Initial movements of re-introduced elk in the Missouri Ozarks.
Although animal re-introductions have been commonplace in the United States since the turn of the 20th century (Enochs, 1998), most attempts to restore wild populations fail (Rodriguez et al, 1995; Clark et al., 2002; Teixeira et al., 2007; Jachowski et al, 2011; Parlato and Armstrong, 2013). As restoration efforts are expensive and time intensive (Spinola et al, 2008), managers must optimize protocols to maximize success. Re-introduced wildlife face many challenges post release, including predation, habitat and food shortages, disease, and human conflicts (Collazo et al., 2003; Vandel et al, 2006; Benson and Chamberlain, 2007; Jachowski et al., 2011; Moseby et al., 2014). Elk (Ceruus elaphus) were re-introduced to the Missouri Ozarks in 2011-2013 after being extirpated from the state in 1865 (Missouri Department of Conservation, 2001). Elk restorations in eastern North America face similar challenges to other wildlife re-introductions. Early attempts to restore elk often failed due to habitat constraints, legal or illegal harvest, agrarian conflicts, vehicle collisions, predation, or meningeal worm (Parelaphostrongylus tenuis) infections (Witmer, 1990; Virginia Department of Game and Inland Fisheries, 2010; Popp et al., 2014). The challenges to translocated wildlife, potential lack of available animals, and cost of re-introduction programs call for a complete understanding of factors contributing to restoration success.
Initial movements of re-introduced animals may determine population growth and therefore restoration success. Dispersal from the release site decreases population growth because animals may die or fail to reproduce (Hardman and Moro, 2006; Devineau et al., 2010; Parlato and Armstrong, 2013). Small populations must remain at a density that provides resiliency against stochastic events, protection against inbreeding, and prevents temporary Allee effects (Komers and Curman, 2000; Larkin et al., 2002; Williams et al., 2002; Armstrong and Wittmer, 2011; Parlato and Armstrong, 2013). However, large mammals often leave release sites quickly and are capable of making extensive movements (Fritts et al., 1984; Ruth et al., 1998; Moehrenschlager and Macdonald, 2003; Spinola et al., 2008; Yott et al., 2011). Elk and other large mammals that disperse long distances often experience high mortality rates (Ruth et al, 1998; Benson and Chamberlain, 2007; Haydon et al., 2008; Devineau et al., 2010; Scillitani et al., 2013). Because adult survival is key to the population growth of long-lived species (Nelson and Peek, 1982; Raithel et al., 2007; Devineau et al., 2010), large mammal restorations should be structured to encourage animals to remain near the release site.
Determining which factors affect movement ecology is important because encouraging reintroduced animals to stay near the release site is key for restoration success (Yott et al., 2011). Sex, age, presence of young, release site, and release year may all affect post-release movements in translocated mammals. Typically, males disperse farther than females in mammal re-introductions (Moehrenschlager and Macdonald, 2003; Hardman and Moro, 2006; Spinola et al., 2008; Devineau et al., 2010; Ryckman et al., 2010), although sex is not always an influential variable for elk (Larkin et al., 2004). Age is also an important factor in mammalian post-translocation movements, with older individuals usually traveling farther than younger ones (Fritts et al., 1984; Ruth et al.. 1998; Clark et al., 2002; Larkin et al., 2004; Ryckman et al., 2010). Maternal females may restrict movements post release compared to individuals without dependent offspring (Clark et al., 2002). This may be especially true for maternal elk and other ungulates that reduce movement post-parturition to tend hiding young (Vore and Schmidt, 2001). Release site and release year can also affect post-release movement and spatial ecology in elk and other large mammals (Larkin et al, 2004; Benson and Chamberlain, 2007; Devineau et al., 2010).
As elk re-introductions to eastern North America become more widespread, site-specific studies of factors influencing initial movements are needed to maximize success. Elk reintroduction programs have seen renewed interest the past two decades (Virginia Department of Game and Inland Fisheries, 2010). Of the eastern states and provinces that have implemented elk restoration efforts, seven have released elk in the past 20 y (Kentucky, Missouri, North Carolina, Ontario, Tennessee, Virginia, and Wisconsin) (O'Gara and Dundas, 2002; Rosatte et al., 2007; Virginia Department of Game and Inland Fisheries, 2010; Missouri Department of Conservation, 2010). Understanding factors influencing elk movements post release would allow managers to optimize protocols for initial population growth (Komers and Curman, 2000), select areas with an appropriate landscape structure (Larkin et al., 2004; Ryckman et al, 2010), provide food sources (Ryckman et al., 2010), mitigate human disturbance on the site (Larkin et al, 2004), and attempt to reduce elkhuman conflict in agrarian areas (Rosatte et al., 2007) to increase restoration success.
Our objectives were to study the initial movements of elk re-introduced to Missouri by assessing range size and two site fidelity metrics, maximum distances moved and range shifts. Because disorientation subsides and translocated mammals establish long-term space use habits 5-156 d post release (Ruth et al., 1998; Moehrenschlager and Macdonald, 2003; Clapp et al., 2014), we evaluated movement responses for the first 6 mo (183 d) post release. We hypothesized site fidelity would be high because elk were released in suitable habitat and human disturbance in the area was minimized immediately post release (Missouri Department of Conservation, 2000; Larkin et al., 2004). In addition, the Missouri elk restoration held animals in pens to form social bonds and acclimate to their new environment after transport for a soft release, which is thought to increase site fidelity (Rosatte et al, 2007; Missouri Department of Conservation, 2010; Ryckman et al., 2010; Armstrong and Wittmer, 2011; Mcintosh et al., 2014). Because transient or exploring individuals are thought to traverse larger areas than resident individuals (Turchin, 1998), and we expected elk to settle quickly due to release site conditions, we hypothesized range sizes would decrease over time. We also predicted initial movement responses would vary by sex, age, maternal status, release site, and release year.
