservation. Several criticisms of reintroduction programs include poor success, high costs that can pre-empt other recovery techniques, and poor monitoring of reintroduction success (Snyder et al., 1996). However, contrary examples also exist and some captive-breeding and reintroduction programs in Canada have shown promise. For instance, the swift fox (Vulpes velox) was extirpated from Canada in the 1930’s (Committee on the Status of Endangered Wildlife in Canada, COSEWIC, 2009). From 1983 to 1997, 942 foxes were reintroduced to southern Alberta and Saskatchewan (Moehrenschlager and Moehrenschlager, 2001). Most foxes were captive-bred at the Cochrane Ecological Institute in Alberta, Canada, but approximately 10% of foxes were translocated from wild populations in Wyoming, Colorado, and South Dakota (sources in Waters, 2010). By 2009, a wild population had persisted for over a decade without any further reintroductions and the Canadian population was thus downlisted from ‘extirpated’ to ‘threatened’ (Committee on the Status of Endangered Wildlife in Canada, COSEWIC, 2009). Managers reintroducing captive-bred species must make a suite of decisions, from the selection of appropriate release sites to the composition of release groups and the procedures with which groups are released (Kleiman, 1989). When release programs are in their infancy, decisions must be made without statistical support, and are usually based on a combination of trial-and-error and the assumptions of ‘experts’, based on their observations of wild populations, congeneric species or other release programs. However, false assumptions can lead to a misallocation of resources or program failures in the absence of active monitoring  1  and analysis of preliminary data on the factors influencing release success. For this reason, it is important to challenge the hypotheses behind release decisions to ensure that captive-bred animals are released in the best possible conditions in order to maximize subsequent survival and reproductive success.  1.2  Research objectives  The recovery project for the Vancouver Island marmot (Marmota vancouverensis, Swarth, 1911) is perhaps one of Canada’s best known captive-breeding programs, with a total of 301 marmots having been released to the wild from 2003-2010. My research objectives were twofold: first, to evaluate the success of the reintroduction program based on short-term release goals of ‘site fidelity’ (hibernation within 1km of the release burrow), ‘active season survival’ (survival from release to hibernation) and ‘overwinter survival’ (survival through hibernation into the following spring). Second, I aimed to identify the traits of candidate animals, release sites, and procedures most likely to maximize first year site fidelity, active season survival and overwinter survival in captive-bred marmots released to the wild. My overall goal was to apply the results of my research to improve release success and accelerate the re-establishment of functioning and persistent metapopulations on Vancouver Island.  1.3  Study species  There are 14 extant species of marmots recognized across the Holarctic (Steppan et al., 1999), with six species found in North America and the remaining eight distributed across Eurasia. Three species are presently classified as ‘threatened’ (M. menzbieri, M. sibirica, and M. vancouverensis, International Union for the Conservation of Nature, IUCN, 2011).  2  Marmots contribute to environmental heterogeneity and ecosystem function through burrow excavation, herbivory, and as secondary prey for several species, and one species (M. sibirica, Zahler et al., 2004, in Murdoch et al., 2009) has been proposed as a keystone species. Marmots are also of ecological interest for a variety of reasons related to their life history, adaptations to living in harsh environments, and the variety of social structures and mating systems that have evolved in the group as a whole. Moreover, because habitat selection by marmots is strongly tied to climate conditions that influence survival (e.g. snowfall, spring melt, and tree line dynamics) they have the potential to act as sentinels of environmental change and, in some regions, have already responded to climate change by emerging early from hibernation (M. flaviventris, Inouye et al., 2000). The Vancouver Island marmot is a large, herbivorous ground squirrel and one of Canada’s five endemic species of land mammal (Nagorsen, 2004). Its status as a true species was initially debated because of mitochondrial similarities with M. caligata (Kruckenhauser et al., 1999; Steppan et al., 1999) but has since been confirmed by extensive comparisons of morphological and behavioural traits (Blumstein, 1999; Cardini, 2003; Cardini and O’Higgins, 2004; Cardini et al., 2009, 2005; Heard, 1977; Polly, 2003). The Vancouver Island marmot inhabits terrain complexes of meadows, talus slides, and cliffs (Heard, 1977), usually above 1050m elevation (Bryant and Janz, 1996) and within the Alpine Tundra and Mountain Hemlock biogeoclimatic zones (Bryant, 1998). Ideal marmot habitat supports the construction of complex burrows and hibernacula, features a variety of suitable forage plant species, and provides marmots with consistent microclimatic conditions that facilitate successful hibernation (Vancouver Island Marmot Recovery Team, 2008). Microclimate may be crucial to successful hibernation in Vancouver Island marmots,  3  which hibernate for an average of 210 days each year (Bryant and McAdie, 2003). Ideal habitat also includes boulders or rock piles which can be used as viewing platforms, refuges from predators, and for thermoregulation. Marmots display bimodal daytime activity patterns: foraging and resting in the sun during mornings and evenings, but avoiding hot midday temperatures by retreating underground (Heard, 1977). Vancouver Island marmots live in colonies of one or more family groups, each typically composed of an adult male, one or more adult females, and varying numbers of 2year-olds, yearlings and ‘pups’ (also called ‘young of the year’, ‘juveniles’, ‘infants’, e.g. Bryant and Janz, 1996). Sexual maturity is achieved relatively slowly (e.g. 3-4 years, Heard, 1977) and evidence of territoriality is minimal. Marmots benefit from coloniality via group vigilance and conspecific alarm calls (Blumstein et al., 2001) and possibly through the energetic benefits of joint hibernation (e.g. M. marmota, Arnold, 1990b). At landscape scales, Vancouver Island marmots exhibit a metapopulation structure (Bryant, 1998; Bryant and Janz, 1996). Although philopatric marmots usually remain within 1000m of their home burrow (Bryant and Page, 2005), some disperse up to 25km from their natal colony to join existing or found new colonies (J. MacDermott, pers. comm.). Most long-distances dispersers are males two or more years of age (Bryant, 2005). Dispersal is thought to maintain gene flow within metapopulations (Kruckenhauser et al., 2009) as colonies undergo local extinction and recolonization related to spatial and temporal variation in reproduction and predation (Bryant and Janz, 1996). Predators of Vancouver Island marmots include cougars (Puma concolor vancouverensis), wolves (Canis lupus crassodon), and golden eagles (Aquila chrysaetus canadensis; e.g. Bryant and Page, 2005), and less commonly, black bears (Ursus americanus vancouveri, pers. obs.). Historically, Vancouver  4  Island wolverines (Gulo gulo vancouverensis) are also likely to have been predators of marmots but are now thought to be extinct. Marmot bones in Vancouver Island caves and oral history accounts also indicate that humans preyed on marmots prior to European colonization (Nagorsen et al., 1996).  1.4  Population history  The Vancouver Island marmot has experienced dramatic fluctuations both in distribution and in population size. The population was thought to be historically small, limited foremost by habitat size, availability, and connectivity (Bryant and Janz, 1996). Cave-bound faunal remains, however, suggest larger colony sizes and much wider prehistoric distribution than was noted in recent years (Nagorsen et al., 1996). Between 1864 and 1969, marmots were reported at natural colonies on at least 25 mountains (Bryant and Janz, 1996). In the following three decades only 15 mountains supported active colonies, and the northern metapopulation was comprised of a single colony (Bryant and Janz, 1996). The Vancouver Island marmot was first protected by the BC Wildlife Act in 1973 and listed as ‘endangered’ in 1980 (Munro et al., 1985), and has since been confirmed as ‘endangered’ by COSEWIC (Munro, 1979), the International Union for the Conservation of Nature (Thornback and Jenkins, 1982), and the United States Endangered Species Act (1984; Vancouver Island Marmot Recovery Team, 2008).  1.5  Recent literature  Relatively little was known about the Vancouver Island marmot prior to 1970. Heard (1977) carried out the first comprehensive study on the Vancouver Island marmot, documenting its  5  habitats, activity patterns, vocalizations and social behaviours. Since then, researchers have conducted more detailed studies on habitat use and forage plants (Martell and Milko, 1986; Milko, 1984), community-level change in subalpine vegetation (Hebda et al., 2004; Laroque, 1998), vocalizations (Blumstein, 1999), and the reproductive biology and behaviour of wild (Bryant, 2006, 2005, 1996) and captive marmots (Casimir et al., 2007; Keeley et al., 2011). Research has also attempted to identify potential threats related to population persistence, including disease (Bryant et al., 2002), parasites (Mace and Shepard, 1980), genetic homogeneity (Bryant, 1990; Kruckenhauser et al., 2009), atmospheric contaminants (Lichota et al., 2004) and the effects of non-consumptive alpine recreation (Dearden and Hall, 1983). Most recently, attention has focused on the complex relationships between the Vancouver Island marmot and its main predators, the wolf, cougar, and golden eagle. Blumstein et al. (2006) confirmed that Vancouver Island marmots can identify terrestrial predators, and Blumstein et al. (2001) observed vigilant patterns similar to six congeners when foraging, including comparable flush distances, alarm call frequencies and the creation and use of numerous escape burrows. Nevertheless, predation was identified as a proximate cause of population decline from 1992-2004 (Bryant and Page, 2005), and was the most common cause of mortality for both wild-born and captive-bred marmots from 1987-2007 and 2003-2007, respectively (Aaltonen et al., 2009). Captive-born marmots incurred greater mortality than wild-born marmots throughout the year, particularly during the ‘active season’ (early May-early October) when marmots were not hibernating, and thus spent large amounts of time above ground (Aaltonen et al., 2009). In some areas, predation has been associated with the establishment of marmot colonies in young, high elevation cutblocks (Bryant et al., 2002). The first colonization of a  6  cutblock was recorded in 1981, and 10 other cutblocks were subsequently colonized, most within 1km of natural habitat (Bryant and Janz, 1996). Bryant (1996) reported that colony persistence was reduced in cutblocks as compared to nearby subalpine meadow and talus habitats. Bryant and Page (2005) suggest several mechanisms for the observed ‘source-sink dynamics’ in cutblocks and their effect on adjacent natural colonies, but there is still uncertainty linked to the apparent extirpation of colonies in protected areas like Strathcona Provincial Park, where cutblocks were confined to the outer edges of park boundaries (Janz et al., 2000).  1.6  Captive-breeding and recovery program  Between 1997 and 2004, 54 wild marmots were taken into captivity for the purpose of developing breeding populations (McAdie, 2004). Marmots were bred at four conservation facilities: the Toronto and Calgary zoos, the Mountainview Conservation and Breeding Centre, and the Tony Barrett Mt. Washington Marmot Recovery Centre on Vancouver Island. Mount Washington is a functioning ski hill with a wild marmot colony established in close proximity to the facility. Release candidates not born at the Mt. Washington facility were brought there at least eight months prior to their release for quarantine purposes, and to give them the opportunity to become familiar with natural forage species and experience a typical, high-elevation hibernation in their release groups for the following summer. Further details of the captive-breeding program and its history can be found in McAdie (2004). The goal of the Vancouver Island Marmot Recovery Team is to establish three self-sustaining, geographically distinct metapopulations totaling 400-600 wild marmots (Vancouver Island Marmot Recovery Team, 2008).  7  1.7  Releases and monitoring  Marmots were released to plywood nest boxes containing food and water. Nest boxes were built with exit holes at opposite ends and installed at burrow entrances or natural talus piles to extend the marmots’ refuge space and to provide a viewing platform for predator detection. Field crew provided supplemental food (Buckerfields® Specialty Feed rabbit pellets and Mazuri® leaf-eater biscuits with peanut butter) at the release site for two or more days post-release. On days when feeding occurred, field crews also ‘shepherded’ marmots in an attempt to reduce predation. Shepherding involved sitting within ~750 m of the release site during daylight hours, prepared to launch bear bangers at predators if detected. Marmots habituated rapidly to the presence of field crew (see also Heard, 1977). All Vancouver Island marmots release candidates had transmitters surgically implanted into the peritoneal cavity at least two weeks prior to release (model A1-2TH, Holohil Systems Ltd., Carp, Ontario; model IMP-300, Telonics Inc., Mesa, Arizona). Transmitter pulse strength provided directionality enabling field crew to track transmitters to an exact location. Pulse speed indicated the temperature of a transmitter, a proxy for the marmot’s survival status because living marmots maintain transmitter temperatures over 30˚C when not hibernating (D. Doyle, pers. comm.). Mortalities were confirmed by field crews by recovering transmitters or tracking them to a single burrow, cliff, tree, or other inaccessible and invariant location. All marmots were located by ground-based or aerial surveys at least twice after their release and prior to hibernation, but most marmots were monitored more often (mean = 4.6 relocations per animal in 2009-2010). Location efforts generally increased in late summer  8  and early fall, when multiple flights were taken to record hibernacula locations and search for marmots that had abandoned their release sites or could not be detected in ground surveys.  