Mapping erosion risk at the basin scale in a Mediterranean environment with opencast coal mines to target restoration actions

34 páginas, 9 figuras

River basin restoration and management is crucial for assuring the continued delivery of ecosystem services and for limiting potential hazards. Human activity, whether directly or indirectly, can induce erosion processes and drastically change the landscape and alter vital ecological functions. Mapping erosion risk before future restoration-management projects will help to reveal the priority areas and develop a hierarchy ordered according to need. For this purpose, we used the Revised Universal Soil Loss Equation (RUSLE) erosion model. We also applied a novel technique called GPVI (Genetic Programming Vegetation Index) in the Martín River basin in NE Spain (2,112 km2), which has a large coalfield located in the southern part of the basin. Approximately two-thirds (69%) of the area of the Martín basin presents low and medium soil loss rates, and one-third (31%) of the area presents high (18%), very high (10%), and irreversible (3%) erosion rates. The southern part of the basin is the most degraded and is strongly influenced by the topography. This work allows us to locate areas prone to erosional degradation processes to help create a buffer around the river and locate “spots” in need of restoration. We also checked the error estimation of the methodology because our soil maps do not include rock and bare rock areas. The usefulness of applying RUSLE for predicting degraded areas and the consequent directing of soil conservation–restoration actions at the basin scale is demonstrated. We highly recommend a field survey of the selected areas to prove the goodness of the model estimations.

This work is part of the research and assistance agreement between Endesa S.A. and CSICPyrenean Institute of Ecology (IPE-CSIC). Funding for this study was provided by Endesa S.A. A special acknowledge is given to Endesa Centro Minero Andorra (Teruel). Thanks are given to, J. M. Garcia Ruiz, S. Begueria, E. Nadal, E. Moran-Tejera, and J.J. Jimenez for reviewing and general advises during the development of this work, M. P. Errea, J. Zabalza, L. C. Alatorre for assistance with GIS analysis, M. Angulo for R factor map, M. Pazos with statistical analysis, and F. Reverberi for laboratory work. M. Trabucchi was in receipt of grant from JAE-CSIC (Ref. I3P-BPD-2006).

Peer reviewed