Elk were re-introduced to an 896 [km.sup.2] Elk Restoration Zone (ERZ, Fig. 1) in parts of Carter, Shannon, and Reynolds Counties in the southeastern Missouri Ozarks (91[degrees]24' to 90[degrees]58'W and 37[degrees]0' to 37[degrees]19'N). The Missouri Department of Conservation (MDC), National Park Service, and United States Forest Service managed 49% of the ERZ. The Nature Conservancy owned 3% of the ERZ, and an additional 27% was held by the L-A-D Foundation, a sustainable forest products initiative (Missouri Department of Conservation, 2010). Combined, 79% of the ERZ was publicly accessible for recreation and hunting.
The ERZ was comprised largely of forest and woodland habitat (93%, Missouri Department of Conservation, 2010). Oak (Quercus alba, Q. stellata, (J. coccinea, Q. velutina), hickory (Caiya tomentosa, C. texana, C. glabra), and shortleaf pine (Pinus echinata) dominated the mature, second-growth forests of the region (Brookshire and Dey, 2000, Shiflev et al., 2000). Sparse open lands (5%) were held mostly in pasturelands and food plots, rather than cropland (0.1%). MDC had restored and maintained glades and woodland understory on their ERZ lands through landscape-level prescribed burns (Missouri Department of Conservation, 2005; Missouri Department of Conservation, 2010; Tousignant, 2011). Other large mammals occurring in the region included white-tailed deer (Odocoileus virginianus), black bears (Ursus americanus), coyotes (Canis latrans), bobcats (Lynxrufus), and transient cougars (Puma concolor). Climate in the ERZ was typically mild. During the study period, there was limited depth and duration of snow, especially during winter 2011-2012. However, the region experienced a drought and excessive heat during the summer of 2012. The average annual temperature for the area was 55.7[degrees]F (U.S. Climate Data, 2017). The annual average high temperature was 70.4[degrees]F; the average annual low temperature was 41.4[degrees]F (U.S. Climate Data, 2017). The vegetative season (temperatures > 50[degrees]F) was from May to September (U.S. Climate Data, 2017). Average rainfall in the study area was 47.57 inches (U.S. Climate Data, 2017).
Elk were released on MDC's Peck Ranch Conservation Area (Peck Ranch, 2011-2013) and The Nature Conservancy's Chilton Creek Preserve (Chilton Creek, 2012) within the ERZ (Fig. 1). Although the release sites were approximately 7.6 km apart, the Chilton Creek release site was only 2.5 km from the Peck Ranch border. The 9327 ha Peck Ranch was made up of 8445 forested ha, 662 ha of glades, savannah, old fields, and wetlands, and 221 ha of food plots (Missouri Department of Conservation, 2009). Ridge elevation ranged between 274.3 to 304.8 m above sea level, but the maximum elevation on Peck Ranch was Stegall Mountain at 410.9 m above sea level (Missouri Department of Conservation 2009). The southern boundary of Peck Ranch aligned with part of the southern boundary of the Elk Restoration Zone (Missouri Department of Conservation, 2010). The almost entirely forested 2277 ha Chilton Creek bordered Peck Ranch and the Current River (The Nature Conservancy, 1997; Fig. 1).
We captured elk in January 2011, 2012, and 2013 using corral traps or darting from the ground near Stoney Fork, Kentucky. Darted elk were immobilized using carfentanil citrate; the immobilization drug was reversed with naltrexone. We fitted all elk with GPS-PPT/VHF collars with blow-off devices (RASSL 3D cell collar. North Star Science and Technology, LLC, King George, Virginia; G2110E Iridium/GPS series model, Advanced Telemetry Systems, Insanti, Minnesota). We held elk in Kentucky 102-116 d (2011), 105-129 d (2012), and 107-128 d (2013) to complete health testing. We transported elk to Missouri overnight on a commercial stock trailer and held them for 27 (2011), 32-34 d (2012), and 19 d (2013) at Peck Ranch before release. Holding periods varied due to disease testing logistics and interagency release approvals. In 2011 we released all animals (n = 34, 15 adult females, six adult males, five yearling females, eight yearling males) at Peck Ranch on 1 June (Fig. 1). In 2012, we released all but one gestating and post-parturient female on Peck Ranch on 19 June (n = 19 adult females); we transported remaining animals (n = 14, three adult females, four adult males, three yearling females, four yearling males) to Chilton Creek on 21 June and held them there until release on 23 June. In 2013 we released all animals (n = 39, 20 adult females, 16 yearling females, three yearling males) at Peck Ranch on 7 June. We captured juveniles and fitted them with expandable VHF collars (custom M4200 series model, Advanced Telemetry Systems, Isanti, Minnesota) to monitor survival. Elk locations were collected using the GPS-PTT collars at 2.5 (n = 3), 3 (n = 41), 4 (n = 2), and 5 (n = 46) h intervals. All elk use was approved by University of Missouri (Animal Care and Use Committee protocol 6909).
We characterized movement patterns over the initial 6 mo (183 d) post release for animals surviving that time. Elk that died during the study period were necropsied and tissue samples were sent to the University of Georgia Southeast Cooperative Wildlife Disease Study or the University of Missouri Veterinary Medicine Diagnostic Laboratory to determine presence of disease. Elk that did not survive often displayed lowered movements due to neurologic disease (Parelaphostrongylus tenuis, Carpenter et al., 1973) and were excluded from the analysis to avoid biasing results. To capture the effect of translocation on animal movements consistently between release groups with different release dates, we evaluated time frames based on days post release rather than biological seasons. In 2011 an apparent change in behavior occurred at approximately 10 d post release; nonmaternal elk returned to the release site following initial long-range movements and movement rates decreased (Bleisch, 2014). For this reason we hypothesized movement patterns 0-10 d post release may differ from other periods. We based other time frames on the literature to facilitate comparisons between the Missouri re-introduction and re-introductions elsewhere (Larkin et al., 2004; Yott et al., 2011). Therefore, we evaluated responses for each elk 0-10 d, 11-30 d, 31-61 d, and 62-183 d post release. For brevity hereafter we use the term elk-period to denote one time frame for one elk.
