(Table 1) Heavy minerals in silt and sand samples of IODP Site 301-U1301

The northwestern Cascadia Basin of western North America accumulated high-sedimentation-rate sequences during the Pleistocene sea-level low-stands. The continental shelf was largely exposed at that time, and rivers and estuaries delivered large sediment fluxes directly to the deep ocean. The IODP EXP1301 core, which was taken from the middle portion of the Cascadia Basin, is well preserved and exhibits the deeper and - more distal sedimentary facies. The lithology in this location is composed of two units, 1) hemipelagic mud with a thin sand layer and 2) thick, coarsening upward silt-sand turbidites with a small proportion of granules at the top. We will focus on the detailed sand-grain proportions in order to understand the origin of these sediments. We determined the modal proportions of the heavy minerals, and the chemical composition of olivine and orthopyroxene in fourteen samples. These are characterized by an abundance of amphibole, pyroxenes and epidote, and the presence of minerals derived from peridotite. There is no drastic change in the modal and mineral compositions of the sands and silts between the turbidite and hemipelagic sequences. There were two probable drainage systems on the continent, the Frazer and Columbia rivers, which shed turbidites into the Cascadia Basin after 1.6 Ma, especially at 0.46-0.76 Ma. Based on a comparison of the modal and mineral compositions, the Northern Cascadia Basin has been supplied with sediments, mainly from the Frazer River, through the Straits of Juan de Fuca, by Pleistocene to Holocene turbidites.

(Table 1) Nesting characteristics of greater snow geese (Chen caerulescens atlanticus) on Bylot Island between 1995-2005

1. Habitat heterogeneity and predator behaviour can strongly affect predator-prey interactions but these factors are rarely considered simultaneously, especially when systems encompass multiple predators and prey. 2. In the Arctic, greater snow geese Anser caerulescens atlanticus L. nest in two structurally different habitats: wetlands that form intricate networks of water channels, and mesic tundra where such obstacles are absent. In this heterogeneous environment, goose eggs are exposed to two types of predators: the arctic fox Vulpes lagopus L. and a diversity of avian predators. We hypothesized that, contrary to birds, the hunting ability of foxes would be impaired by the structurally complex wetland habitat, resulting in a lower predation risk for goose eggs. 3. In addition, lemmings, the main prey of foxes, show strong population cycles. We thus further examined how their fluctuations influenced the interaction between habitat heterogeneity and fox predation on goose eggs. 4. An experimental approach with artificial nests suggested that foxes were faster than avian predators to find unattended goose nests in mesic tundra whereas the reverse was true in wetlands. Foxes spent 3-5 times more time between consecutive attacks on real goose nests in wetlands than in mesic tundra. Their attacks on goose nests were also half as successful in wetlands than in mesic tundra whereas no difference was found for avian predators. 5. Nesting success in wetlands (65%) was higher than in mesic tundra (56%) but the difference between habitats increased during lemming crashes (15%) compared to other phases of the cycle (5%). Nests located at the edge of wetland patches were also less successful than central ones, suggesting a gradient in accessibility of goose nests in wetlands for foxes. 6. Our study shows that the structural complexity of wetlands decreases predation risk from foxes but not avian predators in arctic-nesting birds. Our results also demonstrate that cyclic lemming populations indirectly alter the spatial distribution of productive nests due to a complex interaction between habitat structure, prey-switching and foraging success of foxes.