Conservation and Ecology of Mammals – Ungulates

Contributed Paper
ROOM: Rooms 27 – Picuris, 29 – Sandia and 31 – Santa Ana Combined
SESSION NUMBER: 26
 

1:10PM Using Occupancy to Understand the Process of Moose Recolonization in Northern Wisconsin.
Lucas Olson; Max Allen; Timothy Van Deelen
Modern wildlife management, through regulation, land protection, and science-informed decision-making, has facilitated the recolonization of fauna throughout North America. Among these successes has been the recolonization of moose (Alces alces) to much of its historic range throughout both New England and the Northern Great Lakes Region. Despite recent population declines throughout Northern Minnesota and the Upper Peninsula of Michigan (UP), Northern Wisconsin has reported increasing observations of moose by members of the public since 1991. This raises the question whether moose recolonization is occurring in Northern Wisconsin, and whether its forests may provide a connective band of habitat for moose populations in Northern Minnesota and the UP. We report initial results using an occupancy framework to understand spatiotemporal patterns of citizen-reported moose observations at the county level, with models of occupancy based on habitat quality, metapopulation dynamics, climate effects, urban avoidance, and interspecific interactions, coupled with covariates for detection such as road density, wildland-urban interface area, and human population density. Whether or not moose recolonization is occurring in Northern Wisconsin, our research will help to understand the drivers in occupancy of a low-density species across large spatial extents; emphasizing Northern Wisconsin’s role in populations of moose in the Northern Great Lakes Region.
1:30PM Archery Hunting and Scouting Drives Early Fall Migration and Reduces High Quality Forage for Elk with Access to Protected Area Refuge
Nate Mikle; Tabitha A. Graves; Edward M. Olexa
Alteration of wide-ranging wildlife migrations can drastically impact the structure and function of ecosystems, yet the causes and consequences of shifting migration patterns remain largely unknown. Decisions made in one portion of a landscape may induce spatial and temporal shifts of wildlife use in another, creating tension among private, state, and federal lands with varying missions and feedback loops that can influence wildlife population sizes and the resources on which wildlife and humans depend. The timing of autumn migration has received limited attention due to the difficulty in assessing the extreme asynchrony in autumnal events, although nutrition during this time period is crucial to winter survival and reproduction. Here, we use 5 years of data from 73 female elk (Cervus canadensis) which utilize a landscape managed by 4 federal agencies, a state, and private landowners, to identify the driving factors behind the initiation of fall migration in two subpopulations, one of which migrates to a protected area where hunting is prohibited. Most elk departed summer range prior to frost or snow, with 67% of protected area elk migrating prior to the onset of archery hunting season (1 September), preemptively avoiding risk, while no elk from the other subpopulation left prior to archery season. However, departing from productive summer range prior to frost or snow and nearly two months before vegetation senescence to be afforded protection during hunting season decreased access to late summer-fall forage by 21%. Our results suggest that in areas where migratory ungulates span multiple jurisdictions, the benefits of migratory behavior may be dramatically impacted by unevenly distributed anthropogenic disturbance. As this is a common scenario globally, our work highlights the urgent need to improve our understanding of subtle changes in migratory behavior, both spatially and temporally, which may erode the resilience of migration to future change.
1:50PM A Framework for Estimating Elk Abundance in Arizona
Kirby Bristow; Mathew Clement; Michelle Crabb; Larisa Harding
Elk (Cervus elaphus) populations in Arizona have historically been managed using relative estimates of abundance. Recent concerns over the influence of large wildfires, impacts to aspen regeneration, and predator-prey relations and management, have caused the Arizona Game and Fish Department to seek methods to more accurately estimate elk abundance. Between 2014 and 2016, we conducted intensive helicopter surveys of selected areas to compare several methods designed to model elk detection and provide algorithms to account for animals missed on subsequent surveys. We conducted annual fall helicopter surveys in 3 areas, which contained a subset of radio-collared elk, and recorded information on covariates affecting both sightability (i.e. vegetative cover, vegetation type, burn category, group size, activity, and ambient light) and observer bias (i.e. observer position, pilot experience). We used information theory to rank a set of candidate a priori models to determine which covariates affected detection and select the most parsimonious models among sightability, double observer, and hybrid modeling methods. We then used the top model from each method to calculate annual site-specific elk abundance estimates for comparison to concurrent mark-recapture abundance estimates. The best supported models included all sightability covariates, and influence of each covariate followed predictable patterns relative to elk detection. The best performing hybrid model consistently provided a more accurate and precise abundance estimate, relative to mark-recapture estimates, than either sightability or double observer models. Application of our hybrid model to future helicopter survey data should improve both precision and accuracy of elk abundance estimates in Arizona.