We calculated maximum distances traveled, range size, and range shifts for each elk-period. We calculated maximum distances traveled from the release site per elk-period using the Near function in ArcInfo 10.2 (ESRI, Redlands, California). To assess range size and shifts, we created fixed kernel utilization distributions (UDs) for each elk-period (Van Winkle, 1975; Seaman et al, 1999) in program R using packages KernSmooth, ks, and mvtnorm (R Version 2.10.0, www.r-project.org, Accessed 26 Jan. 2012). We selected bandwidths using plug-in methods (Gitzen et al., 2006). We evaluated range shifts by comparing UD volume overlap for all within-individual elk-periods (UDs from each period compared to UDs from each other period for each individual, e.g., 0-10 vs. 11-30 d, 0-10 vs. 31-61 d, 0-10 vs. 62-183 d) using Seidel's (1992) volume of intersection index (Millspaugh et al., 2004), which yields a number ranging from zero to one indicating the proportion of overlap in the volume of UD pairs. We used 100% volume UDs to calculate the Volume of Intersection Index in order to capture peripheral range overlap. We report 95% volume UDs to assess range size.
We developed 18 candidate models representing different hypotheses explaining reintroduced elk movement patterns (Bleisch, 2014). We constructed models using the biological variables sex (Ryckman et al., 2010), age (Larkin et al., 2004; Ryckman et al, 2010), and maternal status, as well as release variables release site and release year. We considered females to be maternal for the elk-periods in which their captured young were born and survived. Because females altered their movement patterns following the death of their offspring, we considered females nonmaternal for the elk-period during which a juvenile died. We noted elk ages as yearling or adult. One candidate model was a global model that assumed all five variables plus time post release contributed to movement patterns. We also included univariate models to determine if a single factor drives movement patterns. Because we expected time post release to be an important contributor to elk movement patterns, we included the variable in all subsequent models. Additional models tested the hypotheses that movement patterns were driven by release factors or biological factors. After plotting the data, we included four multivariate models that combined release and biological factors. We fit each of the 18 models to each of the three movement response variables.
We used an information theoretic approach (Burnham and Anderson, 2002) to rank candidate models. We fit repeated measures, mixed-effects models for each movement response using the MIXED procedure in SAS 9.3 (SAS Institute, Inc., Cary, North Carolina; Littell et al., 2006). We used Akaike's Information Criterion corrected for small sample sizes ([AIC.sub.C]) to evaluate support for each model (Burnham and Anderson, 2002). Based on [AIC.sub.C] values from restricted maximum likelihood global models of each response variable, we used an autoregressive covariance structure for range size models and an unstructured covariance structure for maximum distance moved and range shift models.
We used a two-step process to find parameter estimates (see Bonnot et al., 2011). First, we fit candidate models using maximum likelihood methods to achieve [AIC.sub.C] values comparable across models with varying fixed effects. Models that had at least one-eighth of the relative support of the top model were included in the confidence set for each response (Burnham and Anderson, 2002). Second, we refit models using restricted maximum likelihood to achieve unbiased estimates and standard errors for the fixed effects. When there was at least one model within one-eighth of the top model, we model-averaged using Akaike weights (Burnham and Anderson, 2002; Bonnot et al., 2008). Confidence intervals were constructed from the unconditional variance estimates.
We collected 69,209 locations on 32 surviving elk in 2011 (n = 26,715), 21 surviving elk in 2012 (n = 15,621), and 31 elk surviving in 2013 (n = 26,873). Typical causes of death included stress-related bacterial and viral infections, especially during the 2012 drought, as well as secondary conditions caused by neurologic disease (Parelaphostrongylus tenuis). We captured four juveniles in 2011, 12 juveniles in 2012, and 10 juveniles in 2013 from females surviving the initial 6 mo post release. We conducted a GPS collar error analysis following the methods of Sager-Fradkin et al. (2006). Collar location error averaged 5.99 meters (range: 3.00-11.35m). Collars functioned properly through 6 mo post release.
Elk displayed distinctly different initial movement behavior in each release (Fig. 1C). Immediately following the 2011 release, we observed several nonmaternal animals traveling southwest together approximately 9.8 km from the release site; all animals returned to Peck Ranch within 4 d. By contrast maternal elk remained near the release site. After the release in Peck Ranch in 2012, elk did not make extensive initial movements; only two individuals left Peck Ranch within the first 10 d post release for this group. Animals in the 2012 Chilton Creek release group divided into subgroups and solitary individuals, each of which traveled in a separate direction. Three small groups of elk released at Chilton Creek traveled 2.6, 4.1, and 8.2 km from the release site within 10 d post release. Within 6 mo all surviving elk released at Chilton Creek joined other elk on Peck Ranch. In 2013 elk left the release site more gradually than in in previous releases. One group of eight nonmaternal elk traveled beyond the Peck Ranch border 400 m to the east one day after release but returned to Peck Ranch the same day. No other elk released in 2013 left Peck Ranch during the first 10 d post release.
MAXIMUM DISTANCE MOVED
Re-introduced elk stayed near the release site through 6 mo post release. The maximum distance moved from the release site was less than 10 km 62-183 d (6 mo) post release for 94% of animals released in 2011. Although this number dropped to 57% for elk released in 2012, this is likely because animals released at Chilton Creek in 2012 either were still making extensive movements during this time frame or joined other elk at Peck Ranch and instead demonstrated fidelity to (hat site. No elk released in 2012 at Peck Ranch moved more than 10 km from the release site within 6 mo post release but all elk released at Chilton Creek did. In 2013 97% of elk remained within 10 km during 62-183 d post release. The mean maximum displacement for each animal in any time frame was 6.1 [+ or -] 4.0 km in 2011 (n = 32, range = 0.4-11.8 km), 10.4 [+ or -] 12.2 km in 2012 (n = 21, range = 0.6-37.0 km), and 8.0 [+ or -] 10.6 km in 2013 (n = 31, range = 3.7-31.2 km). Females with young stayed closer to the release site immediately following release than nonmaternal elk. Nonmaternal elk were twice as far away from the release site than maternal elk for the first 10 d, and daily average displacement was first equal between maternal and nonmaternal elk at 21 d post release (Fig. 2).
Individual maximum distances from the release site were stable for the first three elk-periods and greater for the final elk-period (Fig. 3). Compared to 0-10 d post release, maximum displacement was 22% greater for days 11-31, 14% greater for days 32-61, and 100% greater for days 62-183. Supported models contained all five covariates (Table 1). Confidence intervals for estimated results based on age, maternal status, and year overlapped. However, females were estimated to be 1.1 km closer to the release site than males in each elk-period, and animals released al Peck Ranch were estimated to achieve 3.8 km lower maximum distances per elk-period than those released at Chilton Creek.
Estimated elk range sizes were stable for the first two periods and increased over the last three periods; range sizes for 62-183 d [lost release were roughly nine limes larger than first period (Fig. 4). Average range sizes for 62-183 d post release were 17.6 [+ or -] 5.4 [km.sup.2] for elk released in 2011 (n = 32), 25.2 [+ or -] 11.0 [km.sup.2] for elk released at Peck Ranch in 2012 (n = 12), 59.5 [+ or -] 60.4 [km.sup.2] for elk released at Chilton Creek in 2012 (n = 9), and 13.9 [+ or -] 17.2 for elk released in 2013 (n = 31). All predictor variables were in the confidence set (Table 2). Site emerged as an important determinant of range size due to movements by elk released at Chilton Creek before finding Peck Ranch. Compared to estimated ranges for elk released at Chilton Creek during the first two elk-periods (20.6 [+ or -] 2.2 [km.sup.2] and 19.6 [+ or -] 2.2 [km.sup.2]'), estimated ranges for elk released at Peck Ranch were small (2.6 [+ or -] 2.2 [km.sup.2] and 1.7 [+ or -] 2.2 [km.sup.2]).
Elk used previously explored range over sequential time periods during the first 6 mo post release. The mean range overlap for successive time periods for elk released in 2011 was 0.230 [+ or -] 0.320 (n = 32), the mean for elk released in 2012 was 0.290 [+ or -] 0.323 (n = 21), and the mean for elk released in 2013 was 0.26 [+ or -] 0.24 (n = 31). Just 9 of 252 (4%) consecutive elk-period UD pairs yielded no range overlap. Range overlap increased for each successive pair of elk-periods, with the lowest range overlap between UD pairs occurring when the first elk-period (days 0-10) was contrasted with die last (days 62-183) (Fig. 5). There was no uncertainty as to which model fit best, and the top model was die global model ([w.sub.i] = 0.944). Maternal status, release site, and release year had the greatest effect. Nonmaternal elk had 34% more range overlap from 0-10 to 11-31 d post release than maternal elk. Elk released at Peck Ranch demonstrated 0.176 more range overlap in each elk-period than those released at Chilton Creek. Elk released in 2011 had 0.18 less range overlap in each elk-period than those released in 2012 and 0.116 less range overlap in each elk-period than animals released in 2013. Due to the initial movements off Peck Ranch, there was virtually no range overlap between the first and last elk-periods for elk released in 2011. Higher estimated range overlap for elk released in 2012 may reflect the high fidelity exhibited by elk to the release site from the start.
Missouri elk exhibited a multiphasic movement strategy post release and quickly settled in the area. Elk appeared to acclimate to their environment in discrete phases, including: (1) immediate departure from the release site and elevated movement rates, followed by (2) establishing a home range and gradually expanding it incorporating previously used area. Multiphasic movement patterns have previously been observed in translocated raccoons (Procyon lotor), swift foxes (Vulpes velox), and cougars (Puma concolor) (Ruth et al., 1998; Mosillo et al., 1999; Moehrenschlager and Macdonald, 2003). Whereas previous studies found re-introduced elk took 3 y to transition from a dispersive to a home range phase (Fryxell et al, 2008), Missouri elk acclimated to their surroundings much faster. Steady displacement levels and range sizes may indicate that elk were transient for just 10 d before settling into a home range phase. After 10 d, elk established a range and expanded it over time, as evidenced by the fact that range sizes and maximum displacement were largest for days 62-183, but range overlap steadily increased over the study period.
Release site was an important factor contributing to movement patterns in re-introduced elk. Elk released at Chilton Creek moved farther and had larger ranges, indicating they exhibited transient behavior longer than elk released at Peck Ranch (Turchin, 1998). The movement responses of elk released at Chilton Creek may be due to differences in habitat, absence of conspecifics, demography of the release-group, or release strategy. Previous studies have suggested re-introduced elk demonstrate higher release site fidelity when release areas have greater edge densities (Larkin et al., 2004) and forage-rich open land habitats (Ryckman et al., 2010). Our results were consistent with these findings. Peck Ranch contains hundreds of open areas and associated edge habitat; Chilton Creek lacks both. Further, previous studies have indicated if a re-introduced species has a high affinity to locally rare habitat patches, dispersal will be limited (Attum et al., 2013). Isolated open patches on Peck Ranch and limited connectivity to similar habitats may have contributed to the establishment of elk on Peck Ranch (Parlato and Armstrong, 2013; Moscby et al., 2014). The presence of conspecifics has also been thought to be a potential factor in release site fidelity and post-release acclimation of gregarious ungulates (Dolev et al, 2002; Ryckman et al., 2010; Scillitani et al., 2013; Clapp et al., 2014). Our study corroborates these claims. Two groups were released at sites without elk in residence; both immediately left. One group did not find conspecifics and returned to the release site (Peck Ranch, 2011). The other found conspecifics, integrated with them, and did not return to the release site (Chilton Creek, 2012). Most elk released in the presence of other elk never left (Peck Ranch, 2012, 2013). Alternatively, elk released at Chilton Creek may have demonstrated different behavior due to demographic characteristics of the group. The release group was predominantly young males, who may disperse long distances even without translocation (Ryckman et al., 2010). Only two females released at Chilton Creek survived for 6 mo post release. Because elk herds are structured by dominant females (Edge et al., 1986), a lack of mature females at the release site may have contributed to the immediate disbanding of the Chilton Creek elk release group into small groups and lone individuals. Re-introduced elk travel farther when solitary (Haydon et al., 2008), which may have exacerbated differences between release sites. Finally, elk relocated to Chilton Creek were held together for just two more days after transport. This process may have interrupted social bonds and led to diffusion of individuals across the landscape. Additionally, these elk may have habituated to Peck Ranch and attempted to return to the site to which they had become acclimated.
Our study provides additional evidence that large mammals are more likely to remain at the release site if they are transported while pregnant and released before bearing young (Clark et al., 2002). In some species this effect is presumably caused by lower mobility of young offspring; in elk, this may be due to hiding behavior of juveniles. Although the impacts of maternal status were not strongly supported, its effect may have been understated due to our inability to capture more juveniles. Managers should weigh the benefits and drawbacks of translocating gestating wildlife for re-introductions. Re-introduction programs are inherently stressful for wildlife (Teixeira et al., 2007; Dickens et al., 2010), and stress could adversely affect both mother and offspring, particularly late-term (Shelton and Huston, 1968; Braastad, 1998; Couret et al., 2009). We observed one possible case of abandonment each release year. In 2011 and 2013, the juvenile was never nursed; in 2012 a hiding juvenile was left in the holding pens post release. All three juveniles subsequently died. Holding female elk together to the date of parturition also carries the potential for mismatching mother/offspring pairs. Because juveniles will nurse from any female, older juveniles can take milk from multiple dams, and dominant females that have lost a juvenile may adopt another female's young (Hudson et al., 2002). We observed both situations in our holding pens pre-release. Finally, because establishing a long-term range typically takes 5-156 d post release (Ruth et al, 1998; Moehrenschlager and Macdonald, 2003; Clapp et al, 2014), preparturient female elk may not have adequate time to find suitable parturition habitat, which may lead to lower juvenile survival. Therefore, managers must weigh the benefits of increased site fidelity against the possible negative effects on females and their offspring.
Although previous studies found sex and age to be important influences on re-introduced elk movements (Larkin et al, 2004; Ryckman et al., 2010), our study only found sex to contribute to movement patterns. Other studies have shown that the life history of translocated animals may affect their movement ecology post release (Moseby et al, 2014). For instance females may travel farther from the release site in species with female-biased dispersal (Kemink and Kesler, 2012). Life history may have contributed to movement patterns in elk released to Missouri. In our study effects of age were supported but were never measurably different over time. However, sex contributed to maximum distances moved. These results may indicate that elk demonstrated typical social behavior within 6 mo post release. Male elk were observed forming bachelor herds, which do not vary by age. Also, it is normal for young male elk to disperse, and most translocated male elk were yearlings at the time of release. Further, familiarity with other individuals at the release site may discourage dispersal in social species (Moseby et al., 2014). Same-sex bonds may have been encouraged by Missouri's release strategy. During the holding period at the release site, elk were separated into same-sex pens. This soft release strategy may have helped to build stronger social bonds between same-sex individuals and encouraged them to aggregate post release.
Acknowledgments.--Funding for this research was provided by the Federal Aid in Wildlife Restoration Program. We thank the Rocky Mountain Elk Foundation, particularly T. Toman, for their enthusiastic support of elk research and management in Missouri. We thank the Missouri Department of Conservation for their elk restoration efforts and collaboration with this research. R. Dent, S. McWilliams, J. Rieken, and S. Crider helped coordinate the capture, care, and transport of elk for release. Peck Ranch Conservation area staff R. Houf, P. Mabry, M. Price, and P. Vessels provided logistical support and area access. Thank you to T. Bixler, J. Stunners, J. Sartwell, J. Fleming, and P. Marley for database management. Thank you to the Kentucky Department of Fish and Game for providing elk, and the University of Kentucky, especially A. Hildreth, for sharing elk capture and holding data. We acknowledge the National Park Service employees at the Ozark National Scenic Riverways for their cooperation with the elk restoration and research. Thank you to technicians S. Snow, T. Wolf, T. Schrautemeier, C. Sebright, J. Dillon, J. Ashling. D. Neel, S. Raiman, J. Foggia, J. Behrens, M. Thomas, C. Wright, D. Payne, D. Haan, N. Oakley, K. Stonehouse, J. Leonard, D. Russell, and S. G. McKee for help with field work.
Armstrong, D. and H. Wittmer. 2011. Incorporating Alice effects into reintroduction strategies. Ecol. Res., 26:687-695.
Attum, O., C. D. Cutshall, K. Eberly, H. Day, and B. Tietjen. 2013. Is there really no place like home? Movement, site fidelity, and survival probability of translocated and resident turtles. Biodivers. Conserv., 22:3185-3195.
Benson, J. F. and M. J. Chamberlain. 2007. Space use, survival, movements, and reproduction of reintroduced Louisiana black bears, J. Witdl. Manage., 71:2393-2403.
Bleisch, A. D. 2014. Initial movements and disturbance response of a newly re-introduced elk herd in the Missouri Ozarks. M.Sc. Thesis, University of Missouri, Columbia. 100 p.
Bonnot, T. W., J. H. Schulz, and J. J. Millspaugh. 2011. Factors affecting mourning dove harvest in Missouri. Wildl. Soc. Bull., 35:7(5-84.
--, M. A. Rumble, and J. J. Millspaugh. 2008. Nest success of black-backed woodpeckers in the forests with mountain pine beetle outbreaks in the Black Mills, South Dakota. Condor, 110:450-457.
Braastad, B. O. 1998. Effects of prenatal stress on behavior of offspring of laboratory and farmed mammals. Appl. Anim. Behav. Sci., 61:159-180.
