Soil properties and susceptibility to preferential solute transport in tilled topsoil at the catchment scale A. Ghafoor, J. Koestel, M. Larsbo, J. Moeys, N. Jarvis / Department of Soil and Environment, Swedish University of Agricultural Sciences (SLU), Box 7014, 750 07 Uppsala, Sweden article info Article history: Received 5 June 2012 Received in revised form 26 March 2013 Accepted 28 March 2013 Available online 11 April 2013 This manuscript was handled by Peter K. Kitanidis, Editor-in-Chief Keywords: Macropores Breakthrough curve Hydraulic conductivity Solute transport Cultivated soil Organic carbon summary Preferential water flow and solute transport can have dramatic effects on the leaching of contaminants to groundwater and surface waters (via subsurface drainage) and is therefore of major concern to policy and decision-makers in the realm of water resources management. Unfortunately, we cannot measure these processes at the landscape scales that are relevant for management (farms, catchments, regions), which implies that an approach based on pedotransfer functions is needed to support model predictions. How- ever, the extent to which susceptibility to preferential solute transport can be predicted from proxy site and soil attributes that can be observed and mapped at the landscape scale is still largely unknown. We therefore carried out non-reactive solute breakthrough experiments on 45 topsoil columns sampled from the contrasting soil types found in a 13 km 2 agricultural catchment in Sweden. Non-parametric indicators of preferential solute transport were derived from the shapes of the solute breakthrough curves and related to soil physical and hydraulic properties measured in the same columns. The results showed that preferential transport was weakly (and negatively) correlated with the saturated macropore hydraulic conductivity. In contrast, it was much more strongly controlled by the size of the largest water-filled pore, which in turn was significantly correlated to the saturated hydraulic conductivity of the soil matrix and soil textural classes. Preferential transport was also weakly expressed in three fine-textured soils of large organic carbon content. We conclude that the spatial pattern of preferential transport across the studied catchment should show a clear deterministic component since it depended on soil properties (e.g. clay content) that are expressed relatively uniformly across larger areas of land. Ó 2013 Elsevier B.V. All rights reserved. 1. Introduction Water and dissolved solutes may move rapidly along certain high-conductance pathways in soils bypassing a large fraction of the porous matrix. This preferential transport can occur at the pore-scale in structural macropores (i.e. biopores, fissures and voids created by tillage implements; Jarvis, 2007) and in matrix- sized pores at the Darcy scale due to differences in hydraulic prop- erties resulting from spatial variation in texture, bulk density or water repellency (e.g. Kung, 1990; Ritsema et al., 1993). The fact that preferential flow and transport can have dramatic effects on contaminant leaching (Jarvis, 2007) has prompted the develop- ment of many models that can deal with these processes (e.g. Šim- u ˚ nek et al., 2003; Gerke, 2006). Model applications have mostly been restricted to column and small plot experiments at well- investigated sites, where input parameters can be derived by a combination of direct measurements and calibration (Köhne et al., 2009a,b; Jarvis and Larsbo, 2012). However, public authori- ties need models and decision-support tools that can account for the effects of macropore flow on leaching at the much larger scales (e.g. farms, catchments or even regions) that are relevant for man- agement (Vanclooster et al., 2004). Models must then be used to make predictions without direct measurements of input parame- ters or site data for calibration. An important unresolved question is whether this can be done with acceptable uncertainty. Evidence from measurements of saturated hydraulic conductivity suggests that macropore networks are often highly variable and spatially uncorrelated at short distances (e.g. Lauren et al., 1988; Mallants et al., 1996). Indeed, it has been speculated as to whether macro- pore flow must be considered as an essentially unpredictable pro- cess (Beven, 1991). Despite the complexity of soil pore networks, application of modern non-invasive X-ray imaging techniques (Wildenschild et al., 2002) has enabled some significant progress towards predic- tion of solute transport at the column scale. For example, various metrics of soil macropore geometry and topology have been com- pared statistically with transport characteristics measured in the same columns (e.g. Luo et al., 2010a), while pore network models re-constructed from 3D X-ray images can support direct predic- tions of solute transport (e.g. Köhne et al., 2011). However, from a pragmatic management point of view, the geometry and topol- 0022-1694/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhydrol.2013.03.046 / Corresponding author. Tel.: +46 18 672465; fax: +46 18 672795. E-mail address: nicholas.jarvis@slu.se (N. Jarvis). Journal of Hydrology 492 (2013) 190–199 Contents lists available at SciVerse ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol