How does an old landscape learn new tricks? Exploring topographic disequilibrium and stream-soil coupling in the critical zone

Abstract:

How do deeply weathered, soil mantled hillslopes in the tropics respond to sudden changes in the erosion rate? Do they re-adjust gradually, or in flashy, threshold driven pulses? One method for addressing this is to examine the soil profile, by asking the question: is the rate of production of soil in equilibrium with the rate of erosion of topsoil? We measure in-situ produced cosmogenic ¹⁰Be in the topsoil, at the soil-saprolite interface, and at several intermediate depth intervals to obtain a quantitative measure of soil mantle disequilibrium in a landscape responding to a rapid change in erosion rate in the Luquillo Critical Zone Observatory, Puerto Rico. Geomorphic features in the tropical Luquillo Mountains indicate a well-defined elevation boundary separating a quickly eroding landscape below from a slowly eroding ‘relict’ landscape above. A high-energy knickzone extends from the floodplain to ~600 m elevation on all of the channels draining the ridgelines, but above this elevation streams are sandy and low gradient. Catchment wide erosion rates measured upstream and downstream of the 600 m knickpoint using cosmogenic ¹⁰Be confirm that the knickzone defines a sudden change in the rate at which materials are exported. This wave of enhanced erosion triggered by the knickzone migration provides a natural experiment to study the change in morphology and response of soils to a change in rate over time. Ridgelines in the relict zone above 600 m are typically broad with deep, nutrient depleted soil profiles, while ridgelines below the knickzone are sharp, narrow features with shallow, poorly developed soil horizons in the adjusted portion of the landscape. Our study targets ridgetops above and below 600 m that appear to be narrowing in response to the backward propagation of the knickzone. We provide a spatially detailed description of ¹⁰Be concentrations with depth in soil profiles and use the deviation from the production curve for in-situ ¹⁰Be under site conditions to quantitatively describe the response of the soil profile to landscape scale changes in erosion rate.