Brookshire, B. L. and D. C. Dev. 2000. Establishment and data collection of vegetation-related studies on the Missouri Ozark Ecosystem Project study sites, p. 1-18. In: S.R. Shifley and B. L. Brookshire (eds.). Missouri Ozark Forest Ecosystem Project site history, soils, landforms, woody and herbaceous vegetation, down wood, and inventory methods for the landscape experiment. United States Department of Agriculture, Forest Service, North Central Forest Experiment Station, St. Paul, Minnesota. 314 p.
Burnham, K. P. and D. R. Anderson. 2002. Model selection and multi-model inference: a practical information-theoretic approach. Second edition. Springer, New York, New York. 488 p.
Carpenter, J. W., H. E. Jordan, and B. C. Ward. 1973. Neurologic disease in wapiti naturally infected with meningeal worms. J. Wildl. Dis., 9:148-153.
Clapp, J. G.,J. L. Beck, and K. G. Gerow. 2014. Post-release acclimation of translocated low-elevation, non-migratory bighorn sheep. Wildl. Soc. Bull., 38:657-663.
Clark, J. D., D. Huber, and C. Servheen. 2002. Bear reintroductions: Lessons and challenges: Invited paper. Ursus, 13:335-345.
Collazo, J. A., T. H. White Jr., F. J. Vilelia, and S. A. Guerrero. 2003. Survival of captive-reared Hispaniolan parrots released in Parque Nacional del Este, Dominican Republic. Condor, 105:198-207.
Couret, D., A. Prunier, A. Mounter, F. Thomas, I. Oswald, and E. Merlot. 2009. Comparative effects of a prenatal stress occurring during early or late gestation on pig immune response. Physiol. Behav., 98:498-504.
Devineau, O., T. M. Shenk, G. C. White, P. F. Doherty Jr., P. M. Lukacs, and R. H. Ruin. 2010. Evaluating the Canada lynx reintroduction programme in Colorado: patterns in mortality. J. Appl. Ecol., 47:524-531.
Dickens, M. J., D. J. Delehanty, and L. M. Romero. 2010. Stress: An inevitable component of animal translocation. Biol. Conserv., 143:1329-1341.
Dolev, A., D. Saltz., S. Bar-Damd, V. Yom-Tov. 2002. Impact of repeated releases on space-use patterns of Persian fallow deer. J. Wildl. Manage., 66:737-746.
Edge, W. D., C. Les Marcum, S. L. Olson, and J. F. Lehmkuhl. 1986. Nonmigratory cow elk herd ranges as management units. J. Wildl. Manage., 50:660-663.
Enochs, C. R. 1998. Here today, gone tomorrow: policies and issues surrounding wildlife reintroduction. W. Northwest J. Environ. Law and Policy, 1996-1998:91-106.
Fritts, S. H., W. J. Paul, and L. D. Mech. 1984. Movements of translocated wolves in Minnesota. J. Wildl. Manage., 48:709-721.
Fryxell, J. M., M. Hazell, L. Borger, B. D. Dalziel, D. T. Haydon, J. M. Morales, T. McIntosh, and R. C. Rosaite. 2008. Multiple movement modes by large herbivores at multiple spatiotemporal scales. PNAS, 105:19114-19119.
Gitzen, R. A., J. J. Millspaugh, and B. J. Kernohan. 2006. Bandwidth selection for fixed-kernel analysis of animal utilization distributions. J. Wildl. Manage.. 70:1334-1344.
Hardman, B. and D. Moro. 2006. Optimising reintroduction success by delayed dispersal: Is the release protocol important for hare-wallabies? Biol. Conserv., 128:403-411.
Haydon, D. T., J. M. Morales, A. Yorr, D. A. Jenkins, R. Rosatte, and J. M. Fryxell. 2008. Socially informed random walks: incorporating group dynamics into models of population spread and growth. Proc. II. Soc. B., 275:1101-1109.
Hudson, R. J., J. C. Hugh, and A. B. Bubenik. 2002. Physical and physiological adaptations, p. 199-257. In: D. E. Toweill and J. W. Thomas (eds.). North American elk: ecology and management. Smithsonian Institution Press, Washington, D.C. 1128 p.
Jachowski, D. S., R. A. Gitzen, M. B. Grenier, B. Holmes, and J. J. Millspaugh. 2011. The importance of thinking big: Large-scale prey conservation drives black-footed ferret reintroduction success. Biol. Consent.. 144:1560-1566.
Kemink, K. M. and D. C. Kesler. 2012. Using movement ecology to inform translocation efforts: a case study with an endangered lekking bird species. Anim. Conserv., 16:449-457.
Komers, P. E. and G. P. Curman. 2000. The effect of demographic characteristics on the success of ungulate re-introductions. Biol. Conserv., 93:187-193.
Larkin, J. L., D. S. Maehr, J. J. Cox, M. W. Wichrowski, and R.D. Crank. 2002. Factors affecting reproduction and population growth in a restored elk Cervus elaphus nelson population. Wildl. Biol., 8:49-54.
--,J.J. Cox. M. W. Wichrowski, M. R. Dzialak, and D. S. Maehr. 2004. Influences on release-site fidelity of translocated elk. Restor. Ecol., 12:97-105.
Littell, R., G. Milliken, W. Stroup, R. Wolfinger, O. Schabenberger. 2006. SAS[R] for mixed models. SAS Institute, Gary, North Carolina. 840 p.
McIntosh, T. E., R. C. Rosatte, J. Hamr, D. L. Murray. 2014. Patterns of mortality and factors influencing survival of a recently restored elk population in Ontario, Canada. Restor. Ecol., 22:806-814.
Millspaugh, J. J., R. A. Gitzen, B.J. Kernohan, M. A. Larson, and C. L. Clay. 2004. Comparability of three analytical techniques to assess joint space use. Wildl. Soc. Bull., 32:148-157.
Missouri Department of Conservation. 2010. Elk restoration in Missouri. 25 p.
--. 2009. Peck Ranch Conservation Area, <http://mdc4.indc.mo.gov/applications/moatlas/ AreaSummaryPage.aspx?txtAreaID=5203>. Accessed 1 Apr. 2014. 2 p.