1.8  Chapter descriptions  My objectives were to evaluate and identify ways to improve upon the short-term success of the Vancouver Island marmot reintroduction program. The most immediate measures of release success for captive-bred marmots during their first year in the wild are release site fidelity, active season survival, and overwinter survival. In Chapter 2, I evaluate success of captive-bred marmots at achieving these metrics. I identify the most important predictive variables from a candidate set including marmot characteristics, release practices and procedures, and habitat attributes, and I discuss these results within the context of management decisions for future releases. Finally, in Chapter 3, I summarize my results, describe additional predictors described in literature that were not evaluated in my thesis, and discuss the implications of my results for future reintroductions of captive-bred Vancouver Island marmots.  9  Chapter 2: Determinants of reintroduction success in captive-bred Vancouver Island marmots 2.1  Introduction  Captive-breeding and reintroduction programs are widely used in endangered species recovery (Kleiman, 1989; Seddon, 1999; Seddon et al., 2007) and undertaken with a goal of creating viable wild populations (Griffith et al., 1989; IUCN, 1998). However, relatively few reintroduction projects have succeeded by that definition (Beck et al., 1994; Fischer and Lindenmayer, 2000; Griffith et al., 1989; Wolf et al., 1996). Seddon (1999) recommended the use of incremental measures of success to evaluate progress following the release of captive-bred animals. Because many captive-bred species show ‘release effects’ such as high mortality during their first year in the wild (e.g. Dama dama mesopotamica, Bar-David et al., 2005; Gyps fulvus fulvus, Sarrazin et al., 1994; Ovis canadensis, Ostermann et al., 2001), evaluating reintroduction success earlier should facilitate adaptive management approaches in order to identify best practices and accelerate species recovery. The Vancouver Island marmot (Marmota vancouverensis, Swarth, 1911) is a large, hibernating ground squirrel endemic to the Insular Mountains of Vancouver Island, in British Columbia, Canada (Heard, 1977). The species was listed as endangered by COSEWIC in 1978 (Munro, 1979). By 1997, marmot abundance in the wild was estimated at ~150 individuals, of which at least 25 were pups (Bryant et al., 2002). Although the ultimate causes of decline remain uncertain, recent declines were attributed to predation by cougars (Puma concolor vancouverensis), wolves (Canis lupus crassodon) and golden eagles (Aquila chrysaetus canadensis) at historic and cutblock colonies (Bryant and Page, 2005). To protect  10  the species from extinction and provide marmots to be released to the wild, a captivebreeding program was established in 1997 (McAdie, 2004). Aaltonen et al. (2009) used radiotelemetry and mark-recapture methods to estimate survival in captive-bred and wild marmots, and confirmed predation as the most frequent cause of mortality in both groups. Annual survival (S) of captive-bred marmots released between 2003 and 2007 was 24% less (S = 0.61, 95% CI = 0.51 – 0.70) than annual survival of wild-born marmots (S = 0.85, 95% CI = 0.76 – 0.92), largely due to poor survival during the active season when marmots were vulnerable to predation by terrestrial and avian predators. Since 2007, an additional 212 captive-bred marmots have been released to the wild. I re-evaluated reintroduction success (2003-2010) using three short-term measures of success for a marmot’s first year in the wild: ‘site fidelity’ (fidelity to release site), ‘active season survival’ (survival to hibernation), and ‘overwinter survival’ (survival through hibernation to late spring). I use the term ‘reintroduction’ to describe any release of a captivebred marmot to the wild, although some releases were technically ‘supplementations’ or ‘conservation introductions’ (IUCN, 1998). Specifically, I aimed to predict post-release site fidelity and survival using variables linked to marmot qualities of sex and age, experience in captivity, management practices, and the local and landscape-level attributes of release habitats. A number of candidate variables were previously demonstrated to affect fidelity or survival in marmots and all are potentially amenable to management, including ‘marmot characteristics’ of sex, age class, and birth facility, and ‘release practices’ related to group size, within-group relatedness (via a pedigree), inclusion of pups within release groups, and the presence of resident established or wild-born marmots at release sites. To evaluate release habitats, I analyzed aerial imagery to  11  describe the extent of continuous or harvested forest, and used meadow and talus cover as proxies for forage value and safety terrain. Release date was included to indicate the length of active season available to marmots between their release and first wild hibernation, and distance between release and hibernation locations measured to test for links between postrelease dispersal and survival. There was great variation in hibernation length achieved by marmots in the winter prior to their release (89-175 days). To predict overwinter survival, therefore, I included pre-release hibernation days, and, as a measure of environmental harshness, elevation at hibernation. I used two grouping factors (year and mountain) and aspect at the hibernation site as random effects. Literature on this and closely related species suggests that sex (Aaltonen et al., 2009; Armitage, 1999; Arnold, 1990b; Bryant, 1998), age (Bryant, 1998, 1996), and social cohesiveness (Armitage et al., 2011; Blumstein et al., 2009; Downhower and Armitage, 1981) would all be strong predictors of release site fidelity, that active season survival would be highest for site-faithful marmots (Arnold, 1990b; Van Vuren and Armitage, 1994), and that age (Blumstein and Arnold, 1998; Bryant, 1996; Lenihan and Van Vuren, 1996) would predict overwinter survival.  2.2 2.2.1  Methods Study species  Vancouver Island marmots live in colonies comprised of one or more family groups, made up of an adult male, one or more adult females, and related subadults, yearlings, and pups (Bryant, 1990; Heard, 1977). Females can breed as yearlings (M. McAdie, pers. comm.) but most produce first litters at > 3 yrs of age (mean = 3.6 yrs, SD = 1.2, Bryant, 2005). Litters  12  are small (mean = 3.4, SD = 1.1, n = 58) and produced annually or biennially (Bryant, 2005). Vancouver Island marmots are typically active above ground April - October, but hibernate socially for an average of 210 d/yr (Bryant and McAdie, 2003). Hibernation requires the accumulation of fat reserves during the active season because marmots do not store food (Armitage et al., 2003; Legaarden et al., 2001; Ortmann and Heldmaier, 2000). Burrows are used as hibernacula, for sleeping and birthing, and as refugia from thermal stress (Heard, 1977) and predators (Blumstein et al., 2001). Vancouver Island marmots are usually found at elevations of 1000-1400m (Bryant and Page, 2005) in steep subalpine meadows and talus slides. These habitats provide forage early in spring and are maintained by winter avalanches and snow creep which limit tree encroachment and enhance visibility (Heard, 1977). Marmot habitat is patchily distributed but dispersal, most frequently in males at or after age two, is common and likely contributes positively to gene flow and the viability of sub-populations (Bryant, 2005, 1998; Bryant and Janz, 1996).  2.2.2  Reintroduction program  Between 1997 and 2004, 54 wild marmots were taken into captivity (McAdie, 2004). Marmots were bred at four conservation facilities: the Toronto and Calgary zoos, the Mountainview Conservation and Breeding Centre, and the Tony Barrett Mount Washington Marmot Recovery Centre on Vancouver Island, British Columbia, Canada. Mount Washington is a functioning ski hill with a wild marmot colony established in close proximity to the facility. Release candidates not born at the Mt. Washington facility were brought there at least eight months prior to their release, giving them the opportunity to  13  become familiar with natural forage species and to experience winter hibernation at highelevation before their release the following summer. Further details of the captive-breeding program and its history can be found in McAdie (2004). In total, 301 captive-bred marmots were reintroduced to the wild between 2003 and 2010 (Figure 1; Appendix A). All marmots used in logistic regressions were captive-bred, aged > 1yo and were released between late June and early September using standard procedures described in section 2.2.3. My analyses excluded 21 marmots released using nonstandard procedures and 23 pups released between 2008 and 2010. I analyzed pups separately because I expected them to exhibit different site fidelity and survival (e.g. Aaltonen et al., 2009), and because there were too few pups with data (n = 19, 19, and 16 for site fidelity, active season survival and overwinter survival, respectively) to allow inclusion as a separate age class in my main analyses. An additional 43 marmots were excluded from analysis because they were ‘missing’ by the September after their release. Thus, three sample groups of 214, 199, and 144 marmots contributed to my primary analyses of release site fidelity, active season survival, and overwinter survival, respectively (Table 1). Overwinter survival had the smallest sample size because it also excluded 31 marmots that died during the active season and 24 marmots that had been located in the fall prior to hibernation but were missing in the following spring.  2.2.3  Release procedures  Release groups were formed to maximize genetic benefits in anticipation of future reproduction by group individuals. Candidate marmots were introduced to their release groups at various times depending on age, relatedness, and birth facility, but all release  14  groups were formed either before or during the pre-release hibernation period at the Mt. Washington facility. At least two weeks prior to release, all candidates were surgically implanted with temperature-calibrated radio telemetry transmitters (model A1-2TH, Holohil Systems Ltd., Carp, Ontario; model IMP-300, Telonics Inc., Mesa, Arizona; see Bryant and Page, 2005, for methods). Activated transmitters were expected to last for up to 36 months. Candidates were examined immediately prior to release to confirm their full recovery from transmitter implantation. Release groups varied in size and composition, particularly in sex and age ratios and in relatedness of each marmot to the other members of its group. Marmots were transported to release sites by some combination of helicopter, 4x4 vehicles, and occasionally on foot by field crew. Plywood nest boxes lined with straw were installed as part of a naturally-occurring or marmot-excavated refuge. Marmots were released to nest boxes and the outer exit blocked for 30-120 minutes to discourage immediate site abandonment, although marmots sometimes created new exits before established exits were re-opened. Boxes were provisioned with supplemental food (Buckerfields® Specialty Feed rabbit pellets and Mazuri® leaf-eater biscuits) and water for at least two days following release. During this time, field crews remained in the vicinity to observe new releases and protect them from predators.  2.2.4  Definitions of success  I defined site fidelity as the presence of a captive-bred marmot within 1km of its release burrow on or after October 1st of its release year. If there were multiple burrows onsite and within 50m of one another at the time of release, I treated their midpoint as the release location. Marmots that could not be located after September but had previously been detected  15  >2km away from the release burrow earlier in the season were assumed not to have returned. In 43 cases, marmots went ‘missing’ and were not detected on- or off-site in subsequent searches of the area (Table 1). I excluded those marmots from my analysis of fidelity because several documented transmitter malfunctions (e.g. premature failure) made me reluctant to assume that non-detection was a true sign of site abandonment. Five marmots that abandoned their release sites during the active season were trapped and either returned to their release colony or relocated to another site. Those animals were included in my analysis of site fidelity, but not for active season or overwinter survival. I defined active season survival as the survival of a marmot from release to hibernation. Because the start of hibernation is frequently difficult to detect, I assumed that a marmot survived to hibernation if it was detected alive in October and if its transmitter was not found above ground the following spring, as would be expected in the event of late-season predation. Overwinter survival was achieved when an animal that entered hibernation was observed to successfully emerge and survive to the end of May in the following spring.  