Conservation and Ecology of Mammals III

Contributed Paper
ROOM: CC, Room 16

1:10PM Yellowstone Bison Manipulate the Green Wave
Chris Geremia; Jerod A. Merkle; Mark Hebblewhite; Daniel R. Eacker; Rick L. Wallen; Matthew J. Kauffman
Recent studies have shown that migrating large herbivores closely time their movements during spring to extend access to forage plants at early phenological stages – a strategy termed “surfing the green wave.” In classic green-wave surfing, the wave of green forage progresses irrespective of the level of herbivory on the landscape. Yet, large herbivores have been known to alter productivity of forage patches where they concentrate their use (i.e., grazing lawns). Dense aggregations of large herbivores may thus be capable of altering phenology at larger scales, which could alter green-wave surfing. We evaluated green-wave surfing in bison (Bison bison), a species that still forms large groups in Yellowstone National Park. We show that daily movements of collared bison (n = 209 animal-years) from 2005 to 2015 are generally consistent with large-scale green wave surfing at some point during spring. Yet, bison also tended to slow down their movements towards the end of spring and let the green wave pass them by. Despite this, based on fecal samples (n = 203), bison maintain remarkably high-quality diets even after the green wave recedes and broad-scale forage quality declines. An analysis of small-scale herbivore exclosures deployed across key meadows in the bison range (n = 24) indicate that grazing is intense enough to suppress plant biomass and enhance forage quality in grazed versus ungrazed plots. Finally, we combined bison use metrics from GPS collar data and plant phenology metrics from satellite imagery across 12 years, and demonstrate that bison grazing intensity is capable of creating landscape-scale grazing lawns that increase overall productivity and alter plant phenology. Bison appear to manipulate the green wave through the intense herbivory they create while foraging in large groups. Such behavior may have cascading effects on ecosystem properties that are reliant on plant phenology and productivity.
1:30PM Impacts of a High-Intensity Hurricane on Habitat for the Imperiled Big Cypress Fox Squirrel
Kira L. Hefty; John Koprowski
Climate change has led to an increase in hurricane intensity and frequency for regions within and surrounding the Gulf of Mexico. Environmental damage caused by these storms can be particularly harmful to wetland habitat and associated wildlife species, however, few studies have measured how high intensity hurricanes may be impacting these unique ecosystems. In the fall of 2017, we investigated how Hurricane Irma affected habitat for the Big Cypress fox squirrel (Sciurus niger avicennia), a threatened arboreal species endemic to slash pine and cypress mosaic communities in southwest Florida. S. n. avicennia are geographically isolated by extensive anthropogenic landscape alteration and are sensitive to change that may result in lowered habitat quality. The focus of our study was to determine the level of acute degradation to habitat for S. n. avicennia caused by high intensity hurricane activity and how that degradation may impact occupancy of S. n. avicennia throughout its range. We used a combination of direct vegetation measurements one month post-hurricane, remote sensing, repeat photography, and data from pre-hurricane occupancy analysis to assess changes to habitat structure and composition and potential impact to S. n. avicennia occupancy. Results indicate hurricane maximum sustained wind speed, water level, vegetative species, vegetative structure, and distance to fragmented edge were significant predictors of extent of damage to habitat for S. n. avicennia. Continued monitoring of these sites is necessary to determine long-term ecosystem resilience, changes in space use of S. n. avicennia, and potential compounded effects of multiple high intensity storms over consecutive hurricane seasons.
