Episodic sediment delivery and landscape connectivity in the Mancos Shale badlands and Fremont River system, Utah, USA

The Fremont River drains about 1000 km 2 of Mancos Shale badlands, which provide a large percentage of the total sediment load of its middle and lower reaches. Factors controlling sediment movement include: weathering that produces thin paralithic soils, mass movement events that move the soil onto locations susceptible to fluvial transport, intense precipitation events that move the sediment along rills and across local pediments, and finally Fremont River floods that move the sediment to the main-stem Colorado River. A forty-year erosion-pin study has shown that down-slope creep moves the weathered shale crust an average of 5.9 cm/yr. Weather records and our monitoring show that wet winters add large slab failures and mudflows. Recent sediment-trap studies show that about 95% of sediment movement across pediments is accomplished by high-intensity summer convective storms. Between 1890 and 1910, a series of large autumn floods swept down the Fremont River, eroding its floodplain and transforming it from a narrow and meandering channel to a broad, braided one. Beginning about 1940, the Fremont's channel began to narrow. Sequential aerial photos and cross-sections suggest that floodplain construction since about 1966 has stored about 4000 to 8000 m 3 of sediment per kilometer per year. These data suggest that it will take two centuries to restore the floodplain to its pre-1890 condition, which is in line with geologic studies elsewhere on the Colorado Plateau. The various landscape elements of slope, pediment, and floodplain are semi-independent actors in sediment delivery, each with its own style. Accelerated mass movement on the slopes has an approximate 20-year recurrence. Sediment movement from slope across pediments to master stream is episodic and recurs more frequently. The slope-to-pediment portion of the system appears well connected. However, sediment transport through the floodplain is not well connected in the decadal time scale, but increases in the century and millennial time scales, and changes over time depending on the cycle of arroyo cutting and filling.