2.2.5 2.2.5.1  Data collection Re-sighting  We used ground- or air-based radiotelemetry to track marmots and confirm survival at least three times between release and hibernation, and we attempted to recover transmitters and determine cause of death in all cases of mortality (see Bryant and Page, 2005). Similar methods were used in October and November to locate all marmot hibernacula. Although many marmots were still active above ground at that time, I treated the last known location as the hibernacula, and confirmed locations the following spring. All coordinates were recorded  16  using various global positioning systems (GPS) devices to an expected accuracy of ±12m. I used ArcGIS (version 9.2, Environmental Systems Research Institute, California, USA) to estimate aspect and elevation from a 25m resolution Digital Elevation Model (Department of Geography, University of British Columbia) created from mass points and break lines collected by the Province of British Columbia (release 1.0, Geographic Data BC, 1997).  2.2.5.2  Habitat characteristics  To characterize habitat type near each release site, I used ArcGIS to examine fused SPOT-5 (5m resolution) and Landsat 5 (30m resolution) panchromatic images (GeoBC, Base Mapping and Cadastre Branch, 2007, 2006, 2004; Province of British Columbia, 2011). Because philopatric marmots spend most of their time within 100-1000m of a home burrow (Bryant and Page, 2005), I digitized polygons as ‘forest’, ‘meadow’, and ‘talus’ within a 1km radius around each release burrow. Habitat types were easily distinguished by colour and texture, and reviewed by field crew with >15 years of experience hiking to and flying over release sites. Using 1:5,000 scale images, I digitized all habitat polygons with a patch size >25m² in area and comprised of >75% cover of a single habitat type. Habitat types were then summed as the percentage of each within a 1km radius, and standardized as z-scores to ensure parameter estimates of similar magnitude. To evaluate the landscape-level context of habitat attributes, I estimated the extent of ‘cutblocks’ (age 0-15 yrs) within a 3km radius of marmot release sites (cf., Bryant, 1998) using maps provided as digital shapefiles by landowners (© Island Timberlands Limited Partnership, 2011; ©Western Forest Products Incorporated, 2011), and then calculated the amount of recently logged habitat in the vicinity of each burrow in each year. TimberWest  17  Forest Corporation (©2011) provided these calculations in spreadsheet form. Between July and early September of 2010-2011, I also sampled fine-scale forage and visibility characteristics within 60m of release burrows at a subset of 39 release burrows on 21 mountains.  2.2.5.3  Statistical analyses  I used an information-theoretic approach to evaluate alternative models based on KullbackLeibler relative distance, an estimate of the information lost when using one probability distribution to approximate another (Burnham and Anderson, 2002). Because n/K was <40 in each of my datasets (214/19, 199/17, and 144/15 for site fidelity, active season survival, and overwinter survival, respectively), I used the second-order form of Akaike’s Information Criterion (AICc, Hurvich and Tsai, 1989) to estimate the K-L relative distance. AICc uses an additional bias-correction term to accommodate small sample sizes relative to model parameters estimated. All analyses were performed in R (version 2.13.2, R Development Core Team, 2005), except for model-averaging, which was carried out manually using Microsoft Excel (2007). Response variables were binary in each case, and each set of models included both fixed and random effects, comprising generalized linear mixed-effects models with a logistic transformation of the predictors (glmer function, lme4 package, Bates et al., 2011). Models were fitted using a Laplacian approximation to the log-likelihood, which accommodated both the number of random effects in models and the volume of models fitted. Prior analyses of captive-bred Vancouver Island marmots tested for sex, age, and release effects on active season survival (Aaltonen et al., 2009). My analyses also included some variables potentially  18  in the control of managers that had not been previously evaluated for their influence on reintroduction success. In order to explore all relationships of interest, therefore, I used an ‘all-subsets analysis’ (Montgomery and Peck, 1992; Neter et al., 1996) to model every possible combination of literature- or management-supported covariates. I calculated the ‘events per variable ratio’ (EPV) for each dataset as the smaller of the number of outcome events per parameter estimated for fixed effects (Vittinghoff and McCulloch, 2007). A small number of site fidelity (4 of 5,121) and active season survival models (24 of 1,025) were excluded from analysis because their estimated variance-covariance matrices became singular (non-invertible), and so a unique solution could not be attained. I used differences in AICc (Δi) to calculate Akaike weights (wi), estimates for the likelihood of a model being ‘best’ given the other candidates. As expected, no model emerged with ‘good’ support (wi >0.9, Burnham and Anderson, 2002), so I selected a subset of models with strong to moderate support (differing by the AICc ‘best’ model by <7, Burnham and Anderson, 2002) for further exploration. To evaluate relative support for covariates, I summed the subset-adjusted Akaike weights for each model in which they were included. ‘Relative importance’ for predictors therefore ranged from 0 (not included in any models) to 1 (included in all models). I next applied weighted multi-model inference to estimate parameters and predict random effects. This also allowed me to avoid the bias inherent in selecting a single model to use in predictions. Finally, I used the revised equation in Burnham and Anderson (2004) applied with the R MuMIn package (Bartoń, 2012) to calculate the unconditional standard errors for parameter estimates and to construct 95% confidence intervals around fixed effects with greater than 0.5 relative support.  19  I explored changes in the predicted probability of success for each response variable by varying the values of the most influential fixed effects (>0.5 relative importance) by quartiles, while maintaining covariates at their medians or modes (Shaffer and Thompson, 2007, in Peak, 2007). To evaluate the predictive ability of averaged models, I dichotomized continuous success predictions at a threshold of 0.5 (<0.5 = failure, >0.5 = success) and used R (caret package, Kuhn, 2012) to calculate ‘sensitivity’ (the proportion of correct predictions of success, Wilson et al., 2005), ‘specificity’ (the proportion of correct predictions of failure, Wilson et al. 2005), and Cohen’s kappa coefficient (the proportion of correct predictions after accounting for chance, Cohen, 1960). I also used sensitivity and specificity to calculate the ‘true skill statistic’ (TSS = sensitivity + specificity – 1; Allouche et al., 2006), a measure of accuracy independent of prevalence that evaluates the improvement of a model on the accuracy that could be expected by chance prediction alone (0 = chance, 1 = perfect accuracy; Allouche et al., 2006). TSS has also been referred to as ‘Youden’s Index’ in medical literature (Youden, 1950), and the ‘Peirce Skill Score’ (Peirce, 1988) or ‘HanssenKuipers discriminant’ in weather forecasting (Hanssen and Kuipers, 1965).  2.3 2.3.1  Results Site fidelity  Of 214 marmots tracked to hibernation, 67.8% (SE = 3.2, n = 145) exhibited site fidelity by hibernating within 1km of their release burrow (Table 3). Of 69 dispersers, 71.0% (SE = 5.5, n = 49) traveled straight-line distances of 1-5km and 29.0% (SE = 5.5, n = 20) dispersed further. The longest straight-line distance recorded in a release year was just over 20km, by a 2 yr-old male in 2006. Four pups were ‘missing’ at hibernation, but this was suspected to be  20  a function of transmitter issues and not site abandonment or mortality. Of the 19 located pups, 100% exhibited site fidelity. My all-subsets analysis included 5,121 models to predict site fidelity and resulted in 392 models with ΔAICc<7 used in a subset to estimate model-averaged parameters. Although the best model in the subset had an Akaike weight of only 0.025, the frequency of variable selection suggested marked variation in influence overall. In particular, release date emerged as a variable with strong relative importance and model influence through its inclusion in 100% of subset models (Figure 2). Marmots released in late June were 49% less likely to remain onsite than marmots released in mid-September (Figure 3). I also found strong support for a negative influence of resident female presence on site fidelity, which predicted a 29% higher likelihood of site fidelity at sites where there were no resident females. Resident male presence near release sites was less important to fidelity, as were group composition variables of age class, sex, group size, relatedness, or release of adults with their own pups (Figure 2). Habitat variables and landscape-level disturbance metrics were also weak predictors of site fidelity, and I found no correlation between site-specific forage species and visibility characteristics and site fidelity, active season survival, or overwinter survival. A sex by age interaction was not selected in any site fidelity model, but all other predictors were selected at least once. Model-averaged parameter estimates for variables with >0.5 relative importance in the 392 model ΔAICc<7 subset are summarized in Table 3. Appendix B lists the complete set of model-averaged parameter estimates for all predictors.  21  2.3.2  Active season survival  Marmots experienced high active season survival overall (84.4%, SE = 2.6, n = 199; Table 3). Predation was the implied cause of mortality in 61% of 31 deaths recorded, including seven by eagles, five by cougars, one each by a wolf and black bear, and five deaths attributed to predation that were not assigned to species. One marmot died by falling and one via a suspected bacterial infection (M. McAdie, pers. comm.). Ten marmots confirmed dead by telemetry were not recovered. Ninety-five percent of pups (SE = 5.3, 18 of 19) achieved active season survival, but the deceased pup was not recovered and so the cause of mortality is unknown. Of 1,025 models used to predict active season survival, 106 had Δ AICc<7, with the best model having a weight of 0.077. Four variables received greater than 0.5 relative importance and all were selected at least once (Figure 4). The averaged model was most sensitive to age class and release date (Table 3, Appendix B). Two-year olds were predicted to survive 5% better than yearlings and 21% better than adults (Figure 5a). Marmots released at the end of June were 31% less likely to survive to hibernation than those released in midSeptember. In addition, females were predicted to be 9% more likely to survive than males (Figure 5b), and marmots released to sites with the maximum local representation of talus habitat were 9% more likely to survive the active season than those released at sites with the least talus (Figure 5c). Birth facility, release group size, and the amount of disturbance around each release burrow were all poorly supported across models as being influential of active season survival.  22  2.3.3  Overwinter survival  Fifty-eight percent of 144 captive-bred marmots used in my primary analyses survived overwinter (SE = 4.1, n = 84, Table 3). Between 2006 and 2010, survival of new releases ranged from 29-92%, but was <40% in both 2009 (14 of 41) and 2010 (6 of 21). In contrast, survival of wild-born and established marmots averaged 99% of 122 and 98% of 118 marmots, respectively (Figure 6). Fifty percent of 16 captive-bred pups survived overwinter (SE = 12.9). I used 129 models to predict overwinter survival and included 51 in the Δ AICc <7 subset, with the best model weighted at 0.133. Release date was the only variable with greater than 0.5 relative support (Figure 7), and was negatively associated with overwinter survival (Figure 8; averaged parameter estimates in Table 3, Appendix B). The model subset included the null model, which differed from the best model by 3.6 AICc. Distance class traveled, pre-release hibernation length at Mt. Washington, age class, sex, elevation at hibernation, and birth facility were all suggested to be less important to overwinter survival than release date. Weighted average survival of marmots from release through the active season and overwinter periods was predicted to be highest for marmots released in late June or early July (Figure 9).  2.3.4  Model assessment  I aimed for a global EPV of at least five (Vittinghoff and McCulloch, 2006), and achieved it for site fidelity (EPV = 5.4), and overwinter survival (EPV = 8.6), but not for active season survival (EPV = 2.9). The averaged models for release site fidelity and active season survival showed high degrees of sensitivity, but poor specificity (Table 4). The site fidelity model had  23  a kappa coefficient of 0.44 and a TSS of 0.40, and the model for active season survival had a TSS of 0.28, and a kappa coefficient of 0.39. The overwinter survival model had the best discrimination ability, with a sensitivity of 0.90, a specificity of 0.85, and a kappa coefficient and TSS of 0.75. Models with perfect predictive ability would have received values of 1 for each of those model performance tools, whereas models with no predictive ability (equivalent to chance) would have had a sensitivity and specificity of 0.5, and a kappa coefficient and TSS of 0.  