Abstract
Topographic evidence requires that some rivers actively meander in bedrock, yet the way in which rivers can erode laterally and meander within bedrock banks is not well understood. Lateral channel migration, and especially lateral channel migration via active bedrock meandering, is commonly responsible for the preservation of unpaired strath terraces. A process-level understanding of lateral channel migration and active meandering in bedrock rivers is key to interpreting the climatic and tectonic significance of unpaired strath terraces and the planform shape of bedrock rivers. In this study, we compare erosional processes in two adjacent bedrock channels in the Santa Cruz Mountains, California. The main differences between these channels are that Pescadero Creek actively meanders within mudstone, while Butano Creek is straight and incises sandstone. Laboratory rock strength and slake durability tests show that while the two lithologies have similar tensile strengths before drying, the meander-supporting mudstone loses strength dramatically when dried and rewet (slakes), while the sandstone does not. The slaked mudstone bank rock was easily detached without the need for bed-load tools during in situ erosion tests, while mudstone that had not dried and sandstone were not detachable. The depth of bank rock detached solely from rewetting of previously dried mudstone ranges between 1 and 8 mm, which is well in excess of annual background erosion in the Santa Cruz Mountains. In addition, hand samples of the mudstone rapidly disintegrated upon wetting and drying in the laboratory, whereas sandstone hand samples remained intact. In the meandering stream, there is a consistent pattern of scoured bedrock (exposed to drying and slaking) along the outside “cutbank” of meander bends and forced bars that grade into soil and vegetation, which protect the bedrock from slaking along the inside of bends. Additionally, in the meandering stream, subaerially exposed mudstone clasts are often found to be disintegrating on the surface of bars. Taken together, these observations suggest that slaking allows for bedrock meandering in two fundamental ways. First, by rapidly disintegrating coarse hillslope-derived sediment that is deposited in the channel, slaking suppresses the negative feedback on lateral channel migration that would otherwise result from the buildup of talus along a retreating bedrock valley wall on the outside of a meander bend. Second, at cutbanks where scour exposes bare bedrock to drying, slaking weakens a layer of bank rock to the point where it can be eroded by clear-water flows. In these ways, slaking enables erosion into bedrock banks in response to curvature-driven fluid shear stress perturbations, as in alluvial rivers.