Habitat use by Dall sheep and an interior Alaska mammal community
Date
2016
Authors
Dertien, Jeremy S., author
Doherty, Paul F., advisor
Aldridge, Cameron, committee member
Bagley, Calvin F., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Anthropogenic disturbances are increasingly recognized for effects on the behavior and physiology of wildlife species. Military training, a potential source of disturbance, has shown mixed behavioral and physiological effects on wildlife, including mountain ungulates. Dall sheep (Ovis dalli dalli) are an important species for hunting and wildlife viewing in Alaska and have shown an aversion to some forms of human disturbance such as direct overflights. Military training is expanding into potential Dall sheep habitat on two training areas of Fort Wainwright, Alaska; Molybdenum Ridge and Black Rapids Training Area. I placed camera traps in expected optimal and sub-optimal Dall sheep habitat to estimate the spatiotemporal habitat use of sheep and to make training recommendations to the U.S. military. Then, I further explored the available data and estimated the habitat use of species in four different mammalian guilds and the co-occurrence of habitat use between apex predators and potential prey species. In Chapter 1, I introduce the impetus for the study, the use of 54 camera traps in respect to mountain sheep, and the overall study design. My cameras captured over 8,000 images of sheep during the continuous 15-month sampling period. I successfully captured images of sheep traveling, foraging, resting, and interacting with other individuals. Occupancy models of detection-non-detection data suggest that abiotic covariates including slope, snow depth, and distance to escape terrain were the most important factors determining habitat use. Seasonal differences in habitat use suggested higher use of the Molybdenum Ridge study site during pre-rut, rut, winter, and lambing seasons with limited use during the summer, while habitat use estimates of Black Rapids were too imprecise to make broader inferences. Detection probabilities were temporally constant, but were positively correlated with cameras on a wildlife trail. From these results, I recommend that the U.S. Army concentrates training on Molybdenum Ridge during the early-July to early-September period and minimize training on both study sites during the lambing periods of May and June. If training were to occur on Molybdenum Ridge outside of this period, training should be concentrated around the easternmost valley/bowl of the ridge and the eastern half of the major south-facing slope of the ridgeline. In Chapter 2, I expanded upon the analysis of Dall sheep habitat use and investigated the alpine habitat use of ten species within four mammalian guilds. I analyzed how spatial covariates and temporal patterns correlated with habitat use of these species within and between guilds. Further, I modeled two-species occupancy of grizzly bears and wolves with different prey species (e.g., caribou and sheep). My results suggest that small and large herbivore habitat use positively correlated with vegetation and rock ground coverages, while large herbivores also correlated with broader abiotic covariates. Meso- and apex predator detections were sparse possibly leading to imprecise estimates of habitat use and little support for most habitat covariates. Detection probabilities of Dall sheep and predators were improved by cameras on trails. Two-species models suggested co-occurrence of habitat use between grizzly bear/caribou and wolf/caribou and independence of habitat use between grizzly bear/squirrel and wolf/sheep.