--. 2005. Current River Hills Conservation Opportunity Area. 4 p.
--. 2000. Missouri elk reintroduction feasibility study. 37 p.
Moehrenschlager, A. and D. W. Macdonald. 2003. Movement and survival parameters of translocated and resident swift foxes Vulpes velox. Anim. Conserv., 6:199-206.
Moseby, K, B. M. Hill, and T. H. Lavery. 2014. Tailoring release protocols to individual species and sites: one size does not fit all. PLoS One. 9:1-12.
Mosillo, M., E. J. Heske, and J. D. Thompson. 1999. Survival and movements of translocated raccoons in northcentral Illinois. J. Wildl. Manage.. 63:278-286.
Nature Conservancy, The. 1997. The Nature Conservancy's Chilton Creek Management Area: A case study of the effects of fire-based ecosystem management in the Missouri Ozarks. Missouri Ozarks Office, Van Buren, Missouri. 8 p.
Nelson, L. J. and J. M. Peek. 1982. Effect of survival and fecundity on rate of increase of elk. J. Wildl. Manage., 46:535-540.
O'Gara, B. W. and R. G. Dundas. 2002. Distribution: Past and present, p. 67-119. In: D. E. Toweill and J. W. Thomas (eds.). North American elk: ecology and management. Smithsonian Institution Press, Washington. D.C. 1128 p.
Parlato, E. H. and D. P. Armstrong. 2013. Predicting post-release establishment using data from multiple reintroductions. Biol. Conserv., 160:97-104.
Popp, J. N., T. Toman, F. F. Mallory, and J. Hamr. 2014. A century of elk restoration in eastern North America. Restor. Ecol., 22:723-730.
Raithel, J. D., M. J. Kaufman, D. H. Pletsciier. 2007. Impact of spatial and temporal variation in calf survival on the growth of elk populations. J. Wildl. Manage., 71:791-803.
Rodriguez, A., L. Barrios, and M. Delibes. 1995. Experimental release of an Iberian lynx (Lynx pardinus). Biodivers. Conserv., 4:382-394.
Rosatte, R., J. Hamr, J. Young, I. Filion, and H. Smith. 2007. The restoration of elk (Cervus elaplius) in Ontario, Canada: 1998-2005. Restor. Ecol., 15:34-43.
Ruth, T. K., k. A. Logan, L. L. Sweanor, M. G. Hornogker, and L. J. Temple. 1998. Evaluating cougar translocation in New Mexico. J. Wildl. Manage., 62:1264-1275.
Ryckman, M. J. R. C. Rosatte, T. McIntosh, J. Hamr, and D. Jenkins. 2010. Postrelease dispersal of reintroduced elk (Cervus elaphus) in Ontario, Canada. Restor. Ecol., 18:173-180.
Sager-Fradkin, R. A., K. J. Jenkins. R. A. Hoffman, P. J. Happe, J. J. Beecham, and R. G. Wright. 2006. Fix success and accuracy of global positioning system collars in old-growth temperate coniferous forests. J. Wildl. Manage., 71:1298-1308.
Scillitani, L., G. Darmon, A. Monaco, G. Cocca, E. Sturaro, L. Rossi, and M. Ramanzin. 2013. Habitat selection in translocated gregarious ungulate species: an interplay between sociality and ecological requirements. J. Wildl. Manage., 77:761-769.
Seaman, D. E., J.J. Millspaugh, B.J. Kernohan, and G. C. Brundige. 1999. Effects of sample size on kernel home range estimates. J. Wildl. Manage., 63:739-747.
Seidel, K. D. 1992. Statistical properties and applications of a new measure of joint space use for wildlife. M.Sc. Thesis, University of Washington, Seattle. 108 p.
Shelton, M. and J. E. Hi ston. 1968. Effects of high temperature stress during gestation on certain aspects of reproduction in the ewe. J. Anim. Sci., 27:153-158.
Shifley, S. R., L. M. Roovers, R. G. Jensen, and D. R. Larsen. 2000. Composition and structure of woody forest vegetation in the Missouri Ozarks, p. 71-106. In: S. R. Shifley and B. L. Brookshire (eds.). Missouri Ozark Forest Ecosystem Project site history, soils, landforms, woody and herbaceous vegetation, down wood, and inventory methods for the landscape experiment. United States Department of Agriculture, Forest Service. North Central Forest Experiment Station, St. Paul, Minnesota. 314 p.
Spinola, R. M., T. L. Serfass, and R. P. Brooks. 2008. Survival and post-release movements of river otters translocated to western New York. Southeastern Nat., 15:13-24.
Teixeira, C. P., C. S. De Azevedo, M. Mendl, C. F. Cipreste, and R. J. Young. 2007. Revisiting translocation and reintroduction programmes: the importance of considering stress. Anim. Behav., 73:1-13.
Tousignant, J. 2011. The Current River Hills of southern Missouri: a CCPI EQIP proposal. 21 p.
Turchin, P. 1998. Quantitative analysis of movement: measuring and modeling population redistribution in animals and plants. Sinauer Associates, Sunderland, Massachusetts. 396 p.
U.S. Climate Data. 2017. U.S. climate data. Van Buren, Missouri, <http://www.usclimatedata.com/ climate/van-buren/missouri/united-states/usmo0892>. Accessed 25 Jan. 2017.
Van Winkle, W. 1975. Comparison of several probabilistic home-range models. J. Wildl. Manage., 39:118-123.
Vandel, J.-M., P. Stahl, V. Herrenschmidt, and E. Marboutin. 2006. Reintroduction of the lynx into the Vosges mountain massif: from animal survival and movements to population development. Biol. Conserv., 131:370-385.
Virginia Department of Game and Inland Fisheries. 2010. Elk restoration and management options for southwest Virginia. 77 p.
Vore, J. M. and E. M. Schmidt. 2001. Movements of female elk during calving season in northwest Montana. Wildl. Soc. Bull., 29:720-725.