2.4  Discussion  Captive breeding is now employed worldwide to re-introduce and improve the likelihood of persistence of rare species in wild settings, including the Vancouver Island marmot (Janz et al., 2000). Long-term surveys suggest that reintroduction has played a critical role in the recovery effort for Vancouver Island marmots to date, given that an estimated 150 wild marmots in 1997 (Bryant et al., 2002) had increased to roughly 300 marmots by late 2010 (D. Doyle, pers. comm). This recovery program has also achieved two of three criteria proposed by Seddon (1999) as interim measures of success: survival of released animals, and successful breeding by released animals and their offspring. By 2010, 15 of 30 release mountains had produced pups, and some mountains had produced multiple litters in multiple years. I set out to evaluate the short-term success of marmot re-introductions with the goal of identifying ways to improve site fidelity, active season survival, and overwinter survival of newly released captive-bred marmots in future. Researchers performing all-subsets analysis are encouraged to build focused models linked to tractable hypotheses (see Burnham and Anderson, 2002; Dochtermann and Jenkins,  24  2011; Symonds and Moussalli, 2011) because including large numbers of predictors increases the likelihood of spurious correlation and model uncertainty (Burnham and Anderson, 2002; Dochtermann and Jenkins, 2011). Similar risks can arise when using second-order Akaike Information Criterion (AICc) to select the ‘best model’ because AICc evaluates models by comparison; thus, even the best-ranked models may perform poorly if there are no strong models in the set (Guthery et al., 2005). I focused on variables under management control and previously identified in the literature as potentially affecting site fidelity, and active season and overwinter survival, and then averaged parameter estimates to evaluate effect sizes and minimize model selection uncertainty (‘zero method’, see Burnham and Anderson, 2002; Nakagawa and Freckleton, 2010). Had I reduced the list of explanatory variables before analysis, I would have risked removing variables of relative importance and thereby under-fitting explanatory models, and also failed to examine untested hypotheses underlying current management practices. While a less model-intensive approach might be preferred (Burnham and Anderson, 2004, 2002), my approach allowed me to compare a complex set of existing hypotheses being actively discussed by managers with goal of improving the predictions of future management experiments (Burnham and Anderson, 2002; Symonds and Moussalli, 2011). For these reasons, and because I felt confident that my variables were supported by published results and management experience, it was an appropriate and useful application of the method. Evaluation of the averaged models revealed that overwinter survival was more predictable than site fidelity or active season survival, but all three models were more effective predictors than chance alone (Table 4). The prevalence of success and failure in the dataset describing overwinter survival was fairly balanced, resulting in good model  25  sensitivity and specificity (Table 4). Moreover, kappa and TSS values of 0.76 and 0.75, respectively, suggested ‘substantial agreement’ of predicted and observed outcomes according to one set of interpretation guidelines (Landis and Koch, 1977). In contrast, both site fidelity and active season survival had higher prevalence of successes than of failures (68% and 84%, respectively), making it less appropriate to assess sensitivity, specificity and kappa at a 0.5 threshold (Allouche et al., 2006; McPherson et al., 2004). TSS, which is similar to kappa and calculated independently of prevalence (Allouche et al., 2006) suggested only ‘fair’ support for site fidelity (TSS = 0.40) and active season survival (TSS = 0.28). Ideally, I would have evaluated model performance using independent data (Fielding and Bell, 1997), but all data were used in model development to accommodate the number of literature- or management-supported covariates. Future releases of captive-bred marmots will enable collection of independent data with which to better validate the predictive abilities of my models for site fidelity, and active season and overwinter survival.  2.4.1  Site fidelity  Site fidelity of captive-bred marmots was most influenced by release date. The length of the active season for marmots in the wild is made finite by hibernation, and predictably, marmots released later in the active season were much more likely to hibernate within 1km of their release site. Natal dispersal by wild-born marmots is also uncommon late in the active season, peaking before August for M. marmota (Arnold, 1990a) and female M. flaviventris (Armitage et al., 2011; Downhower and Armitage, 1981), and being inhibited entirely in juvenile S. beldingi (Nunes et al., 1998). Early dispersal could be advantageous to marmots because of the fitness consequences for animals unable to find a mate or suitable  26  hibernaculum prior to hibernation. Increases in site fidelity late in the active season could also be explained in part by circannual rhythms of the genus Marmota, wherein wild individuals typically spend less time above ground in late summer or early fall as immergence approaches (M. caligata, Taulman, 1990; M. flaviventris, Armitage et al., 1996; M. marmota, Perrin et al., 1993; M. vancouverensis, Heard, 1977). Therefore, my observation of greater site fidelity amongst animals released later in the active season is both biologically and statistically plausible. My results also suggest a decline in site fidelity of 29% among captive-bred marmots released as groups within 500m of established or wild-born resident females (Figure 3, Table 3). By contrast, although managers structured release groups to maximize the likelihood of social cohesion and breeding, I found little support for an influence of group structure on site fidelity (Figure 3). This suggests that conspecific interactions with marmots other than those included in the release group may be more influential of site fidelity than those within, a relationship also documented in reintroduced Lutra lutra (Sjoasen, 1997) and suspected in translocated Martes americana (Davis, 1983). Where possible, managers may want to consider maintaining a greater spatial buffer between new release groups and resident females. Based on several studies suggesting the importance of habitat safety characteristics to survival in M. flaviventris (Blumstein et al., 2006; Carey, 1985; Carey and Moore, 1986), M. caligata (Holmes, 1984), and M. vancouverensis (Milko, 1984), managers of Vancouver Island marmots have attempted to select release sites that maximize access to safety terrain and foraging habitat. Surprisingly, I found little influence of landscape-level habitat descriptors on site fidelity, perhaps due to the limited range of site quality exhibited by  27  release sites, or because fidelity was not strongly affected by the proximity of talus, meadow or old or cutblock forest within 1km of release burrows. I also found no correlation between site fidelity and fine-scale forage and visibility characteristics measured at a subset of release sites, but acknowledge that my dataset including these variables was limited and potentially influenced by differences in conditions related to release date and year of sampling. Captive-bred animals are often released with the goal of site fidelity, but experience shows that many released animals go through a phase of unpredictable movement before establishing a stable home range (e.g. Mandrillus sphinx, Peignot et al., 2008; Myadestes palmeri, Tweed et al., 2003; Vulpes vulpes, Robertson and Harris, 1995). Movement may expose newly released animals to higher risk of predation or accidental death than experienced by site-faithful animals (but see Banks et al., 2002) and it may also reduce a hibernators’ ability to accumulate body fat. However, I was unable to detect a relationship between dispersal distance and subsequent survival in newly released Vancouver Island marmots (Figs. 4 and 7). Although release site fidelity may be desirable when attempting to establish new colonies, my results suggest that managers may gain greater benefits by focusing on predictors linked to increased survival.  2.4.2  Active season survival  Predation was the predominant cause of active season mortality for recently reintroduced captive-bred marmots but was not the main source of mortality for new releases overall. My models suggested that release date was a strong positive predictor for active season survival, perhaps because marmots released later in the active season were vulnerable to predation for a fewer number of days (Figure 5, Table 3). Predation was a key cause of mortality for wild-  28  born marmots between 1992 and 2004 (Bryant and Page, 2005). Although I found similar patterns to Aaltonen et al. (2009) with regard to the cause of active season mortality for captive-bred marmots, in particular confirming that eagles and cougars were often implicated, Aaltonen et al. (2009) also suggested that predation was the most influential factor affecting annual survival of marmots. My results suggest that managers can increase active season survival by managing release dates to limit exposure to risk. In addition to release date, the models also predicted that age at release and sex influenced active season survival. Aaltonen et al. (2009) estimated that captive-bred marmots released as 2-year-olds or older survived better annually than yearlings (0.774, 95% CI = 0.649-0.864 vs. 0.602, 95% CI = 0.459-0.729, respectively), but did not use a separate age class for adults. I also found that 2-year-olds survived best (S = 97%), but estimated that adults were much more susceptible to mortality during the active season (S = 77%) than either 2-year-old marmots or yearlings (S = 93%). My models predicted that female survival was 9% greater than that of males. Because Vancouver Island marmots sometimes breed in polygamous groups (Bryant, 1996; D. Doyle, pers. comm.), modest skews in the sex ratio of surviving reintroduced adults may be of limited consequence to population growth if balanced by sex ratios of pups and sex-biased survival of wild and established adults. Finally, greater cover of talus habitat near release sites is thought to improve active season survival because boulder piles facilitate predator detection and act as refugia (e.g. M. broweri, Bee and Hall, 1956, in Gunderson et al., 2009; M. caligata, Holmes, 1984; Barash, 1973; M. flaviventris, Svendsen, 1974). Talus cover at each release site ranged from 0 to 11% in 1km buffers around release burrows and was associated with predicted survival probabilities of 90% to 100%, respectively. This suggests that marmots released to sites with  29  little available talus may suffer higher mortality on average, perhaps because a lack of talus impeded their ability to detect or escape from predators. Aside from release date, age, sex, and talus cover, other factors previously identified in literature as influential of active season survival were poorly supported, as were the remaining variables included for management purposes (Figure 4).  2.4.3  Overwinter survival  My results indicate that poor overwinter survival is a key factor limiting first-year reintroduction success for captive-bred Vancouver Island marmots (Table 3). This finding is particularly interesting given that Vancouver Island marmots are reported to be efficient hibernators because of low daily mass loss in captivity (Bryant and McAdie, 2003) and high overwinter survival in wild-born individuals (Figure 6; Bryant and Page, 2005). Hibernation is a behavioural adaptation to the food shortages of winter (Heldmaier et al., 2004), and all species of marmots hibernate. During hibernation, metabolic, heart, and ventilation rates are greatly reduced, allowing marmots to survive at lower body temperatures and to expend less energy than when active (Heldmaier et al., 2004). Nevertheless, marmots must cycle through periods of torpor, arousal, and euthermia in order to maintain their body temperature above a critical survival threshold (Arnold, 1995), and both arousal and euthermia are energetically costly (Heldmaier et al., 2004). For instance, individual studies of M. marmota and M. caudata found that marmots lost an average of 30% (SD = 6.2, n = 250; Arnold, 1990b) and 41% (SE = 2.7, n = 16; Blumstein and Arnold, 1998) of their respective fall body mass during the course of hibernation. It is plausible, then, that my observation of relatively poor  30  overwinter survival in newly released animals was related to their body condition at immergence or their ability to hibernate efficiently. Release date had a strong relative influence on overwinter survival, and marmots that were released early in the active season were predicted to survive better than those released at later dates. This result was contrary to my expectations. Because energy expenditure in captivity is likely to be lower than that in the wild, captive marmots may store calories more rapidly or efficiently than wild or released animals. Therefore, it seemed reasonable to assume that marmots kept in captivity later in the summer would weigh more at immergence than marmots released earlier, giving late-released marmots a relative advantage in body condition when entering hibernation. However, my results suggested that releasing marmots earlier in the active season could greatly increase their likelihood of overwinter survival. Overwinter survival of captive-bred marmots was consistently lower than that of established (captive-bred marmots in their second and subsequent wild hibernations) or wild-born marmots (Figure 6). This implies that the differences in overwinter survival that I observed between newly released and wild-born or established marmots exist primarily in the first year post-release. Anecdotal reports from the field suggest that survival differences could be related in part to the timing of immergence. Most wild-born or established marmots were confirmed in torpor or in plugged