Elsevier

Journal of Hydrology

Volume 393, Issues 1–2, 27 October 2010, Pages 105-113
Journal of Hydrology

Investigations of soil cracking and preferential flow in a weighing lysimeter filled with cracking clay soil

https://doi.org/10.1016/j.jhydrol.2010.03.007Get rights and content

Summary

An improved understanding of deep drainage processes in irrigated cracking soils is needed for sustainable irrigation management. To investigate the effect of crack dynamics and macropore flow on drainage in cracking soils, a series of irrigation experiments was carried out in a weighing lysimeter. Subsurface soil cracks of the initially very dry soil were investigated with a videoscope and changes in the surface expression of cracks in response to irrigation events were monitored by time-lapse photography. A bromide tracer was applied to one irrigation event. Variations in the combined soil and moisture mass and the volume of drainage out of the soil column was logged and the drainage EC and bromide content were determined. No drainage occurred out of the soil column during the first 3 out of 6 irrigation events, even though substantial surface runoff into the cracks occurred and, at least initially, soil cracks provided an uninterrupted flow path through the profile. The breakthrough of the bromide tracer, as well as an initially low EC of the drainage water indicate that preferential flow accounted for a substantial part of the first of the two drainage events, even though the soil cracks were sealed on the surface at the onset of the irrigation causing the drainage. The results show that lateral infiltration of macropore flow into the soil matrix can be substantial and should not be neglected while simulating macropore flow and deep drainage in cracking soils. The results also indicate that soil cracks can remain pathways for preferential flow even after they are closed at the soil surface. The type of water application appears to have an impact on the location of crack formation, with flood irrigation favouring reappearance of cracks at previous crack locations and simulated rainfall resulting in shifting crack locations.

Introduction

Deep drainage (DD) of irrigation water can impose a variety of effects on irrigation management. Irrigation water that becomes DD is lost for the growing plants and therefore reduces the application efficiency (Smith et al., 2005). The occurrence of DD potentially increases leaching of agrochemicals, such as various forms of Nitrogen, which constitutes a loss of nutrients from the farming system and potentially degrades the quality of receiving aquifers and surface waters (Bronswijk et al., 1995, Silburn and Montgomery, 2001). DD reaching the saturated groundwater zone can lead to rising water tables, which, once the capillary fringe of the ground water gets close to the surface and starts to evaporate, can severely increase soil salinity (Jorenush and Sepaskhah, 2003). Worldwide ∼1.5 million hectares of irrigated agricultural land is lost to salinity each year (Foley et al., 2005). At the same time, some DD can be beneficial in salinity management as it drains accumulated salts out of the top soil (Vervoort et al., 2003).

Fine grained cracking clay soils are prevalent in irrigation agriculture in many parts of the world. Due to their high nutrient content and water holding capacity they are highly productive soils. Because of their high clay content and thus low permeability, DD of irrigation water in these soils was long assumed minimal (Hearn, 1998, Vervoort et al., 2003). However, in recent years evidence of DD in cracking clay soils became widely accepted, yet estimates of its magnitude still vary greatly (Silburn and Montgomery, 2001, Smith et al., 2005, Acworth and Timms, 2009). For instance in a field investigation detected almost instantaneous arrival of irrigation water to a depth of 1.2 m as a response to flood irrigation on cracked soil.

Irrigation literature provides different management recommendations for irrigation and DD management on cracking soils. Mitchell and van Genuchten (1993) highlighted the advantage of flood irrigation, to make use of the high infiltration rates before crack closure, while Chen et al. (2002) recommended the use of low intensity sprinkler irrigation to avoid DD and leaching through soil cracks. Smith et al. (2005) emphasized that DD could be controlled by ensuring that the irrigation applications do not exceed the soil moisture deficit and Bethune (2004) recommended drying out fields after the irrigation season to prevent that water added by winter rains exceeds the soil moisture deficit and causes deep drainage. However, shrinkage cracks that form during drying of a soil could also favour the occurrence of DD due to the high importance of preferential flow through cracks in cracking clay soils (Bronswijk, 1988). It is known that preferential flow through soil cracks can cause rapid transport of irrigation water, solutes and agrochemicals through the unsaturated zone (Kosmas et al., 1991, Harris et al., 1994, Bronswijk et al., 1995, Lin and McInnes, 1995, Weaver et al., 2005), hence lowering the soils capacity of storage, adsorption and transformation of potential pollutants and greatly reducing irrigation efficiency.

To optimize irrigation management an improved understanding of the hydrology of these cracking clay soils is needed. In this study the process of soil crack formation and preferential flow was investigated in detail in a cracking clay soil in a weighing lysimeter. This controlled environment enabled the study of preferential flow and drainage in a highly cracked soil column under simulated strong rainfall and flood irrigations to determine the relationship between crack dynamics, bulk soil moisture content and macropore flow.

Section snippets

Methods and materials

The presented experiment took place over a time period of ∼5 years and consisted of three major parts, which are described in detail below. The investigated soil column was set up between September 2002 and February 2003. After a drying period of 30 month, a first set of measurements was carried out in 2005, followed by a further 20 month of soil drying and a second set of measurements in 2007.

Results

In November 2005, all cracks that were large enough to fit the 6 mm fibreglass tube of the videoscope were found to penetrate the entire soil profile of 450 mm. At two out of 20 insertion locations the lower 50 mm of the crack were filled with loose soil aggregates, presumably fallen in from above. The angle in which the cracks extended downwards, altered with up to ±15° from the vertical. In addition to the vertical crack network, near horizontal shrinkage planes were observed that intersected

Discussion

Based on the results in this study there are two interesting observations to be noted in regards to flow dynamics in cracking soils: Firstly, that drainage out of the soil column was not observed during the first three water applications and secondly that the drainage water that was collected after application 4 did show signs of preferential flow. As the observed crack dynamics play an important part in explaining these processes, they will be discussed before the flow dynamics are examined.

Conclusions

Crack dynamics and preferential flow were investigated during six irrigation events on an initially very dry and cracked soil in a weighing lysimeter. No drainage occurred out of the soil column during the first 3 out of 6 irrigation events, even though substantial surface runoff into the cracks occurred and, at least initially, soil cracks provided an uninterrupted flow path through the entire soil profile. This shows that lateral infiltration of macropore flow into the soil matrix was

Acknowledgments

This study was funded by the Cotton Catchment Communities CRC and the Cotton Research and Development Cooperation. The lysimeter was set up by Dr. Justin Bell.

References (33)

  • R. Acworth et al.

    Evidence for connected water processes through smectite-dominated clays at Breeza, New South Wales

    Australian Journal of Earth Sciences

    (2009)
  • C. Appelo et al.

    Geochemistry, Groundwater and Pollution

    (2005)
  • M. Bethune

    Towards effective control of deep drainage under border-check irrigated pasture in the Murray–Darling basin: a review

    Australian Journal of Agricultural Research

    (2004)
  • H. Booltink et al.

    Sensitivity analysis on processes affecting bypass flow

    Hydrological Processes

    (1993)
  • J. Bouma et al.

    Characterizing soil water regimes in swelling clay soils

  • J. Foley et al.

    Global consequences of land use

    Science

    (2005)
  • Cited by (0)

    View full text