1:50PM Estimating Wild Horse Abundance Across 53 Million Acres
L. Stefan Ekernas; Bruce Lubow
Approximately 82,000 feral wild horses (Equus caballus) and burros (Equus asinus) range across 53 million acres of land managed by the Bureau of Land Management (BLM), spanning 10 states and 177 Herd Management Areas. As introduced species with rapidly growing populations that by federal law cannot be harvested, wild horses & burros pose unique management challenges. The 1971 Wild Free-Roaming Horses and Burros Act dictates that BLM shall maintain a current inventory of wild free-roaming horses and burros. To achieve that policy, BLM partnered with U.S. Geological Survey (USGS) and independent scientists to develop accurate and precise methods for estimating current wild horse & burro abundance across their range. USGS tested several methods in areas with known wild horse abundance and found that simultaneous double-observer aerial surveys are accurate to < 0.7 standard errors. BLM currently uses these validated methods to survey most of the 177 Herd Management Areas every second or third year, with approximately 20 million acres surveyed annually. Flights cover 100% of each Herd Management Area, plus any surrounding lands potentially occupied by wild horses & burros. We analyze data from surveys using Huggins mark-recapture models in program R, estimating parameters for numerous detection covariates with the package RMark, and then applying these detection parameters to observed animal groups to calculate abundance. We use a combined parametric and non-parametric bootstrap procedure to measure uncertainty around the abundance estimate, setting the observed number of animals as a floor for uncertainty estimates. These more robust survey results will enable us to evaluate herd growth rates over time. Beginning in 2018, USGS will publicly release R code and raw data for each new wild horse & burro survey to facilitate transparency, external review, and application to other species and situations where simultaneous double-observer surveys may be conducted.
2:10PM How Do They Weather the Storm: White-Tailed Deer Habitat Selection and Movement during Hurricane Irma
Heather N. Abernathy; Daniel A. Crawford; Elina P. Garrison; Richard B. Chandler; Michael L. Conner; Karl V. Miller; Michael J. Cherry
Extreme climatic events (ECEs) are defined as weather disturbances that are relatively intense and infrequent, which alter ecosystem structure and/or function outside of the bounds of typical stochasticity. Such events are increasing in frequency and intensity and therefore, understanding factors that influence a population’s vulnerability to ECEs are critical to predicting how populations will respond to such events. Hurricane Irma hit southwestern Florida where we were using GPS tracking collars to monitor 54 (34 F, 16 M) adult white-tailed deer (Odocoileus virginianus, deer). We examined the hurricane’s effect on deer survival and mitigation strategies utilized by deer to offset potential negative impacts of the hurricane. During Irma, deer selected for upland habitats such as pine forests and avoided lower elevation marshes. Further, many deer (F:53% and M:25%) left their seasonal home ranges during the storm, likely to seek higher elevation habitat. Females increased (p=0.001) movement rates 55% during the storm, while males did not alter movement rates (p=0.41). No collared deer died during the storm suggesting that deer behavioral strategies effectively mitigated potential negative impacts of Hurricane Irma. Broadly, our findings highlight the importance of mobility and access to refugia, i.e., higher ground, with regard to wildlife vulnerability to ECEs. Documenting such adaptions can inform management and a growing body of ECE literature.
2:30PM Refreshment Break
3:20PM Pronghorn Avoidance Behavior to Different Kinds of Barriers
Benjamin Robb; Jeffrey Beck; Hall Sawyer; Matthew Kauffman
Anthropogenic barriers pose a significant threat to wildlife by limiting access to both forage and critical seasonal ranges. The effects of barriers, however, are species specific because of physiological and behavioral differences. Pronghorn (Antilocapra americana), for example, are known to be extremely sensitive to movement barriers, which can severe migrations and cause long-term reductions in population performance. However, the extent to which different kinds of anthropogenic barriers — from fences to roads— influence pronghorn remains uncertain and therefore difficult to incorporate into management plans. The objective of this study is to quantify the degree and severity of avoidance by pronghorn to different movement barriers. Here, we present an analysis of several different pronghorn herds found in southern Wyoming (N = 167 individuals, N = 3 herds), all of which were GPS collared between 2010 and 2015. Using a step-selection function (SSF) with a conditional logistic regression clustered by individual, we analyzed the severity of avoidance that pronghorn exhibited for a wide variety of barrier types, including: fences, railroads, county roads and interstates. Our results demonstrated pronghorn strongly avoid all types of barriers including fences (odds ratio 0.67, 95% confidence interval 0.62 – 0.72), railroads (odds ratio 0.50, 95% confidence interval 0.39 – 0.66), county roads (0.95, 95% confidence interval 0.92 – 0.99) and interstates (0.36, 95% confidence interval 0.24 – 0.55). Interstates and other state roads were 3 times less likely to be crossed than county roads and had the greatest degree of pronghorn avoidance. The quantification and prediction of these barrier effects will help managers identify critical local barriers that could be mitigated through fencing modifications or other types of crossing structures.