hibernacula between late September and the end of October. In contrast, at new release sites, signs of marmot activity above ground were often observed into November (D. Doyle, pers. comm.). Delayed immergence may cause newly released marmots to rely more heavily on senesced plant material as forage and may expose them to inclement fall weather, both of which may act to deplete body condition before entering  31  hibernation. However, delayed immergence by captive-bred marmots would not account for the dramatically lower overwinter survival of captive-bred marmots released in 2009 and 2010 relative to those released in 2006-2008 (Figure 6). One possible explanation for abnormal hibernation behaviour in newly released marmots might be their existence as an artifact of hibernation patterns in captivity. A recent study of arctic ground squirrels (Urocitellus parryii) found evidence that thermoregulatory behaviours also differ between wild-born squirrels in captivity and free-living squirrels in the wild (Sheriff et al., 2012). A greater proportion of captive squirrels (>60%, n = 99) than freeliving squirrels (25.6%, n = 141) experienced pre-hibernation ‘test drops’, repeated bouts of torpor with progressively lower minimum body temperatures maintained for <24h before arousal (Sheriff et al., 2012). Test drops were also more frequent and shorter in length for captive than wild squirrels. Bryant and McAdie (2003) compared Vancouver Island marmots in the wild and at breeding facilities to find that, on average, captive marmots entered hibernation six weeks later and emerged five weeks earlier than wild marmots. Unfortunately, immergence and emergence dates could not be measured in a similar way for both wild and captive cohorts, so estimates of hibernation timing were confounded by differing definitions and the inability to observe animals continuously in the wild (M. McAdie, pers. comm.). Hibernation behaviour in newly released Vancouver Island marmots might also be influenced by social and environmental stress, particularly ‘release stress’ from their transportation and introduction to natural habitat. Stress produces physiological responses that could impair health, impede cognitive processing, and alter natural behaviour in released animals (Twixeira et al., 2007), but often abates with time. My results predicted that marmots  32  released earlier in the active season would have higher overwinter survival than those released later, perhaps because they had longer to recover from stress and synchronize with wild hibernation patterns. Studies show that arctic ground squirrels in the wild begin preparing for hibernation weeks before initiating torpor; average, minimum, and maximum daily body temperatures started decreasing at least 60 days prior to hibernation (n = 141), with significant declines in temperature documented at 45, 45, and 40 days, respectively (Sheriff et al., 2012). No similar studies exist for the Vancouver Island marmot, but records from the captive population show lower average body temperatures in the first week of September relative to those recorded in August (M. McAdie, pers. comm.). To test for differences in hibernation timing based on origin and release date and evaluate competing hypotheses of stress and the influence of captivity in delaying hibernation in newly released marmots, managers should consider more intensive monitoring in the months prior to hibernation, especially to determine ‘dates of first torpor’ (body temperature <30˚C, M. monax, Zervanos et al., 2010) in released, wild-born, and established animals. Marmots in captivity are not usually implanted with transmitters until a few weeks prior to release; therefore, to enable comparisons with marmots in the wild, a subset of captive animals should also be implanted. If delayed hibernation is a function of short-term release effects, data will show that marmots released late in the active season enter torpor late relative to those released at earlier dates. However, if delayed hibernation is the result of artificial hibernation patterns developed in captivity, I would expect both captive animals and new releases to have later average dates of first torpor relative to established and wild-born marmots.  33  Released captive-bred pups suffered slightly lower overwinter survival rates (50.0%) than the average of other captive-bred age classes (58.3%; Table 3). This difference reflects a pattern also observed in wild-born pups of other marmot species (M. caudata aurea, Blumstein and Arnold, 1998; M. flaviventris, Lenihan and Van Vuren, 1996). Some marmots participate in socially synchronous thermoregulation during hibernation (M. caudata aurea, Blumstein and Arnold, 1998; M. marmota, Arnold 1999, 1998, 1993, 1990b; but see M. flaviventris, Armitage and Woods, 2003; Blumstein et al., 2004). Although hibernation in social groups is believed to reduce the energetic output of all individuals via passive ‘rewarming’ (Arnold, 1988), the largest benefactors are thought to be pups, which often lack the endogenous fat reserves used for rewarming (Armitage et al., 1976). However, social hibernation can simultaneously increase the mass loss of related group members because pups require more frequent rewarming (Arnold, 1990b; but see Blumstein and Arnold, 1998). Given the importance of increasing overwinter survival in newly released captive-bred marmots to enhance overall success, and the potential for pups to create thermoregulatory stress during hibernation, I recommend caution when including pups in future releases and encourage the prioritization of data collection to examine potential links between group composition and survivorship in future.  2.5  Implications and recommendations  My results suggest that release date was highly influential of site fidelity, active season survival, and overwinter survival in reintroduced Vancouver Island marmots. However, because later release dates were positively linked to site fidelity and active season survival, but negatively to overwinter survival, managers face a dilemma when trying to maximize  34  success in the first year post-release. Site fidelity is necessary to accelerate colony establishment and connectivity, and although many marmots that abandoned their release sites failed to contribute to population growth, some joined or established new colonies where they reproduced successfully (pers. obs.). For this reason, maximizing survival of new releases appears to be a more important short-term goal than site fidelity. However, given low average rates of overwinter survival observed during the first-year following release, my results also suggest that managers should achieve greater overall success by focusing on improving overwinter rather than active season survival. Realistically, release dates are limited to a small window of time during the active season, and the best date for overall survival may itself vary annually. Until late June, most release sites are under snow and food and shelter are limited. By late September, forage quality wanes and stormy weather makes helicopter releases unfeasible. The release date chosen to maximize overall survival could be July 16 (Figure 8), the date on which the predicted probabilities of active season and overwinter survival intersect, and at approximately 86% probability of success. This would result in a joint probability of 74% that a marmot would survive the active season and then hibernation. However, this date does not take into account the shorter length of the active season relative to hibernation (Figure 9). Estimates of weighted average survival suggest that concentrating releases in late June to early July may lead to the greatest increase in overall success in future. Moreover, given that early release dates were associated with lower active season survival rates, managers could also consider emphasizing the release of 2-year-old and yearling marmots over adults, and selecting release sites with nearby talus to give marmots the best opportunity to survive the active season. Monitoring newly released marmots should also be continued to better  35  estimate the potential costs of early release on active season survival, and to test ways to improve overwinter survival via group composition. Because it can take many years for endangered species reintroduction programs to establish viable populations (Fischer and Lindenmayer, 2000; Griffith et al., 1989; Kleiman, 1989), it is advisable to conduct early and frequent evaluations of program performance (Kleiman et al., 1994; Seddon, 1999). I identified release date as the variable with the strongest overall influence on first-year reintroduction success in captive-bred Vancouver Island marmots, a finding that may also be influential in other species that display seasonally adaptive behaviours including hibernation, estivation and migration. However, because the influence of release date was opposite to expectations, my results illustrate a point made by Armstrong and Seddon (2008), that all captive release programs should incorporate measures to estimate the influence of management practices on success rather than assuming that strategies will work as expected.  36  Table 1. Summary of the number of captive-bred marmots released in each year of this study that met the criteria for analysis (see 2.2.4), and that were used to predict site fidelity, active season and overwinter survival (in bold). Overwinter survival had the smallest sample size because marmots that died during the active season were excluded, and additional marmots went missing between fall and the following spring.  37  Table 2. Descriptions, range of values, and predicted direction of influence for variables used to predict site fidelity, active season survival, and overwinter survival in newly-released, captive-bred Vancouver Island marmots (Marmota vancouverensis). Covariates were supported by literature on this or similar species, or were potentially amenable to management.  38  Table 3. Rates of site fidelity, active season and overwinter survival in newly released marmots. Estimated success, parameter estimates, unconditional standard errors, and 95% unconditional confidence intervals are provided for variables with >0.5 relative importance to predictive models of site fidelity, active season and overwinter survival. Models for each success metric were constructed using ‘all-subsets analysis’, and parameter estimates were attained by averaging models with AICc Δi <7 (see Statistical analyses, 2.2.5.3).  39  Table 4. Discriminatory performance of averaged models when predicting site fidelity, active season and overwinter survival. Prevalence measures the number of occurrences of success in the data, and unbalanced prevalence is known to bias sensitivity (the proportion of correctly predicted successes), specificity (the proportion of correctly predicted failures), and Cohen’s kappa coefficient (the proportion of correct predictions after accounting for chance). The True Skill Statistic (TSS) operates independently of prevalence (see Discussion, 2.4), and suggested that overwinter survival was the most predictable measure of success, but models for site fidelity and active season survival were still more predictive of success than chance alone.  40  Figure 1. Location of 30 release sites (2003-2010) for captive-bred Vancouver Island marmots, British Columbia, Canada.  41  Figure 2. Relative importance of 13 predictors of site fidelity for newly released, captive-bred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date and resident female received >0.5 relative support as predictors and were included in 100% and 75% of models in the AICc Δi <7 subset, respectively.  42  Figure 3. Predicted probability of site fidelity in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would hibernate within 1km of the release site, and were calculated using model-averaged parameter estimates and varying the values of release date and resident female presence while maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Site fidelity is predicted to increase with release date, but was lower overall for marmots released to sites with resident females within 500m of the release burrow.  43  Figure 4. Relative importance of 10 predictors of active season survival for newly released, captive-bred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date, age class, sex and talus cover received > 0.5 relative support as predictors and were included in 100%, 86%, 74% and 73% of models in the AICc Δi <7 subset, respectively.  44  (a)  (b)  (c)  Figure 5. Predicted probabilities of active season survival in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would survive to hibernation, and were calculated using model-averaged parameter estimates and varying (a) age class, (b) sex, and (c) talus cover, while maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Active season survival is predicted to increase with release date and was higher for females, 2-year-olds, and marmots released to sites with talus cover within 1km of the release burrow.  45  Figure 6. Estimated annual overwinter survival and 95% confidence intervals for wild-born marmots (never in captivity), established marmots (captive-bred and with >1 prior wild hibernation), and newly released captivebred marmots (2006-2010). Captive-bred marmots that were ‘newly released’ in one year were then considered to be ‘established’ in subsequent years. Estimates of survival were much higher for wild-born and established marmots than for captive-bred marmots undergoing their first wild hibernation, particularly in 2009 and 2010.  46  Figure 7. Relative importance of 10 predictors of overwinter survival for newly released, captive-bred marmots that had been identified prior to my analysis (see Table 2). The horizontal bar indicates relative support of 0.5 from models in the AICc Δi <7 subset, the value that predictors were required to exceed in order to be considered for further analyses. Release date was the only variable with >0.5 relative support, and was included in 88% of models in the AICc Δi <7 subset. 