Many different types of riparian areas can be found throughout the seven-State Midwest Region. The ecological and social importance of riparian areas as landscape elements is undisputed (e.g., see Verry et al. 2000); however, there is great uncertainty about the cumulative importance of riparian areas in many landscapes and regions. This is a consequence of inadequate inventory data on the extent, type, and land use/land cover of riparian resources. The seven-State Midwest Region of the continental United States is no exception (figure 1). This region is one of the most water-rich areas in the continental United States, yet we are not aware of any general accounting of amount, type, or land use of riparian resources here. Because of the lack of regional data on riparian resources, it is difficult to determine the degree to which different types of riparian areas receive protection, the extent of change in riparian land use/land cover over time, and the social and economic benefits derived from different types of riparian land use. At issue is the availability and adequacy of riparian resource information used to make policy and management decisions in the region.

Forest management practices imposed at one spatial scale may affect the patterns and processes of ecosystems at other scales. These impacts and feedbacks on the functioning of ecosystems across spatial scales are not well understood. We examined the effects of silvicultural manipulations simulated at two spatial scales of management planning on landscape pattern and assessed the implications for forest-interior bird species. Landscape context was taken into consideration in determining harvest locations in the landscape-base management planning scenario but not in the stand-base planning scenario (where the focus of planning activities was at the level of individual stands and the context in which stands were located was not considered). We also compared ecological implications of patterns created at the stand and landscape levels by even-and uneven-age silvicultural systems. We used a harvest simulator (HARVEST) to simulate even-age, uneven-age and a combination of even-and uneven-age management systems for a period of 5 decades in the two forest management planning scenarios. Clearcuts of 5 to 16 ha were simulated to represent even-age management and small openings of 0.09 to 0.22 ha scattered throughout a stand were simulated to represent uneven-age management. Forest management that considered landscape context generated greater landscape total core area compared to that of the stand-base planning. There was a difference in landscape mean patch size, interspersion index, Simpson's diversity index and total core area for patches defined by stand age between stand-and landscape-base management planning. These results indicate that different landscape patterns can be produced by management planning conducted at different spatial scales. The scale of focus should depend on the management goals. Silvicultural manipulations at the stand level can cause the creation of different patterns at the stand and landscape levels. Such differences can lead to different ecological implications at each of those levels, thereby making it difficult to simply aggregate stand-level responses to the landscape-level.Furthermore, the ecological effects of landscape patterns on processes can be highly variable as the effects depend on how patches are defined. 0 1997 Elsevier Science B.V.

The spatial coincidence between riparian buffers of various widths and extents and potentially unstable ground was quantified using a physically based model for shallow landslide initiation and GIS for two watersheds on the Olympic Peninsula, Washington, USA. The proportion of the potentially unstable ground in each watershed within riparian buffers is a function of both buffer width and the extent of the stream channel network being buffered. While current buffers required by Washington State cover less than 5% of the potentially unstable ground, buffering all stream channels in these watersheds with 100-m buffers covered 75%-90% of the potentially unstable areas. Our analyses further show that: (1) riparian buffers are not efficient mechanisms for protecting potentially unstable ground, and (2) identifying potentially unstable ground using a physically based model should prove more effective for designing methods to reduce shallow landsliding hazards than relying on extensive buffer zones along stream channels.

No abstract