3:40PM Assessment of Ancestry in the Columbia Basin Pygmy Rabbit
Stacey A. Nerkowski; Paul A. Hohenlohe; Kenneth Warheit; Lisette P. Waits
Pygmy rabbits (Brachylagus idahoensis) are the smallest rabbit in North America and are unique among rabbits due to their ability to consume sagebrush. Loss and fragmentation of habitat due to agricultural conversion has led to the near extirpation of the disjunct pygmy rabbit population in the Columbia Basin (CB) of, Washington (WA) State. In 2003, the CB pygmy rabbit was listed as a distinct population segment under the US Endangered Species Act. In 2001, sixteen CB rabbits were taken from the last remaining population in Sagebrush Flats (SBF), WA to start a captive breeding program, and 9 Idaho rabbits were added to counteract the effects of inbreeding. Rabbits were moved to semi-wild breeding enclosures at SBF (2011), and since then ~1947 mixed ancestry rabbits have been released into the wild. To provide insight into genomic diversity and ancestry, we used a single digest restriction site associated DNA sequencing (RADseq) protocol on 189 pygmy rabbit samples from all source populations to generate the first genome-wide set of single nucleotide polymorphism (SNP) markers for this species. SNPs were aligned to the European rRabbit (Oryctolagus cuniculus) genome with an average of 62% alignment. Using STRUCTURE, four regional populations have been identified. The genetic clusters identified at K=4 were the CB, northern Utah/Wyoming, southern Utah, and the Great Basin plus Snake River plain. The establishment of these populations reflects the glacial pluvial lakes that existed in the area and may have restricted gene flow between the populations. A SNP panel to assess ancestry in the CB populations has been identified which will allow us to genetically monitor captive and wild populations of CB pygmy rabbits to guide strategies for conservation and management.
4:00PM Factors Influencing Demography in a Restored Cheetah Population: Context-Dependency of Top-Down and Bottom-Up Drivers
Laura C. Gigliotti; Rob Slotow; Luke Hunter; David S. Jachowski
Understanding drivers of population dynamics is critical for managing wildlife populations, especially for species of high conservation concern. Cheetahs (Acinonyx jubatus) are experiencing large-scale range contractions and population declines. As suitable habitat for cheetahs becomes more fragmented, small fenced reserves represent an important component for the persistence of this species. However, the population trajectories of cheetahs reintroduced into small reserves have varied greatly. Therefore, understanding the drivers of cheetah demography in small reserves is critical. We studied drivers of cheetah reproduction and survival in Phinda Private Game Reserve in South Africa using 25 years of post-reintroduction monitoring data. We investigated the effects of top down (lion density), bottom up (prey density), and density dependent (cheetah density) factors on cheetah reproduction and survival. We used multi-state live-encounter dead-recovery models to estimate monthly survival rates for adult and juvenile cheetahs, and generalized linear models to investigate drivers of litter size. Our results indicate context-dependency in drivers of cheetah demographic rates. We found that litter sizes were larger in the years immediately post-reintroduction, compared to after the population had stabilized. However, on a finer temporal scale there were no strong drivers of litter size. In contrast to the top-down and bottom-up hypotheses, survival was lowest during months with high prey densities and low lion densities, for both adults (0.73 ± 0.13) and juveniles (0.26 ± 0.12). We suggest that spatial arrangement of lions, cheetahs, and prey, as well as behavioral modifications by both cheetahs and prey, might be influencing the observed survival patterns. In addition, management actions, such as lion removals which are required in small restricted reserves, may be affecting cheetah demographics indirectly by altering the behavior of resident lions and cheetahs. Our results highlight the complexities of understanding drivers of cheetah demographic rates, which could improve future conservation efforts.
4:20PM Improving the Conservation of the Threatened Utah Prairie Dog Through the Evaluation of Genomic Variation and Population Structure.
