LandSoil: A model for analysing the impact of erosion on agricultural landscape evolution R. Ciampalini a, ⁎, S. Follain b , Y. Le Bissonnais a a INRA, UMR – LISAH, “Laboratoire d'étude des Interactions Sol - Agrosystème – Hydrosystème”, INRA – IRD – SupAgro, Bat. 24. 2, place Viala – 34060, Montpellier, France b Montpellier SupAgro, UMR – LISAH, “Laboratoire d'étude des Interactions Sol - Agrosystème – Hydrosystème”, INRA – IRD – SupAgro, Bat. 24. 2, place Viala – 34060, Montpellier, France abstract article info Article history: Received 4 November 2010 Received in revised form 16 April 2012 Accepted 16 June 2012 Available online 23 June 2012 Keywords: Quantitative modelling Landscape modelling Landscape scale Medium-term soil redistribution Soil erosion Tillage erosion The purpose of this research was to develop a model for the analysis of agricultural landscape evolution at a fine spatial resolution scale (1–10m) and a medium-term temporal scale (10–100 years). The model (LandSoil: land- scape design for soil conservation under Land use and climate change) works at a catchment scale and it is based on the STREAM soil erosion model. It is spatially distributed, event-based, and considers interrill, rill and tillage erosion as the main processes acting in soil redistribution. The specificities of the model include the use of a de- tailed representation of the agricultural landscape through parameters such as soil surface properties, hydrologic pathways and considering a climate component based directly on rainfall events. In this paper, we present the characteristics of the model and its application to a Mediterranean study area. The model was subjected to a calibration/validation procedure at two different spatial scales (field and catchment) with a medium-term data series of runoff and sediment concentration measurements. A num- ber of medium-term field and catchment simulations were carried out, allowing us to observe landscape evolution under recent and actual agricultural practices and to formulate hypotheses based on changes re- lated to different agricultural patterns and soil uses. Modelling at a field scale using rill, interrill and tillage erosion simulations on reconstructed prior topographic surfaces showed a relationship between observed and simulated topography with a prediction error b15% and a correlation coefficient (r) of 0.605. A catch- ment scale analysis of a degraded scenario, maximising hillslope runoff with the removal of all the grass strips between fields, indicated a global increase of the soil erosion rate (+29%), with spatial variability depending on the specific soil use type. © 2012 Elsevier B.V. All rights reserved. 1. Introduction Soil cover is currently undergoing rapid evolution, due mostly to changes produced by climate and humans. Relevant human-induced changes include those implemented by farmers individually at field- or farm-scales as well as those imposed by policy- and decision-makers (land planners and natural resource managers) in a range of spatial scales (from farm plot to administrative divisions) (Verburg et al., 2002; Rounsevell et al., 2005; Claessens et al., 2009). Climate-induced changes are those related to changes in the seasonal distribution of cli- mate factors and in the frequency of extreme events predicted by pro- jections of future climate change (IPCC Core Writing Team, 2007). All of these rapid modifications may have considerable consequences for cultivated ecosystem productivity and indicate the need for wider usage of technologies and landscape management practices to conserve soil. Thus, modelling is an appropriate methodology for analysing, un- derstanding and simulating space-time evolution of topography and sediment generation. Such modelling presumes the ability to produce quantitative models of soil redistribution in the landscape to test the ef- fects of different scenarios of land management and climate evolution. Among available soil redistribution models, it is possible to distinguish between continuous process-based models dimensioned for large spa- tial and temporal scales and event-based small catchment scale erosion models. Coulthard (2001) and Willgoose (2005) have listed a few recent DEM-based models that can be used to study the evolution of agricul- tural environments in the medium-term perspective. In these models, a physical description of water and sediment fluxes is used to repre- sent the processes and the evolution of terrain physiography at geo- logical time scales (10 4 –10 6 year). The cumulative effects of erosion and sediment yield are considered key to understanding landscape evolution (e.g., Willgoose et al., 1991; Schoorl et al., 2002). These models provide an evaluation of soil redistribution that uses mecha- nistic rules for diffusion and water erosion similar to those proposed by Kirkby (1985), and some of them also include a soil production function (Heimsath et al., 1997; Minasny and McBratney, 1999, 2001). In contrast, event-based catchment erosion models are developed at fine scales and describe the catchment with a combination of Geomorphology 175–176 (2012) 25–37 ⁎ Corresponding author. Tel.: +33 499 612 161; fax: +33 467 632 614. E-mail address: rossano.ciampalini@gmail.com (R. Ciampalini). 0169-555X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2012.06.014 Contents lists available at SciVerse ScienceDirect Geomorphology journal homepage: www.elsevier.com/locate/geomorph