Comparison of slope profiles in areas exhibiting widespread bedrock landsliding with the use of a model for the maximum size of stable hillslopes established that mountain-scale material strength can limit topographic relief. Conventional laboratory values for intact rock greatly exceeded integrative rock strength properties that were back-calculated from the upper limit to hillslope relief and gradient in the northern Cascade Range and Santa Cruz Mountains. Back-calculated strength values, however, were indistinguishable from those obtained through field and conventional laboratory measurements on the weakest members of each rock formation, as well as on glacial sediments along the Cascade front. These results contrast with the conventional assumption that relief is incision-limited and indicate that the relief of mountain ranges can reflect landscape-scale material strength, as well as the interaction of tectonic and climatic processes.

The influence of root reinforcement on shallow landsliding has been well established through mechanistic and empirical studies, yet few studies have examined how local vegetative patterns influence slope stability. Because root networks spread outward from trees, the species, size, and spacing of trees should influence the spatial distribution of root strength. We documented the distribution and characteristics of trees adjacent to 32 shallow landslides that occurred during 1996 in the Oregon Coast Range. Although broadly classified as a conifer-dominated forest, we observed sparse coniferous and abundant hardwood trees near landslide scars in an industrial forest (Mapleton) that experienced widespread burning in the 19th century. In industrial forests that were burned, selectively harvested, and not replanted (Elliott State Forest), swordfern was ubiquitous near landslides, and we observed similar numbers of live conifer and hardwood trees proximal to landslide scarps. We demonstrate that root strength quantified in landslide scarps and soil pits correlates with a geometry-based index of root network contribution derived from mapping the size, species, condition, and spacing of local trees, indicating that root strength can be predicted by mapping the distribution and characteristics of trees on potentially unstable slopes. In our study sites, landslides tend to occur in areas of reduced root strength, suggesting that to make site-specific predictions of landslide occurrence slope stability analyses must account for the diversity and distribution of vegetation in potentially unstable terrain.Key words: slope stability, vegetation, root strength, shallow landslide, debris flow, Oregon Coast Range.