'Pre-rel. hibernation days' refers to the number of days a marmot spent in hibernation prior to release, used as a proxy for quality of past hibernation experience.  47  Figure 8. Predicted probabilities of site fidelity, active season and overwinter survival in newly released captive-bred marmots. The plotted lines represent estimates of the probability that an animal released on that date would achieve each success metric, and were calculated using model-averaged parameter estimates and maintaining covariates at their medians or modes (see Statistical Analyses, 2.2.5.3). Site fidelity and active season survival are predicted to increase with release date, but overwinter survival was predicted to be highest for marmots released early in the active season.  48  Figure 9. Predicted probabilities of active season and overwinter survival, and weighted average survival to the spring following release. Average survival was calculated by weighting the probabilities of active season and overwinter survival on each potential release date by the relative fraction of time required to achieve those individual success metrics (73 and 243 days, respectively, assuming a non-leap year).  49  Chapter 3: Conclusion The Vancouver Island marmot is a critically endangered ground squirrel that has been captive-bred for conservation purposes and reintroduced to the wild since 2003 in an intensive program aimed at creating a self-sustaining, wild population (Vancouver Island Marmot Recovery Team, 2008). Previous research on this species has focused on understanding its behaviour (Blumstein et al., 2001; Brashares et al., 2010; Heard, 1977), life history (Bryant, 2008, 2005; Bryant and McAdie, 2003) and mortality patterns (Aaltonen et al., 2009; Bryant 2000, 1996; Bryant and Page, 2005). In contrast, my research focused on identifying variables influential to reintroduction success, and was therefore the first to ask how the characteristics of individual marmots, release practices, and the local and landscapelevel attributes of release habitats can predict site fidelity, active season survival, and overwinter survival for each released marmot in their first year in the wild. In this concluding chapter of my thesis I summarize my results, present additional predictors described in literature that were not evaluated in my thesis, and discuss the implications of my results for future reintroductions of captive-bred Vancouver Island marmots. Overwinter survival was identified as a key factor limiting reintroduction success in captive-bred Vancouver Island marmots. Released marmots achieved high active season survival and showed moderate fidelity to release sites, but achieved low overwinter survival in their first wild hibernation relative to wild-born and established marmots. I determined that release date was the most influential variable among the effects examined, and a more valuable predictor of success than age, sex, marmot experience in captivity, management practices, and habitat characteristics used to predict site fidelity and active season and overwinter survival. However, whereas later release dates were positively associated to site  50  fidelity and active season survival, they were negatively related to overwinter survival. Although overwinter survival was not linked to other predictors with strong relative importance, site fidelity was also influenced by the presence of resident females at the release site, and active season survival of new releases was influenced by age at release, sex, and talus cover near the release site. Of these variables, release date stands as the best candidate for future research and evaluation given its pervasive appearance in my models of release success. In trying to determine the influence of each explanatory variable on reintroduction success, I employed an information-theoretic approach and all-subsets analysis to predict post-release site fidelity and active season and overwinter survival. Attaining reliable parameter estimates in logistic regression requires that research balance the number of parameters examined to the availability of relevant data to avoid ‘over-parameterization’. Some consequences of over-parameterization include ‘finite-sample bias’ (Greenland, 1989) in parameter estimates, uncertainty in estimated confidence intervals, increased occurrence of Type I errors when using frequentist hypothesis tests, and problems with model convergence (Vittinghoff and McCulloch, 2007). However, recommendations in the literature differ with respect to the general guidelines to which researchers should adhere. Burnham and Anderson (2002) recommend a maximum of n/10 parameters, whereas Peduzzi et al. (1996, 1995) recommend that logistic regressions be supported by 10 events-per-variable (EPV). By this measure, even large datasets with unbalanced outcomes can be limited in the number of variables they attempt to fit. Vittinghoff and McCulloch (2007) showed via simulation that EPV ratios as low as 5-9 can be sufficient to minimize bias and uncertainty. EPV ratios for global models constructed to predict site fidelity and overwinter survival were within that  51  range, and should therefore be relatively robust to bias and uncertainty, especially given my use of model-averaging, which reduces these potential errors (Burnham and Anderson, 2002). In contrast, the EPV ratio for the global model of active season survival was below the threshold of five and probably contributed to minor problems of model convergence reported in Chapter 2. Many aspects of marmot behaviour have been related to the intrinsic traits of individuals (e.g. Heard, 1977; Lea and Blumstein, 2011), to variation in social organization (e.g. Allainé, 2000; Armitage, 1975; Barash, 1989, 1974; Heard, 1977), and to the characteristics of habitats affecting foraging behaviour and predation risk (e.g. Blumstein et al., 2006; Carey and Moore, 1986; Holmes, 1984). I also expected that behaviour of newly released captive-bred marmots would influence the success in their first year in the wild. Accordingly, I included many of the predictors shown previously to influence the behaviour and survival of wild marmots. However, additional predictors described in the literature on marmots could have been evaluated if the appropriate data had been available. In particular, I would expect social cohesion and familiarity within release groups to be predictive of site fidelity. I now describe briefly the mechanisms of these potential relationships and their implications for newly released captive-bred Vancouver Island marmots. Literature on other social species suggests that for many, ‘social cohesion’ is a key factor affecting philopatry and dispersal (e.g. Bekoff, 1977). I included sex, age and relatedness as a necessarily limited set of factors with the potential to influence marmot behaviour and survival post-release, but it is possible that these variables were too general to sufficiently represent the potential effects of group sociality. For instance, M. flaviventris exhibited greater rates of philopatry (females, Armitage et al., 2011; Blumstein et al., 2009)  52  or delayed dispersal (males, Downhower and Armitage, 1981) when they more often displayed affiliative interactions with other marmots in the colony. Direct evaluation of affiliative behaviours, such as play-fighting or allogrooming (M. marmota, Perrin et al., 1993) in release groups prior to release may therefore be useful in determining individual contributions to social cohesion and the likelihood of post-release site fidelity. I would expect higher likelihood of site fidelity in captive-bred Vancouver Island marmots that more frequently participate in affiliative interactions with others in their release groups than for marmots that rarely engage in affiliative behaviour. Familiarity of marmots within a release group may also contribute to the expression of affiliative behaviour. Perrin et al. (1993) reported high cohesion within groups of M. marmota but observed that interactions between groups were frequently agonistic, suggesting that lack of familiarity contributes to aggression in marmots. Similar processes could also influence site fidelity in captive-bred Vancouver Island marmots because members of release groups are introduced to each other at widely varying times prior to release. The length of association depends in practice upon many factors including relatedness, the facilities involved, the enclosures in which marmots were born and housed, their genetic status or relatedness to others in captivity, and their age at release. On some occasions, release groups were formed while the marmots were hibernating. Thus, within a single release group of marmots, it is possible that one or more subgroups with greater familiarity may exist and influence cohesion post-release. Release groups with high familiarity and frequent expression of affiliative behaviours may still suffer from low group cohesion if members of the release group breed in the spring prior to their release. Marmots breed immediately following arousal from hibernation, and  53  breeding activity in many rodents has been associated to higher rates of intraspecific aggression (Ebensperger and Blumstein, 2007). Aggressive behaviour in marmots is often initiated by parous females of species in which infanticide has been documented (e.g. M. caudata aurea, Blumstein, 1997; M. flaviventris, Armitage et al., 1979; M. marmota, Farand et al., 2002), ostensibly as a means to protect their pups from potential threats (Ebensperger and Blumstein, 2007). Similarly, in captive Vancouver Island marmots, aggressive behaviour is most often initiated by pregnant or post-partum females towards their mates (Casimir, 2005; Keeley et al., 2011) and other marmots in their enclosure (M. McAdie, pers. comm.). If site fidelity of released marmots is related to social cohesion within release groups, it could be tempered by reproduction within groups in the months prior to release. Given the importance of social dynamics to behaviour of other species of marmots, a closer examination of the influence of pre-release management practices on social cohesion and release site fidelity may be worthwhile. There are many reasons to feel optimistic about the recovery of the Vancouver Island marmot. The reintroduction program has helped to re-establish colonies and expand the distribution of marmots far beyond the six colonies reported in 2006 (Brashares et al., 2010). Several years of successful reproduction by wild-born and established marmots has also helped to increase the population size. In 2009 and 2011, the number of marmots born in the wild (68 and 65) were essentially equal to the number of marmots released in those years (68 and 66, respectively). Thus, the wild population is contributing positively to population recovery and also being supported by continuing reintroductions. Yet, despite these positive indicators, there are still several processes in effect with the potential to threaten the recovery of the species.  54  Reintroduction literature suggests that identification and elimination of the causes behind a species’ decline are critical steps to achieving success (Fischer and Lindenmayer, 2000; IUCN, 1998; Kleiman, 1989). However, determining the mechanisms of decline can be challenging for species already declining by the time of initial study (Armstrong and Seddon, 2008). This is true for the Vancouver Island marmot, which was first surveyed in 1972 (Bryant and Janz, 1996). Studies suggest that recent population declines can be attributed to episodic predation events (Bryant, 2000) by cougars, wolves, and golden eagles (Bryant and Page, 2005; Aaltonen et al., 2009), and suggest a variety of hypotheses to explain how a secondary prey species could be decimated by its natural predators (Bryant and Page, 2005). Prior to 1972, however, much less is known about the population dynamics of this species. Written records, observations, and specimens (Bryant and Janz, 1996), as well as faunal remains (Nagorsen et al., 1996), suggest that Vancouver Island marmots were historically more widely distributed and more numerous. A geographic contraction of the species’ range could have been symptomatic of an already declining population, or alternately, the driver of population decline itself. Site-specific studies suggest the latter is true at some southern locations, with evidence supporting a transition from open to forested habitat (Hebda, 2004; Milko, 1984), a trend which is projected to continue into the future (Laroque et al., 2000). If tree invasion does continue and the habitat used by Vancouver Island marmots is pushed higher in elevation, this elevation shift could also necessitate a northward shift in distribution on Vancouver Island because the southernmost hills are relatively low in elevation. There is no reason to believe that predation pressure on Vancouver Island marmots will ease in the near future, or that the present availability and distribution of marmot habitat  55  will be maintained in perpetuity. Therefore, it is essential to protect the wild population against extinction through the volume of marmots in the wild and the geographic distribution of these marmots across Vancouver Island. The original Recovery Strategy for the Vancouver Island marmot specified a recovery goal of 400-600 marmots in the wild, distributed amongst three self-sustaining metapopulations on Vancouver Island (Janz et al., 1994). There have now been eight years of releases and 301 captive-bred marmots reintroduced to the wild, however most wild and established adults, colonies with successful immigrants, and mountains with wild-born pup litters are found in the south region. The wild population is still highly vulnerable to stochastic events, and recovery efforts should be continued with a focus on building and strengthening subpopulations in geographic areas other than in the south. There are new options available to managers interested in the recovery of the Vancouver Island marmot that were not possible when reintroductions began. For example, several wild and established colonies are now thought by managers to be large enough to donate wild-born marmots for translocation to new release sites (D. Doyle, pers. comm.). This is a potentially important advance because Griffith et al. (1989) reported that projects that used translocations of wild individuals were slightly more successful than those relying on the reintroduction of captive-bred animals (46% vs. 38%). Rebuilding wild populations could also be accelerated if the overwinter survival in newly released, captive-bred marmots could be increased. My results imply that focused attempts to optimize release dates could represent a key avenue to increased reintroduction success. A priority for future research is identification of the biological mechanisms driving overwinter survival in reintroduced marmots.  