2:10PM Ecological Drivers of Elk Survival in Idaho
Jon Horne; Mark Hurley; Erin Roche; Scott Bergen
Effective management of elk populations is facilitated by an understanding of the factors that influence elk survival. Over the past ~10 years, elk across the state of Idaho have been monitored for mortality and often times cause-specific mortality but to date, these data have not been used in a comprehensive survival analysis. We compiled known-fate survival data from ~2000 radio-collared elk (1120 adult cows and 880 6-month-old calves) that were monitored from 2005 – 2016. Statewide, lion and wolf predation were the main causes of mortality for cows (30% and 29%, respectively) and 6-month old calves (41% and 33%, respectively). However, mortality rates were highly variable across elk populations and years. To examine factors potentially causing this variation, each elk was assigned to one of 27 populations based on its winter range. We then modeled risk of mortality as a function of winter severity, summer nutritional resources, and wolf abundance. We found that elk survival is inherently complex but by utilizing a data set encompassing substantial spatial and temporal variation, we were able to identify the main drivers of elk survival in Idaho.
2:30PM Physiological Stress in Reintroduced Elk Reveals Tolerance to Pulses of Increased Human Activity
Colter Chitwood; Barbara J. Keller; Lonnie Hansen; Aaron M. Hildreth; Joshua J. Millspaugh
Fecal glucocorticoid metabolites (FGMs) provide a noninvasive means to evaluate physiological responses of wildlife to various stressors. Adrenocortical activity (i.e., stress response) is of interest to wildlife managers because stress can alter animal behavior, reduce disease resistance, and affect population performance. Consequently, we quantified FGMs among the free-ranging, newly established elk (Cervus elaphus) population in Missouri. Elk were translocated from Kentucky and released in 3 cohorts from 2011-2013, and we quantified FGMs seasonally from summer 2011 through fall 2014. Additionally, we used a general linear model to assess stress response at a finer scale by quantifying FGMs prior to and following 3 managed white-tailed deer (Odocoileus virginianus) hunts that occurred at the restoration site each fall. FGM levels changed seasonally (Welch’s F [3, 152.85] = 20.351; p < 0.001), with winter levels (3-yr = 24.22 ng/g [SE = 1.07]) lower in pairwise comparisons with all other seasons (all 3 pairwise Tamhane’s T2 p-values ≤ 0.002). Summer FGM levels (4-yr = 38.71 ng/g [SE = 1.94]) were higher than spring levels (3-yr = 32.19 ng/g [SE = 1.82]) and were greater (both pairwise Tamhane’s T2 p-values ≤ 0.008) than fall (4-yr = 32.13 ng/g [SE = 0.48]) and winter. FGM levels were not affected by hunter activity. Given that our study site had limited human disturbance except during the 3 deer hunts, these results combined with a complementary study on elk space use, suggest that the newly established elk population was robust to periodic increases in human activity over short time scales, even shortly after reintroduction. At longer time scales, our results corroborate previous work in other elk herds, suggesting that seasonal changes in elk FGM levels likely reflect responses to prolonged environmental stressors (e.g., forage quality or availability during the summer stress period).