Rachael M. Giglio; Tonie E. Rocke; Jorge E. Osorio; Emily K. Latch
Utah prairie dogs (Cynomys parvidens) are federally threatened and their population size has been in decline due to persecution, habitat destruction, and outbreaks of plague. Today, Utah prairie dogs exist in small, isolated subpopulations, making them less demographically stable and more susceptible to erosion of genetic variation by genetic drift. We evaluated the consequences of continued population decline on the genetic structure and diversity of Utah prairie dogs in order to inform conservation decisions. We generated single nucleotide polymorphisms (SNPs) for 319 individuals across 7 locations in southern Utah using double digest restriction site associated DNA sequencing (ddRAD). We detected genetic structure among all sampling locations, however, two sites had unique alleles (private alleles) and had a higher degree of differentiation from all sampling locations. Understanding patterns of genetic structure and genetic diversity can be used to inform the Utah prairie dog recovery plan by targeting populations with low genetic diversity as potential areas for translocations. Managers can prioritize conservation efforts on populations with high genetic variation and unique alleles in order to preserve that genetic variation.
4:40PM Spatiotemporal Patterns of Human-Panther Conflicts in a Human-Dominated Landscape
Marta Prat-Guitart; David P. Onorato; James E. Hines; Madan K. Oli
The growth of the endangered Florida panther (Puma concolor coryi) population within the anthropogenically altered landscape of southern Florida has increased opportunities for human-panther conflicts. In addition, the number of Florida panther vehicle collisions have increased, potentially threatening both human safety and the long-term persistence of the panther in the wild. Using a long-term dataset (2006-2017) of documented depredations and human-panther encounters, we analyzed the spatiotemporal patterns and ecological and anthropogenic factors driving these conflicts using dynamic occupancy modeling. The total number of human-panther conflicts increased during the study period, with depredations being the most frequent. Probability of a depredation or human-panther encounter was higher at sites that had higher fragmentation and close habitat patches, which could be providing suitable cover for panthers to rest and stalk prey. In addition, the probability of extinction of these depredations and human-panther encounters decreased as the panther population increased in size. Using the same methodology, we also analyzed Florida panther deaths and injuries due to vehicle collision to assess the risk to panthers in southern Florida. The number of vehicle collisions increased during the study period, with larger risks to panthers in areas where habitat was more contiguous. However, sites with wildlife crossings and roads with wider shoulders had higher vehicle collision extinction probabilities. Future management efforts to mitigate and prevent depredations, human-panther encounters, and vehicle collisions should be targeted to those areas with the highest risk. The successful resolution of conflicts will be necessary to promote human-panther coexistence and the long-term persistence and recovery of the Florida panther.
12:50PM Wildlife Movement Models for Conservation Planning along Linear Barriers
Eric Abelson
Habitat fragmentation is increasingly threatening wildlife by reducing areas of suitable home range and creating barriers to movement. Identifying where wildlife movement corridors intersect with anthropogenic movement barriers (e.g., roads) is critical to determining locations for mitigation efforts, including wildlife underpasses. However, application of existing modeling approaches to on-the-ground conservation planning is often difficult due to differences between modeling methods. This study asks: how do various approaches to modeling wildlife movement differ in their identification of movement barriers and possible mitigation locations? I use a matrix of points across the landscape and full-factorial models to model wildlife movement across a ~2000 km2 landscape in the central Sierra Nevada Mountains in California. Resistant kernel, least-cost and circuit-theory based models were used to identify variation in on-the-ground movement paths. Variation was quantified by identifying locations where each movement model approach predicted wildlife movement intersecting existing barriers on the landscape. I present two key discussion points: 1. the importance for conservation planning to model movement across the landscape using factorial modeling approaches – this identifies movement across the landscape with the acknowledgement that wildlife often live and move though sub-optimal habitats. Additionally, this has important implications for on-the-ground locations of mitigation sites. 2. Use of multiple modeling methods in conjunction provide additive, rather than substitutive value for conservation modeling. Different models provide important, and unique, information to identify on-the-ground mitigation locations for applied ecology and conservation. This work identifies data requirements and a workflow for conservation planners to implement landscape connectivity modeling. Specifically, I propose a framework for considering intermediate-scale (hundreds to thousands of sq. km.) landscape modeling to identify where predicted high-use movement paths intersect with barriers (i.e. roads), which can then be targeted for mitigation. These modeling techniques can be used to identify locations most in need of mitigation.


Contributed Paper
Location: Cleveland CC Date: October 9, 2018 Time: 12:50 pm - 5:00 pm