56  References Aaltonen, K., Bryant, A.A., Hostetler, J.A., Oli, M.K., 2009. Reintroducing endangered Vancouver Island marmots: survival and cause-specific mortality rates of captive-born versus wild-born individuals. Biological Conservation, 142(10), 2181-2190. Allainé, D., 2000. Sociality, mating system and reproductive skew in marmots: evidence and hypotheses. Behavioural Processes, 51(1-3), 21-34. Allouche, O., Tsoar, A., Kadmon, R., 2006. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). Journal of Applied Ecology, 43(6), 1223-1232. Anderson, D.R., Burnham, K.P., 2002. Avoiding pitfalls when using information-theoretic methods. The Journal of Wildlife Management, 66(3), 912-918. Armitage, K.B., 1999. Evolution of sociality in marmots. Journal of Mammalogy, 80, 1-10. Armitage, K.B., 1975. Social behavior and population dynamics of marmots. Oikos, 26(3), 341-354. Armitage, K.B., Blumstein, D.T., Woods, B.C., 2003. Energetics of hibernating yellowbellied marmots (Marmota flaviventris). Comparative Biochemistry and Physiology – Part A: Molecular & Integrative Physiology, 134(1), 101-114. Armitage, K.B., Johns, D., Andersen, D.C., 1979. Cannibalism among yellow-bellied marmots. American Society of Mammalogists, 60(1), 205-207. Armitage, K.B., Salsbury, C.M., Barthelmess, E.L., Gray, R.C., Kovach, A., 1996. Population time budget for the yellow-bellied marmot. Ethology Ecology & Evolution, 8, 6795. Armitage, K.B., Van Vuren, D.H., Ozgul, A., Oli, M.K., 2011. Proximate causes of natal dispersal in female yellow-bellied marmots, Marmota flaviventris. Ecology, 92(1), 218-227. Armitage, K.B., 2007. Evolution of sociality in marmots: it begins with hibernation. In Wolff, J., Sherman, P.W. (Eds.), Rodent Societies: An Ecological & Evolutionary Perspective. University of Chicago Press, Chicago, US. Pp. 357-367.  57  Armstrong, D.P., and Seddon, P.J., 2008. Directions in reintroduction biology. Trends in Ecology & Evolution, 23(1), 20-25. Arnold, W., 1995. Social behavior and telemetric assessment of thermoregulation in hibernating marmots. In E. Alleva et al. (Eds.), Behavioural Brain Research in Naturalistic and Semi-Naturalistic Settings, Kluwer Academic Publishers: Netherlands, 395-411. Arnold, W., 1990a. The evolution of marmot sociality: I. Why disperse late. Behavioral Ecology and Sociobiology, 27, 229-237. Arnold, W., 1990b. The evolution of marmot sociality: II. Costs and benefits of joint hibernation. Behavioral Ecology and Sociobiology, 27, 239-246. Arnold, W., 1988. Social thermoregulation during hibernation in alpine marmots (Marmota marmota). Journal of Comparative Physiology B, 158(2), 151-156. Arnold, W., Dittami, J., 1997. Reproductive suppression in male alpine marmots. Animal Behaviour, 53, 53-66. Bar-David, S., Saltz, D., Dayan, T., Perelberg, A., Dolev, A., 2005. Demographic models and reality in reintroductions: Persian fallow deer in Israel. Conservation Biology, 19(1), 131138. Barash, D.P., 1989. Marmots: social behavior and ecology. Stanford University Press: Stanford, CA. pp.368. Barash, D.P., 1974. The social behavior of the hoary marmot (Marmota caligata). Animal Behaviour, 22(1), 256-261. Barash, D.P., 1973. Habitat utilization in three species of subalpine mammals. Journal of Mammalogy, 54(1), 247-250. Bartoń, K., 2012. MuMIn: multi-model inference. R package, version 1.7.2. Retrieved May 15, 2012, from: <http://r-forge.r-project.org/projects/mumin> Bates, D., Maechler, M., Bolker, B., 2011. lme4: linear mixed-effects models using S4 classes. R package version 0.999375-42. Retrieved April 15, 2012, from: <http://r-forge.rproject.org/projects/lme4>  58  Beck, B.B., Rapaport, L.G., Price, M.R.S., Wilson, A.C., 1994. Reintroduction of captiveborn animals, in: Olney, P.J.S., Mace, G.M., Feistner, A.T.C., (Eds). Creative conservation: interactive management of wild and captive animals. Chapman and Hall: London, UK. pp.265-286 Bekoff, M., 1977. Mammalian dispersal and the ontogeny of individual behavioral phenotypes. American Naturalist, 111, 715-732. Blumstein, D.T., 1999. Alarm calling in three species of marmots. Behaviour, 136(6), 731757. Blumstein, D.T., Arnold, W., 1998. Ecology and social behavior of golden marmots (Marmota caudata aurea). Journal of Mammalogy, 79(3), 873-886. Blumstein, D.T., Daniel, J.C., Bryant, A.A., 2001. Anti-predator behavior of Vancouver Island marmots: using congeners to evaluate abilities of a critically endangered mammal. Ethology, 107(1), 1-14. Blumstein, D.T., Ozgul, A., Yovovich, V., Van Vuren, H., Armitage, K.B., 2006. Effect of predation risk on the presence and persistence of yellow-bellied marmot (Marmota flaviventris) colonies. Journal of Zoology, 270(1), 132-138. Blumstein, D.T., Wey, T.W., Tang, K., 2009. A test of the social cohesion hypothesis: interactive female marmots remain at home. Proceedings of the Royal Society B: Biological Sciences, 276(1669), 3007-3012. Blumstein, D.T., Evans, C.S., Daniel, J.C., 1999. An experimental study of behavioural group size effects in tammar wallabies (Macropus eugenii). Animal Behaviour, 58, 351-360. Bowkett, A.E., 2009. Recent captive-breeding proposals and the return of the ark concept to global species conservation. Conservation Biology, 23(3), 773-776. Brashares, J., Werner, J.R., Sinclair, A.R.E., 2010. Social ‘meltdown’ in the demise of an island endemic: Allee effects and the Vancouver Island marmot. Journal of Animal Ecology, 79(5), 965-973. Bryant, A.A., 2005. Reproductive rates of wild and captive Vancouver Island marmots (Marmota vancouverensis). Canadian Journal of Zoology, 83(5), 664-673.  59  Bryant, A.A., 2000. Relative importance of episodic versus chronic mortality in the decline of Vancouver Island marmots (Marmota vancouverensis), in: Darling, L. (Ed.), Proceedings of a conference on the biology and management of species and habitats at risk. Queen’s Printer, Victoria, pp. 189-195. Bryant, A.A., 1998. Metapopulation ecology of Vancouver Island marmots (Marmota vancouverensis). PhD thesis, University of Victoria, pp 1-96. Bryant, A.A., 1996. Reproduction and persistence of Vancouver Island marmots (Marmota vancouverensis) in natural and logged habitats. Canadian Journal of Zoology, 74, 678-687. Bryant, A.A., Janz, D.W., 1996. Distribution and abundance of Vancouver Island marmots (Marmota vancouverensis). Canadian Journal of Zoology, 74(4), 667-677. Bryant, A.A., Janz, D.W., deLaronde, M.C., Doyle, D.D., 2002. Recent Vancouver Island marmot (Marmota vancouverensis) population changes. In Armitage, K.B., Rumianstev, V.U., (eds.): Holarctic marmots as a factor of biodiversity, ABF Publishing House (Moscow). 411 pp. Bryant, A.A., McAdie, M., 2003. Hibernation ecology of wild and captive Vancouver Island marmots (Marmota vancouverensis), in: Ramousse, R., Allainé, D., Le Berre, M., (Eds.), Holarctic marmots as a factor of biodiversity. ABF Publishing House: Moscow, pp.159-166. Bryant, A.A., Page, R.E., 2005. Timing and causes of mortality in the endangered Vancouver Island marmot (Marmota vancouverensis). Canadian Journal of Zoology, 83(5), 674-682. Burnham, K.P., Anderson, D.R., 2002. Model selection and multimodel inference: a practical information-theoretic approach, second ed. Springer-Verlag, New York, NY. Cardini, A., 2003. The geometry of the marmot (Rodentia: Sciuridae) mandible: phylogeny and patterns of morphological evolution. Systematic Biology, 52(2), 186-205. Cardini, A., O’Higgins, P., 2004. Patterns of morphological evolution in Marmota (Rodentia, Sciuridae): geometric morphometrics of the cranium in the context of marmot phylogeny, ecology and conservation. 82(3), 385-407. Cardini, A., Hoffmann, R.S., Thorington, R.W. Jr., 2005. Morphological evolution in marmots (Rodentia, Sciuridae): size and shape of the dorsal and lateral surfaces of the cranium. Journal of Zoological Systematics and Evolutionary Research, 43(3), 258-268.  60  Cardini, A., Nagorsen, D., O’Higgins, P., Polly, P.D., Thorington, R.W. Jr., Tongiorgi, P., 2009. Detecting biological distinctiveness using geometric morphometrics: an example case from the Vancouver Island marmot. Ethology Ecology & Evolution, 21, 209-223. Carey, H.V., 1985. The use of foraging areas by yellow-bellied marmots. Oikos, 44, 273-279. Carey, H.V., Moore, P., 1986. Foraging and predation risk in yellow-bellied marmots. American Midland Naturalist, 116(2), 267-275. Casimir, D.L., 2005. Reproductive behavior of Vancouver Island marmots, Marmota vancouverensis: conclusions from a conservation breeding program. M.Sc. thesis, University of Calgary, pp 1-124. Cohen, J., 1960. A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. COSEWIC, 2009. Canadian wildlife species at risk. Committee on the Status of Endangered Wildlife in Canada. Retrieved April 15, 2012, from: <http://www.cosewic.gc.ca/eng/sct0/rpt/ rpt_csar_e.cfm> Dearden, P., Hall, C., 1983. Non-consumptive recreation pressures and the case of the Vancouver Island marmot (Marmota vancouverensis). Environmental Conservation, 10(1), 63-66. Dochtermann, N.A., Jenkins, S.H., 2011. Developing multiple hypotheses in behavioral ecology. Behavioral ecology and sociobiology, 65(1), 37-45. Downhower, J.F., Armitage, K.B., 1981. Dispersal of yearling yellow-bellied marmots (Marmota flaviventris). Animal Behaviour, 29(4), 1064-1069. Ebensperger, L.A., Blumstein, D.T., 2007. Nonparental infanticide, in: J. Wolff and P.W. Sherman (Eds.), Rodent Societies: An Ecological & Evolutionary Perspective. The University of Chicago Press: Chicago, US. Edelman, A.J., 2003. Marmota olympus. Mammalian Species, 736, 1-5. Farand, É., Allainé, D., Coulon, J., 2002. Variation in survival rates for the alpine marmot (Marmota marmota): effects of sex, age, year, and climatic factors. Canadian Journal of Zoology, 80, 342-349.  61  Fielding, A.H., Bell, J.F., 1997. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation, 24(1), 38-49. Fischer, J., Lindenmayer, D.B., 2000. An assessment of the published results of animal relocations. Biological Conservation, 96(1), 1-11. Freeman, E.A., Moisen, G.G., 2008. A comparison of the performance of threshold criteria for binary classification in terms of predicted prevalence and kappa. Ecological Modelling, 217(1-2), 48-58. Gladders, A.D., 2003. Predation behavior of Vancouver Island Cougar (Puma concolor vancouverensis) and its relation to micro- and macroscale habitat. M.Sc. thesis, University of British Columbia, Vancouver, BC. Goh, K.M.L., 2000. Macrohabitat selection by Vancouver Island cougar (Puma concolor vancouverensis). M.Sc. thesis, The University of British Columbia, Vancouver, BC. Greenland, S., 1989. Modeling and variable selection in epidemiologic analysis. American Journal of Public Health, 79(3), 340-349. Griffith, B., Scott, T.M., Carpenter, J.W., Reed, C., 1989. Translocation as a species conservation tool: status and strategy. Science, 245(4917), 477-480. Grizzell, R.A. Jr., 1955. A study of the southern woodchuck, Marmota monax monax. The American Midland Naturalist, 53(2), 257-293. Gunderson, A.M., Jacobsen, B.K., Olson, L.E., 2009. Revised distribution of the Alaska Marmot, Marmota broweri, and confirmation of parapatry with hoary marmots. Journal of Mammalogy, 90(4), 859-869. Guthery, F.S., Brennan, L.A., Peterson, M.J., Lusk, J.J., 2005. Information theory in wildlife science: critique and viewpoint. Journal of Wildlife Management, 69(2), 457-465. Hanssen, A.W., Kuipers, W.J.A., 1965. On the relationship between the frequency of rain and various meteorological parameters. Meded. Verh., 81, 2-15. Heard, D., 1977. The behavior of Vancouver Island marmots (Marmota vancouverensis). University of British Columbia, Vancouver, BC.  62  Hebda, R.J., McDadi, O., Mazzucchi, D., 2004. Habitat history and the decline of the Vancouver Island marmot (Marmota vancouverensis), in: Hooper, T.D. (Ed.), Proceedings of the Species at Risk 2004 Pathways to Recovery Conference. BC Species at Risk 2004 Pathways to Recovery Conference Organizing Committee, Victoria, BC. pp. 1-11. Heldmaier, G., Ortmann, S. , Elvert, R., 2004. Natural hypometabolism during hibernation and daily torpor in mammals. Respiratory Physiology & Neurobiology, 141(3), 317-329. Holmes, W.G., 1984. Predation risk and foraging behavior of the hoary marmot in Alaska. Behavioral Ecology and Sociobiology, 15(4), 293-301. Inouye, D.W., Barr, B., Armitage, K.B., Inouye, B.D., 2000. Climate change is affecting altitudinal migrants and hibernating species. Proceedings of the National Academy of Sciences of the United States of America, 97(4), 1630-1633. Island Timberlands Limited Partnership, 2011. [Cutblock distribution and extent near Vancouver Island marmot release sites]. Unpublished raw data. IUCN, 1998. Guidelines for Re-introductions. Prepared by the IUCN/SSC Re-introduction Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. IUCN, 2011. The IUCN Red List of threatened species. Version 2011.2. Retrieved April 15, 2012, from: <http://www.iucnredlist.org> Janz, D., Blumensaat, C., Dawe, N.K., Harper, B., Leigh-Spencer, S., Munro, W.T., Nagorsen, D., 1994. National Recovery Plan for the Vancouver Island Marmot. Report No. 10. Recovery of Nationally Endangered Wildlife Committee, Ottawa, ON. Janz, D.W., Bryant, A.A., Dawe, N.K., Schwantje, H., Harper, B., Nagorsen, D., Doyle, D., deLaronde, M., Fraser, D., Lindsay, D., Leigh-Spencer, S., McLaughlin, R., Simmons, R., 2000. National Recovery Plan for the Vancouver Island Marmot (2000 Update). Report No. 19. Recovery of Nationally Endangered Wildlife, Ottawa, ON. Johns, D.W., Armitage, K.B., 1979. Behavioral ecology of alpine yellow-bellied marmots. Behavioral Ecology and Sociobiology, 5(2), 133-157. Jefferies, D.J., Wayre, P., Jessop, R.M., Mitchell-Jones, A.J., 1986. Reinforcing the native Otter Lutra lutra population in East Anglia: an analysis of the behavior and range developments of the first release group. Mammal Review, 16(2), 65-79.  63  Keeley, T., Goodrowe, K.L., Graham, L., Howell, C., MacDonald, S.E., 2011. The reproductive endocrinology and behavior of Vancouver Island marmot (Marmota vancouverensis). Zoo Biology, 29, 1-16. Kleiman, D.G., 1989. Reintroduction of captive mammals for conservation. BioScience, 39(3), 152-161. Kruckenhauser, L., Pinsker, W., Haring, E., Arnold, W., 1999. Marmot phylogeny revisited: molecular evidence for a diphyletic origin of sociality. Journal of Zoological Systematics and Evolutionary Research, 37(1), 49-56. Kuhn, M., 2012. Package ‘caret’: Classification and Regression Training. R package version 5.15-023. Retrieved May 15, 2012, from: <http://caret.r-forge.r-project.org> Laroque, C.P., 1998. Tree invasion in subalpine Vancouver Island marmot meadows. BC Environmental Research Scholarship Committee, Victoria, BC. Laroque, C.P., Lewis, D.H., and Smith, D.J., 2000. Treeline dynamics on southern Vancouver Island, British Columbia. Western Geography, 10/11(43-63). Lea, A.J., Blumstein, D.T., 2011. Age and sex influence marmot antipredator behavior during periods of heightened risk. Behavioral Ecology and Sociobiology, 65(8), 1525-1533. Legaarden, T.K., Walton, J.P., Plaizier, J.C., McBride, B.W., Valdes, E.V., 2001. Serum fatty acid analysis and digestibility study in the Vancouver Island marmot (Marmota vancouverensis) fed a captive diet supplemented with α-linolenic acid. Zoo Biology, 20(4), 251-259. Lenihan, C., Van Vuren, D., 1996. Costs and benefits of sociality in yellow-bellied marmots (Marmota flaviventris): do noncolonial females have lower fitness? Ethology Ecology & Evolution, 8(2), 177-189. Lichota, G.B., McAdie, M., Ross, P.S., 2004. Endangered Vancouver Island marmots (Marmota vancouverensis): sentinels of atmospherically delivered contaminants to British Columbia, Canada. Environmental Toxicology and Chemistry, 23(2), 402-407. Lukacs, P.M., Burnham, K.P., Anderson, D.R., 2010. Model selection bias and Freedman’s paradox. Annals of the Institute of Statistical Mathematics, 62(1), 117-125.  64  Mace, T.F., Shepard, C.D. 1981. Helminths of a Vancouver Island marmot, Marmota vancouverensis Swarth, 1911, with a description of Diandrya vancouverensis sp.nov. (Cestoda: Anoplocephalidae), Canadian Journal of Zoology, 59, 790-792. Manel, S., Williams, H.C., Ormerod, S., 2001. Evaluating presence–absence models in ecology: the need to account for prevalence. Journal of Applied Ecology, 38: 921–931. Marmot Recovery Foundation, 2007. Greig Ridge Marmot Transplant – Strathcona Park 2007. Accessed April 2012 from BC Hydro, Fish and Wildlife Compensation Program website: www.bchydro.com/bcrp/projects/docs/bridge_river/07.W.CBR.01.pdf Martell, A.M., Milko, R.J., 1986. Seasonal diets of Vancouver Island marmots. Canadian Field-Naturalist, 100(2), 241-245. McAdie, M., 2004. History and current status of the Vancouver Island marmot (Marmota vancouverensis) captive-breeding program, in: Hooper, T.D. (Ed.), Proceedings of the Species at Risk 2004 Pathways to Recovery Conference. PRESS??, Victoria, pp. 1-11. Milko, R.J. 1984. Vegetation and foraging ecology of the Vancouver Island marmot (Marmota vancouverensis). M.Sc. thesis, University of Victoria, Victoria, B.C. Moehrenschlager, A., Moehrenschlager, C., 2001. Census of swift fox (Vulpes velox) in Canada and Northern Montana: 2000-2001. Alberta Sustainable Resource Development, Fish and Wildlife Division, Alberta Species at Risk Report No. 24. Edmonton, AB. Montgomery, D.C., Peck, E.A., Vining, G.G. 2012. Introduction to linear regression analysis, John Wiley & Sons Inc., New Jersey, US. Munro, W.T., 1979. COSEWIC status report on the Vancouver Island marmot Marmota vancouverensis in Canada. Committee on the Status of Endangered Wildlife in Canada, Ottawa. 1-12 pp. Munro, W.T., Janz, D.W., Heinsalu, V., Smith, G.W., 1985. The Vancouver Island marmot: status and management plan. BC Ministry of Environment, Wildlife Branch, Wildlife Bulletin No. B-39. Victoria, BC. Murdoch, J.D., Munkhzul, T., Buyandelger, S., Reading, R.P., Sillero-Zubiri, C., 2009. The endangered Siberian marmot Marmota sibirica as a keystone species? Observations and implications of burrow use by corsac foxes Vulpes corsac in Mongolia. Oryx, 43, 431-434.  65  Nagorsen, D.W., 2004. Canada’s endemic mammals at risk: recent taxonomic advances and priorities for conservation. In: Hooper, T.D. (Ed.), Proceedings of the Species at Risk 2004 Pathways to Recovery Conference. BC Species at Risk 2004 Pathways to Recovery Conference Organizing Committee, Victoria, BC. pp. 1-2. Nagorsen, D.W., Keddie, G., Luszcz, T., 1996. Vancouver Island marmot bones from subalpine caves: archaeological and biological significance. Occasional Paper No. 4, Ministry of Environment, Lands and Parks, Victoria, BC. Neter, J., Wasserman, W., 1996 (fourth Ed.). Applied linear statistical models: regression, analysis of variance, and experimental designs. Irwin: Chicago, Illinois, US. Nunes, S., Ha, C.D.T., Garrett, P.J., Mueke, E.M., Smale, L., Holekamp, K.E., 1998. Body fat and time of year interact to mediate dispersal behavior in ground squirrels. Animal Behaviour, 55, 605-614. Ortmann, S., Heldmaier, G., 2000. Regulation of body temperature and energy requirements of hibernating Alpine marmots (Marmota marmota). American Journal of Physiology, 278, R698-R704. Ostermann, S.D., Deforge, J.R., Edge, W.D., 2001. Captive breeding and reintroduction evaluation criteria: a case study of peninsular bighorn sheep. Conservation Biology, 15(3), 749-760. Ozgul, A., Armitage, K.B., Blumstein, D.T., Oli, M.K., 2006. Spatiotemporal variation in survival rates: implications for population dynamics of yellow-bellied marmots. Ecology, 87(4), 1027-1037. Peak, R.G., 2007. Forest edges negatively affect golden-cheeked warbler nest survival. The Condor, 109(3), 628-637. Peduzzi, P., Concato, J., Kemper, E., Holford, T.R., Feinstein, A.R., 1996. A simulation study of the number of events per variable in logistic regression analysis. Journal of Clinical Epidemiology, 49(12), 1373-1379. Peduzzi, P., Concato, J., Feinstein, A.R., Holford, T.R. 1995. Importance of events per independent variable in proportional hazards regression analysis II. Accuracy and precision of regression estimates. Journal of Clinical Epidemiology, 48(12), 1503-1510.  66  Peignot, P., Charpentier, M.J.E., Bout, N., Bourry, O., Massima, U., Dosimont, O., Terramorsi, R., Wickings, E.J., 2008. Learning from the first release project of captive-bred mandrills Mandrillus sphinx in Gabon. Oryx, 42(1), 122-131. Peirce, C.S. 1884. The numerical measure of the success of predictions. Science, 4(93), 453454. Perrin, C., Coulon, J., Le Berre, M., 1993. Social behavior of alpine marmots (Marmota marmot): seasonal, group, and individual variability. Canadian Journal of Zoology, 71, 19451953. Polly, D.P., 2003. Paleophylogeography: the tempo of geographic differentiation in marmots (Marmota). Journal of Mammalogy, 84(2), 369-384. Price, M.R., Soorae, P.S., 2003. Reintroductions: whence and whither? International Zoo Yearbook, 38, 61-75. Province of British Columbia, 2011. GeoBC Base Mapping and Cadastre Branch. Retrieved July 13, 2011, from: <http://geobc.gov.bc.ca/Imagery.html> R Development Core Team, 2005. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Switzerland. Retrieved September 30, 2011, from: <http://www.r-project.org> Robertson, C.P.J., Harris, S., 1995. The behavior after release of captive-reared fox cubs. Animal Welfare, 4(4), 295-306. Ruf, T., Arnold, W., 2000. Mechanisms of social thermoregulation in hibernating alpine marmots (Marmota marmota), in Heldmaier, G., Klingenspor, M. (Eds.), Life in the cold. Springer Verlag, Heidelberg, pp.81-94. Sarrazin, F., Bagnolini, C., Pinna, J.L., Danchin, E., Clobert, J., 1994. High survival estimates of griffon vultures (Gyps fulvus fulvus) in a reintroduced population. The Auk, 111(4), 853-862. Scott, B.M.V., Shackleton, D.M., 1980. Food habits of two Vancouver Island wolf packs: a preliminary study. Canadian Journal of Zoology, 58(6), 1203-1207. Seddon, P.J., 1999. Persistence without intervention: assessing success in wildlife reintroductions. Trends in Ecology & Evolution, 14(12), 503.  67  Seddon, P.J., Armstrong, D.P., Maloney, R.F., 2007. Developing the science of reintroduction biology, 21(2), 303-312. Sheriff, M.J., Williams, C.T., Kenagy, G.J., Buck, C.L., and Barnes, B.M., 2012. Thermoregulatory changes anticipate hibernation onset by 45 days: data from free-living arctic ground squirrels. Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology. Pre-released online, retrieved July 9, 2012, from: <http://springerlink.com> Snyder, N.F.R., Derrickson, S.R., Beissinger, S.R., Wiley, J.W., Smith, T.B., Toone, W.D., Miller, B., 1996. Limitations of captive breeding in endangered species recovery. Conservation Biology, 10(2), 338-348. Stephens, P.A., Frey-Roos, F., Arnold, W., Sutherland, W.J., 2002. Model complexity and population predictions. The alpine marmot as a case study. Journal of Animal Ecology, 71, 343-361. Steppan, S.J., Akhverdyan, M.R., Lyapunova, E.A., Fraser, D.G., Vorontsov, N.N., Hoffmann, R.S., Braun, M.J. 1999. Molecular phylogeny of the marmots (Rodentia: Sciuridae): tests of evolutionary and biogeographic hypotheses. Systematic Biology, 48(4), 715-734. Svendsen, G.E., 1974. Behavioral and environmental factors in the spatial distribution and population dynamics of a yellow‐bellied marmot population. Ecology, 55, 760-771. Swarth, H.A., 1911. Two new species of marmots from British Columbia. University of California Publications in Zoology, 7(6), 201-204. Symonds, M.R.E., Moussalli, A., 2011. A brief guide to model selection, multimodel inference and model averaging in behavioural ecology using Akaike’s information criterion. Behavioural Ecology and Sociobiology, 65(1), 13-21. Taulman, J.F., 1990. Late summer activity patterns in hoary marmots. Northwestern Naturalist, 71(2), 21-26. Tear, T.H., Scott, J.M., Hayward, P.H., Griffith, B., 1993. Status and prospects for success of the Endangered Species Act: A look at recovery plans. Science, 262(5136), 976-977.  68  Teixeira, C.P., de Azevedo, C.S., Mendl, M., Cipreste, C.F., Young, R.J., 2007. Revisiting translocation and reintroduction programmes: the importance of considering stress. Animal Behaviour, 73(1), 1-13. Thornback, J., Jenkins, M., 1982. The IUCN Mammal Red Data Book: Threatened mammalian taxa of the Americas and the Australasian zoogeographic region (excluding Cetacea). IUCN, Gland, Switzerland. pp. 225-227. TimberWest Forest Corporation, 2011. [Cutblock distribution and extent near Vancouver Island marmot release sites]. Unpublished raw data. Tweed, E.J., Foster, J.T., Woodworth, B.L., Oesterle, P., Kuehler, C., Lieberman, A.A., Powers, A.T., Whitaker, K., Monahan, W.B., Kellerman, J., Telfer, T., 2003. Survival, dispersal, and home-range establishment of reintroduced captive-bred puaiohi, Myadestes palmeri. Biological Conservation, 111(1), 1-9. Van Vuren, D., Armitage, K.B., 1994. Survival of dispersing and philopatric yellow-bellied marmots: what is the cost of dispersal? Oikos, 69(2), 179-181. Vancouver Island Marmot Recovery Team, 2008. Recovery strategy for the Vancouver Island marmot (Marmota vancouverensis) in British Columbia. Prepared for the B.C. Ministry of Environment, Victoria, BC. Vittinghoff, E., McCulloch, C.E., 2007. Relaxing the rule of ten events per variable in logistic and Cox regression. American Journal of Epidemiology, 165(6), 710-718. Waters, S.S., 2010. Swift fox Vulpes velox reintroductions: a review of release protocols. International Zoo Yearbook, 44(1), 173-182. Western Forest Products Incorporated, 2011. [Cutblock distribution and extent near Vancouver Island marmot release sites]. Unpublished raw data. Wilson, K.A., Westphal, M.I., Possingham, H.P., Elith, J., 2005. Sensitivity of conservation planning to different approaches to using predicted species distribution data. Biological Conservation, 122(1), 99-112. Wolf, C.M., Griffith, B., Reed, C., Temple, S.A., 1996. Avian and mammalian translocations: update and reanalysis of 1987 survey data. Conservation Biology, 10(4), 1142-1154.  69  Youden, W.J., 1950. Index for rating diagnostic tests. Cancer, 3(1), 32-35.  70  Appendices Appendix A - Representation of newly released captive-bred Vancouver Island marmots across categorical predictors for site fidelity, active season and overwinter survival.  71  Appendix B - Relative importance, parameter estimates, unconditional standard errors, and 95% unconditional confidence intervals for all variables used to predict site fidelity, active season survival, and overwinter survival. Predictors with large standard errors (> parameter estimates) have little reliability.  72  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0072983/manifest

Comment

Related Items