Effect of antecedent soil moisture on preferential flow in a texture-contrast soil Marcus A. Hardie a,e,c,⇑ , William E. Cotching b,e,c , Richard B. Doyle d , Greg Holz a , Shaun Lisson c , Kathrin Mattern d a Tasmanian Institute of Agricultural Research, University of Tasmania, PB 98, Hobart, Tasmania 7001, Australia b Tasmanian Institute of Agricultural Research, University of Tasmania, PO Box 3523 Burnie, Tasmania 7320, Australia c CSIRO Sustainable Ecosystems University of Tasmania, PB 98, Hobart, Tasmania 7001, Australia d School of Agricultural Science, University of Tasmania, PB 54 Hobart, Tasmania 7001, Australia e Secondment from Department of Primary Industries, Parks Water and Environment (DPIPWE), GPO Box 44, Hobart, Tasmania 7001, Australia article info Article history: Received 20 July 2009 Received in revised form 8 December 2010 Accepted 10 December 2010 Available online 24 December 2010 This manuscript was handled by P. Baveye, Editor-in-Chief Keywords: Duplex Dye tracer Finger flow Funnel flow Macropore flow Shrinkage crack summary The effect of soil moisture status on preferential flow in a texture-contrast soil was investigated by applying 25 mm Brilliant Blue dye tracer to soil profiles at high and low antecedent soil moisture. Differences in soil morphology and chemistry between soil profiles had little effect on the depth of dye infiltration and dye distribution down the profile. Antecedent soil moisture strongly influenced the type, depth and rate of dye tracer movement. In the wet treatment, the dye tracer infiltrated to depths between 0.24 and 0.40 m, at an average rate of 120 mm h À1 . Whilst in the dry treatment, the same volume of dye tracer infiltrated to between 0.85 and 1.19 m depth at an average rate of 1160 mm h À1 . In dry antecedent conditions, finger flow developed in the A1 horizon as a result of water repellency. In the wet treatment, the wetting front developed permutations but did not break into fingers. Despite similar particle size distributions, flow in the A2 e was slower than the A1 horizon, due to the absence of macropores. In the dry treatment, the dye tracer ponded on the upper surface of the B21 horizon, which then spilled down the sides of the large clay columns as rivulets, at rates of between 2000 and 3000 mm h À1 . The dye tracer accumulated at the base of the columns resulting in backfilling of the inter column shrinkage cracks, at an estimated rate of 750 mm h À1 . In the subsoil, water movement occurred via shrinkage cracks which resulted in flow by-passing 99% of the soil matrix in the B21 horizon and 94% of the soil matrix in the B22 horizon. Evidence of rapid and deep infiltration in ‘dry’ texture-contrast soils has implications for water and solute management. This knowledge could be used to: (i) improve irriga- tion and fertilizer efficiency (ii) explain variations in crop yield (iii) reduce salinity through improved leaching practices, (iv) reduce the risk of agrochemicals contaminating shallow groundwater. Ó 2011 Elsevier B.V. All rights reserved. 1. Introduction Preferential flow refers to all phenomena where water moves along preferred pathways through the soil profile allowing water to bypass part of the soil matrix. It allows water and solutes to move to greater depths, at faster rates, than predicted by the Richards equation for uniform flow (Hendrickx and Flury, 2001; Simunek and van Genuchten, 2007). Preferential flow is considered to be both common and widespread (Flury et al., 1994) resulting in either enhanced, or reduced capacity of the soil to buffer and filter potential contaminants (Clothier et al., 2008). Although the term preferential flow does not imply any particular mechanism, it usually refers to one or more of three physically distinct processes: macropore flow, finger flow, or funnel flow (Kung, 1993; Ogawa et al., 1999). Finger flow or fingering, results from air entrapment or water repellence causing an instability in the wetting front leading to the formation of thin fingers beneath a more uniformly wetted distribution zone (Glass et al., 1988). Finger flow is distinct from other forms of preferential flow in that it is a fluid pheno- menon, resulting from spatial heterogeneity in water repellence or soil moisture content rather than soil structure (Jury and Horton, 2004; Lemmnitz et al., 2008). Funnel flow, results from lateral redirection of infiltrating water resulting from changes in soil texture or lithographic boundaries (Kung, 1990). Macropore flow or bypass flow results from infiltrating water moving through void spaces in the soil. Macropores are formed in various ways including soil shrinkage, root growth, chemical weathering, cycles of freezing and thawing, or bioturbation (Beven and Germann, 1982). In texture-contrast soils, two classes of macropores are observed: (i) biopores, which are created by soil fauna or flora, resulting in cylindrical, semi-stable voids, and (ii) shrinkage cracks, formed by drying of swelling clays (McCoy et al., 1994). 0022-1694/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jhydrol.2010.12.008 ⇑ Corresponding author. Fax: +61 3 6226 1925. E-mail address: Marcus.Hardie@utas.edu.au (M.A. Hardie). Journal of Hydrology 398 (2011) 191–201 Contents lists available at ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol