Channel morphology and restoration of Sitka spruce (Picea sitchensis) tidal forested wetlands, Columbia River, United States .

摘要

Globally, land conversion has reduced the areal extent of tidal freshwater floodplain forests, swamp ecosystems that are shaped by interactions between local geology and both riverine and oceanic hydrologic processes. Commonly, the lateral connectivity between the main stem and floodplain has been severed through diking for agriculture or urbanization, reducing available habitats for many species. Today, hydrologic connectivity is being restored with the goal of reestablishing swamps, and acquiring knowledge of swamp reference conditions to inform project design and the prediction and assessment of restoration trajectories. Reference conditions are thought to vary both within and among regions; however, few locations remain in which the reference conditions and restoration responses of tidal swamps on temperate zone large river floodplains can be compared. One such location is the 235-km tidal portion of the Columbia River on the West Coast U.S.A. This dissertation assesses geomorphological aspects of tidal freshwater Picea sitchensis (Sitka spruce) wetlands on the floodplain of the Columbia River using methods from fluvial and estuarine sciences. Field survey and Light Detection and Ranging (LiDAR) data were collected from three reference swamps and three restoration sites in the vicinity of Grays Bay some 37 river kilometers from the Pacific Coast in Washington State. On the basis of pool spacing and observed sequences of logjams and pools, channels were found to be consistent with the characteristics of fluvial forced step-pool types. This research documented correlations between channel morphology and catchment-scale features of the channel network that were consistent with previously published hydraulic geometry of tidal salt marsh systems located in the United Kingdom and in San Francisco Bay, California. Restoration sites behind dikes have subsided due to lack of sediment supply and therefore receive more frequent inundation than reference swamps. This produces a greater channel cross-sectional area at the outlet than would be predicted from using data on the hydraulic geometry of the reference sites. Early indications are that hydrologic reconnection does not change up-channel cross-sectional area, which may have effectively "fossilized" with the channel network when diking reduced both frequency of inundation and sediment supply. Though restoration sites generally lack accumulations of large wood, buried logs in channels were observed to produce post-restoration channel morphologies consistent with reference swamp step pools. Logs on the floodplain also appear to influence microtopography and plant associations in the swamps. Site-scale restoration affects ecosystem processes including the production and recruitment of large wood, tidal prism, and sediment supply that produce measurable patterns in hydraulic geometry and channel morphology, and may also have cumulative effects on ecosystem processes at the landscape or riverscape scale that influence new restoration sites. Such cumulative effects should be tested through modeling and experimental spatial clustering and temporal sequencing of hydrologic restoration projects.