2:50PM Refreshment Break
3:20PM Mortality and Human Predation: How Sociality Influences Probability of Mortality of Large Mammals
Brittany Slabach; John T. Hast; Philip H. Crowley; Will Bowling; R. Daniel Crank; Gabriel Jenkins; John J. Cox
Elk (Cervus canadensis) are highly gregarious, large mammals, where females exhibit stable affiliation patterns. Selective take, due to hunter harvest, has been shown to differentially effect recruitment and fecundity more drastically compared to non-human predators. Hunter harvest is also presumed to influence individual behavior and group dynamics, yet how affiliation patterns between individuals may offset probability of mortality has not been explored. We partnered behavioral observations with standard very high frequency (VHF) mortality collars to investigate the interaction between probability of hunter harvest and elk social structure in a population of elk residing in southeastern Kentucky, USA. Cow elk were outfitted with VHF collars and ear tags for identification (n = 94). Individual behavior, group composition, and mortality were monitored from Feb. 1, 2013 to Feb. 1, 2016. Hunter harvest was the primary cause of mortality, as expected, resulting in 54% of all mortality observed (38% modern gun; 12% archery; 4% wounding loss). The survival rate of elk for the duration of the study was 0.66 (0.4 – 0.8). Annual survival rate varied slightly between 2013 (n13 = 45; μ= 0.54, 0.19 – 0.9) and 2014 (n14 = 69; μ= 0.49, 0. 18 – 0.8) yet was highest for 2015 (n15 = 46; μ= 0.83, 0.7 – 0.95). Cox proportional-hazard regression revealed age-class, herd location (e.g., public versus private), and betweeness centrality – a measure of how central an individual is to the overall structure – were all associated with lower hazards. Specifically individuals with higher betweeness values were 10 – 25% less likely to be harvested depending upon herd location (95% CI: 0.55 – 0.9). Our results suggest that underlying herd social structure is an important component to understanding probability of mortality, and further highlight the potential negative effects of selective take of core herd members in ungulate populations.
3:40PM Behavioral and Nutritional Responses of Moose to Beetle-Killed Forests in Southeast Wyoming
Alexander A. B. May; Matthew J. Kauffman; Kevin L. Monteith
Conifer forest, a major habitat within Shiras moose (Alces alces shirasi) range, has experienced widespread mortality following the irruption of mountain pine beetles (Dendroctonus ponderosae). Tree mortality and resulting canopy loss are altering forest composition in ways that could affect moose, including potential benefits from increased forage and detriments through loss of refuges from heat and deep snow. We examined nutritional and behavioral effects of beetle-killed forest on moose in southeastern Wyoming, USA by fitting female moose with GPS collars and recapturing them twice annually for two years to measure nutritional condition and reproduction. We used resource selection functions to compare habitat use in summer and winter between moose currently (19 moose; 50,617 locations) and a GPS dataset collected a decade ago, before extensive beetle kill (8 moose; 42,002 locations). Habitat selection was similar pre- and post-beetle infestation, with selection for riparian and deciduous cover year round, and avoidance of conifer forest in summer. Moose that succeeded in raising a calf to spring lost body fat (-1.47% ± 0.51) over summer, whereas those whose calf did not survive gained fat (0.62% ± 0.5), suggesting inadequate forage to meet the costs of lactation. Annual adult survival was low (81%) and moose with low body fat were less likely to survive (p < 0.001). Low average March body fat (6.2%), virtually no twinning, and a low pregnancy rate (80%) was characteristic of a population experiencing nutritional limitation. The findings indicate that forest changes caused by bark beetles do not appreciably improve habitat quality for moose, and moose continue to employ similar patterns of habitat use regardless of forest changes thus far. Our study highlights the importance of riparian and deciduous habitats in sustaining moose populations in the face of widespread forest disturbances and altered climate.
4:00PM Influence of Beetle-Killed Forests on Elk Habitat Selection
Bryan Lamont; Dr. Matthew Kauffman; Dr. Kevin Monteith; Dr. Jerod Merkle; Matt Hayes; Tony Mong
Landscape disturbances in forested environments have the ability to both provide and hinder wildlife access to high-quality resources, as well as thermal and hiding cover. For nearly two decades, the forests of the Rocky Mountains (USA) have been experiencing a bark beetle epidemic of severity and duration that has not been seen in over a 100 years. As the epidemic changes the structure and characteristics of the forest, the ungulates that inhabit these areas may alter their resource selection patterns to adapt to changing habitats. As has been seen with other forest disturbances (fire and logging), in some cases elk (Cervus elaphus) increase use of disturbed areas as a result of increased understory vegetation, while in other cases elk selection may decrease as a result in an accumulation of down logs (slash). By employing global positioning system technology in conjunction with a satellite-derived land classification (NAIP 1m), we were able to quantify elk habitat selection as it relates to beetle-killed forests in south-central Wyoming. Elk showed an over-all avoidance of beetle-killed forests during all parts of the day, supporting our prediction that due to a loss of thermal cover elk would avoid beetle-killed areas during the hot parts of the day, although not supporting our prediction that due to the loss in canopy we would expect increases in understory vegetation leading to selection for beetle-killed areas during prime feeding times. While our results showed an over-all avoidance by elk for beetle-killed areas of the forest currently, as we move through time and the vast majority of standing dead trees begin to fall, it may be that forests and elk may respond more similarly to more classic, well-understood disturbances to conifer forests in western North America.
4:20PM Drone and Thermal Imaging System as an Innovative Tool for Forest Ungulate Surveys
Stanisław Pagacz; Julia Witczuk; Anna Zmarz; Maciej Cypel
Effective wildlife management require reliable assessments of animal abundance. However, no ungulate monitoring methods is entirely satisfying in terms of cost-effectiveness and accuracy. Combining unmanned aerial vehicles (drones) and thermal infrared imaging has the potential to overcome the difficulties experienced in ground-based ungulate surveys. Drones enable safe operations at low altitudes, and at night – a time that often offers the optimal conditions for wildlife monitoring. In January 2017 we used fixed-wing drone with thermal camera to conduct test surveys in a 5000 ha hunting district in central Poland. Pine-dominated coniferous forest covers ~65% of the area, while the rest is comprised of mainly agricultural land. The area is inhabited by red deer, roe deer, wild boar and moose. Flights in visual line of sight were flown after sunset (16:00-20:00) to increase the chances of detecting animals. Flight altitude was 150 m above the ground, resulting in a spatial resolution of 15 cm. Total of 39 transects, each 2.8 km long, were flown over three sample areas, covering ~20% of the entire district area. The distance between recorded animals and the transect line (image centerline) was used for density estimation by the distance sampling method. A total of 210 thermal signatures of animals were recorded on transects, estimated ungulate density was 24 individuals/100 ha (95% CI: 17-32). Analysis of the distribution of signature lengths indicated, that the most common species in the district are roe deer and wild boar but some red deer and moose were also identified. Results show that thermal surveys from drones are a promising method for ungulate enumeration in forests. They proved to be more efficient than commonly used ground-based counting techniques. The main challenges of the method are difficulties in automatic animal detection, species identification and regulations limiting drone operations to visual line of sight.
4:40PM Fire-History Exerts Persistent Effects on Ungulate Behavior: Daily, Seasonal and Successional Patterns
Derek B. Spitz; Mary Rowland; Darren Clark; Mike Wisdom; Taal Levi; Bruce K. Johnson
In temperate forests, fire is often expected to confer long-lasting benefits to ungulates by improving available nutritional resources. Fire opens the canopy, facilitating growth of understory vegetation that then gradually wanes as trees reestablish. These benefits have, however, remained difficult to test—in part because of their dependence on ungulates’ behavioral response to fire, which may span multiple temporal scales. We hypothesized that fire affects ungulate behavior across many years, that within a year ungulates use burns more in spring than summer, and that within spring ungulates use burns more when foraging. We tested these hypotheses using a long-term telemetry dataset on elk (Cervus elaphus) collected 1997-2012 from Starkey Experimental Forest and Range (n = 274 unique females). Between 2001 and 2003, 26 stands of fir (Abies spp.) and Douglas-fir (Pseudotsuga menziesii) were thinned and burned, while 27 similar stands were held as experimental controls. We used discrete choice models to test for temporal structure in elk responses to fire history and to quantify the magnitude and duration of responses. We found daily, seasonal and successional differences in elk selection for burned stands. Our results support the hypothesis that fire has persistent (>10yrs) effects on ungulate behavior and that these effects peak quickly (<5yrs) then gradually decline. Elk showed stronger selection for burns in spring than summer, but we found no support for the hypothesized relationship between burns and foraging behavior. Instead, we found sharp transitions between nocturnal and diurnal behavior, with elk avoiding burns during daylight hours. The behavioral responses we observed are only possible within a landscape with a heterogeneous fire history. All individuals observed spent more than two thirds of their time in unburned areas, suggesting that elk benefit from the ability to choose between a range of conditions with different attendant risks and rewards.

 

Contributed Paper
Location: Albuquerque Convention Center Date: September 24, 2017 Time: 1:10 pm - 5:00 pm