Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4725824 | Earth-Science Reviews | 2013 | 17 Pages |
Soil erosion is a key driver of land degradation and heavily affects sustainable land management in various environments worldwide. An appropriate quantification of rates of soil erosion and a localization of hotspots are therefore critical, as sediment loss has been demonstrated to have drastic consequences on soil productivity and fertility. A consistent body of evidence also exists for a causal linkage between global changes and the temporal frequency and magnitude of erosion, and thus calls for an improved understanding of dynamics and rates of soil erosion for an appropriate management of landscapes and for the planning of preventive or countermeasures.Conventional measurement techniques to infer erosion rates are limited in their temporal resolution or extent. Long-term erosion rates in larger basins have been analyzed with cosmogenic nuclides, but with lower spatial and limited temporal resolutions, thus limiting the possibility to infer micro-geomorphic and climatic controls on the timing, amount and localization of erosion. If based on exposed tree roots, rates of erosion can be inferred with up to seasonal resolution, over decades to centuries of the past and for larger surfaces with homogenous hydrological response units. Root-based erosion rates, thus, constitute a valuable alternative to empirical or physically-based approaches, especially in ungauged basins, but will be controlled by individual or a few extreme events, so that average annual rates of erosion might be highly skewed. In this contribution, we review the contribution made by this biomarker to the understanding of erosion processes and related landform evolution. We report on recent progress in root-based erosion research, illustrate possibilities, caveats and limitations of reconstructed rates, and conclude with a call for further research on various aspects of root–erosion research and for work in new geographic regions.