| Summary: | Thesis (M.S.) University of Alaska Fairbanks, 2015 The goal of this thesis is to define specific parameters of mesoscale sea-ice seascapes for which walruses show preference during important periods of their natural history. This research thesis incorporates sea-ice geophysics, marine-mammal ecology, remote sensing, computer vision techniques, and traditional ecological knowledge of indigenous subsistence hunters in order to quantitatively study walrus preference of sea ice during the spring migration in the Bering Sea. Using an approach that applies seascape ecology, or landscape ecology to the marine environment, our goal is to define specific parameters of ice-patch descriptors and mesoscale seascapes in order to evaluate and describe potential walrus preference for such ice and the ecological services it provides during an important period of their life-cycle. The importance of specific sea-ice properties to walrus occupation motivates an investigation into how walruses use sea ice at multiple spatial scales when previous research suggests that walruses do not show preference for particular floes. Analysis of aerial imagery, using image processing techniques and digital geomorphometric measurements (floe size, shape, and arrangement), demonstrated that while a particular floe may not be preferred, at larger scales a collection of floes, specifically an ice-patch (< 4 km²), was preferred. This shows that walruses occupy ice patches with distinct ice features such as floe convexity, spatial density, and young ice and open water concentration. Ice patches that are occupied by adult and juvenile walruses show a small number of characteristics that vary from those ice patches that were visually unoccupied. Using synthetic aperture radar imagery, we analyzed co-located walrus observations and statistical texture analysis of radar imagery to quantify seascape preferences of walruses during the spring migration. At a coarse resolution of 100-9,000 km², seascape analysis shows that, for the years 2006-2008, ...
|
|---|