Williams, C. L., T. L. Serfass, R. Cogan, O. E. Rhodes Jr. 2002. Microsatellite variation in the reintroduced Pennsylvania elk herd. Mol. Ecol., 11:1299-1310.
Witmer, G, 1990. Re-introduction of elk in the United Slates. J. Penn. Acad. Sci., 64:131-135.
Yott, A., R. Rosatte, J. A. Sghaffer, J. Hamr, and J. Fryxell. 2011. Movement and spread of a founding population of re-introduced elk (Cennts elaphns) in Ontario, Canada, Restor. Ecol., 19:70-77.
Submitted 31 December 2015
Accepted 3 April 2017
AMY D. BLEISCH (1), BARBARA J. KELLER (2) and THOMAS W. BONNOT
Department of Fisheries and Wildlife Sciences, University of Missouri, Columbia 65211
LONNIE P. HANSEN
Missouri Department of Conservation, Columbia 65201
JOSHUA J. MILLSPAUGH
Department of Ecosystem and Conservation Sciences, University of Montana, Missoula 59808
(1) Corresponding author: e-mail: email@example.com
(2) Present address: Missouri Department of Conservation, Columbia 
Caption: Fig. 1.--Missouri elk restoration zone in parts of Carter, Reynolds, and Shannon Counties where elk were released in 2011-2013 (1A). Elk were released on the Missouri Department of Conservation's Peck Ranch Conservation Area (PCRA) in 2011-2013 and The Nature Conservancy's Chilton Creek Preserve (CCP) in 2012 (1B). The CPS locations collected from the first 10 d of post-release monitoring for elk re-introduced to south-central Missouri (1C)
Caption: Fig. 2.--Daily average distance (m) from the release site for maternal (black) and nonmaternal (gray) elk re-introduced to south-central Missouri 2011-2013
Caption: Fig. 3.--Model-averaged estimates of maximum displacement from the release site flue to biological and release factors for elk re-introduced to south-central Missouri 2011-2013. Legends indicate age classes adult (light gray) and yearling (medium gray), maternal status nonmaternal (light gray) and maternal (medium gray), sexes female (light gray) and male (medium gray), release sites Peck Ranch (light gray) and Chilton Creek (medium gray), and release years 2011 (light gray), 2012 (medium gray), and 2013 (dark gray)
Caption: Fig. 4.--Model-averaged estimates of average range sizes 0-10, 11-31, 32-61, and 62-183 d post release based on biological and release factors for elk re-introduced to south-central Missouri 2011-2013. Legends indicate age classes adult (light gray) and yearling (medium gray), maternal status nonmaternal (light gray) and maternal (medium gray), sexes female (light gray) and male (medium gray), release sites Peck Ranch (light gray) and Chilton Creek (medium gray), and release years 2011 (light gray), 2012 (medium gray), and 2013 (light gray)
Fig. 5.--Model-averaged estimates for volume of intersection indices between utilization distribution pairs representing 0-10, 11-31, 32-61, and 62-183 d post release for elk re-introduced to south-central Missouri 2011-2013. Legends indicate age classes adult (light gray) and yearling (medium gray), maternal status nonmaternal (light gray) and maternal (medium gray), sexes female (light gray) and male (medium gray), release sites I'eck Ranch (light gray) and Chilton Creek (medium gray), and release years 2011 (light gray), 2012 (medium gray), and 2013 (dark gray)
Table 1.--Model selection results for factors influencing maximum displacement from the release site achieved in four sequential time periods (0-10. 11-31, 32-61, and 62-183 d post release) for elk re-introduced to south-central Missouri 2011-2013. The global model included the variables period, site, year, sex, age, and maternal status. We report log likelihood (LL), number of parameters (K), Akaike Information Criterion (AIC), AIC corrected for small sample sizes ([AIC.sub.C]), and the difference in AIC, from the most supported model ([DELTA] [AIC.sub.C]) Rank Model LL K AIC [AIC.sub.C] 1 PERIOD SITE SEX ACE 6277 17 6311 6313 2 PERIOD SITE SEX 6279 16 6311 6313 3 PERIOD YEAR SITE SEX 6276 18 6312 6314 4 GLOBAL 6273 20 6313 6316 5 SEX 6273 20 6313 6316 Rank [DELTA] [w.sub.i] [AIC.sub.C] 1 0 0.361 2 0.598 0.268 3 1.918 0.138 4 2.951 0.083 5 2.952 0.083 Table 2.--Model selection results for factors influencing 95% volume utilization distribution range sizes in four sequential time periods (0-10, 11-31, 32-01, and 62-183 d post release) for elk re-introduced to south-central Missouri 2011-2013. The global model included the variables period, site, year, sex, age, and maternal status. We report log likelihood (LL), number of parameters (K), Akaike Information Criterion (AIC), AIC corrected for small sample sizes ([AIC.sub.C]), and the difference in [AIC.sub.C] from the most supported model ([DELTA][AIC.sub.C]) Rank Model LL K AIC [AIC.sub.C] 1 PERIOD YEAR SITE AGE 12109 10 12129 12130 2 PERIOD SITE YEAR 12111 9 12129 12130 3 PERIOD YEAR SITE SEX 12111 10 12131 12132 4 GLOBAL 12109 12 12133 12133 Rank [DELTA] [w.sub.i] [AIC.sub.C] 1 0 0.419 2 0.228 0.374 3 2.320 0.131 4 3.889 0.060
|Printer friendly Cite/link Email Feedback|
|Author:||Bleisch, Amy D.; Keller, Barbara J.; Bonnot, Thomas W.; Hansen, Lonnie P.; Millspaugh, Joshua J.|
|Publication:||The American Midland Naturalist|
|Date:||Jul 1, 2017|
|Previous Article:||Characterization and isolation of ten microsatellite loci in false map turtles, graptemys pseudogeographica (Emydidae, Testudines).|
|Next Article:||Winter habitat associations for spotted skunks (Spilogale spp) in South-Central Wyoming.|