- Dead wood patterns and dynamics vary with biophysical factors, disturbance history, ownership, and management practices. Through field and modeling studies, I examined the current and potential future amounts of dead wood in two landscapes and region-wide in the Coastal Province of Oregon. The objectives of the first study were to (1) determine whether two landscapes with different recent disturbance histories differ in the amount and characteristics of dead wood; and (2) explore relationships between patterns of dead wood in each landscape to potentially related factors including topography. The objectives of the second study were to (1) describe current regional amounts of dead wood; (2) compare dead wood amounts across ownerships; (3) determine relationships between current dead wood amounts and ownership, current and past vegetation conditions, climate, topography, and soils; and (4) evaluate whether the factors related to dead wood patterns differed according to the scale of analysis. The objectives of the third study were to (1) characterize the projected future change in dead wood amounts in a multi-ownership Province; (2) determine the longevity of present-day dead wood of different types and sizes in relation to amendments from management and stand development; and (3) evaluate differences in management approaches in transitional dynamics and long-term patterns of dead wood. In the first study, I sampled logs and snags at four topographic positions (streams, lower slopes, middle slopes, upper slopes) in the Tillamook State Forest and the Siuslaw National Forest. These two landscapes experienced catastrophic fire at different points in recent history. I developed statistical models relating various attributes of dead wood abundance to biophysical variables related to climate, topography, historical vegetation, current vegetation, soils, and ecoregion. I found that the type and timing of disturbance was important to dead wood amounts and characteristics, and that potential source and sink areas for dead wood were related to topographic position. In particular, lower slopes had higher amounts of logs, and upper slopes had higher basal areas of potential source wood, in the form of snags and legacy (pre-fire) stumps. Climatic factors were of greater relative importance to overall gradients of dead wood in the landscape in which fire occurred less recently. In the second study, I analyzed dead wood data from a region-wide systematic grid of field plots according to ownership and biophysical variables at multiple scales of resolution including plots, subwatersheds. Dead wood abundance and types varied greatly among ownerships, with public lands (Forest Service, Bureau of Land Management, State of Oregon) typically having higher amounts of dead wood and more dead wood in the larger size classes than the private lands (forest industry, non-industrial private). I found that the relative influence of ownership, topography, current and historical vegetation, and climate varied with scale of resolution. Current vegetation was of greater relative importance at finer scales of plots and subwatersheds, whereas climate, topography, and historical vegetation were of greater relative importance at coarser scales of watersheds and subbasins. Ownership was important to overall dead wood gradients at all scales considered. In the third study, by simulating stand development and dead wood dynamics under various forest management scenarios over a 300-year period, I was able to examine the long-term effects of management on dead wood abundance in the Coastal Province. I estimated potential upper bounds for future dead wood amounts. Dead wood amounts increased over time on average across the Province, mainly because of policies on public lands, especially the federal lands under the Northwest Forest Plan. Forest industry, under the Oregon Forest Practices Act and assuming retention of all snags at harvest and thinning, maintained amounts of dead wood that were similar to present-day levels, but size classes shifted toward the smaller sizes as existing large legacy dead wood decomposed. Non-industrial private lands showed increases from very low present-day amounts of dead wood. Across the Province, legacy logs and snags remained present for over a century of the simulation period, and buffered effects of intensive management to dead wood amounts. Variation across landscapes in starting conditions meant that contrasting management approaches had differential effects on long-term dead wood dynamics depending on where they were applied. Current amounts of dead wood and live vegetation patterns in the Province resulted from historical fire and logging. Results of this simulation study indicate that recently established policies oriented toward dead wood production and retention, in the absence of fire or other large- or mid-scale disturbances, are likely to result in increases in dead wood amounts that greatly exceed present-day levels. My results suggest that dead wood patterns of abundance will continue to diverge according to land ownership and that management practices that foster dead wood creation are of increasing importance to the long-term abundance of large dead wood as legacy dead wood is lost through decomposition.