PARENT SESSION
Contributed Oral Session 4: Forest Management
Monday, August 8, 8:00 AM - 11:30 AM, Meeting Room 514 A, Level 5, Palais des congrès de Montréal

The use of remote sensing, GIS and multivariate statistical tools to explain the distribution of riparian forest communities at multiple spatial scales.

Mollot, Lauren ^{1, 2}, ¹ University of Washington, Seattle² Cedar River Watershed, Seattle

ABSTRACT- Timber management has reduced the diversity of forest seral age classes in watersheds throughout the Pacific Northwest. The harvest of conifer in riparian stands has resulted in a disproportionate share of red alder (Alnus rubra) and shrub dominated stands. This loss of conifer has been associated with a decline in anadromous fish habitat. Under what conditions should we restore conifer to stands currently dominated by deciduous communities in our effort to rehabilitate fish habitat? This research provides an analytical framework for addressing this question at multiple spatial scales by incorporating three interdisciplinary tools: remote sensing, GIS, and multivariate vegetation analyses. Remote sensing techniques are used on a hyperspectral/high resolution image to characterize the current range of riparian habitat conditions at the watershed scale and create a streamside forest cover map. This technique provides a cost effective means of mapping the distribution of streamside forest cover at an accuracy of > 75% when compared with field observations. Using the riparian forest cover layer derived from the imagery, a habitat suitability model is then developed using GIS to identify a suite of potentially suitable sites for conifer restoration based on the quality of salmonid habitat as defined by stream channel gradient, confinement, and streamside vegetation. Once these potential sites have been identified, a multivariate statistical approach is used to examine the spatial distribution of riparian plant communities as they relate to alluvial landform types. This portion of the research identifies microscale landforms and site conditions that would have the greatest probability of supporting conifer retention long term.

Key words: Riparian, GIS, Remote Sensing, Restoration