Do different methods for measuring the hydrophobicity of soil aggregates give the same trends in soil amended with residue?

Most soils have a low level of water repellency and several studies have shown this to be a major property underlying soil structural stability. Various approaches have been developed to measure low levels of repellency, but selecting the one most appropriate is hampered by the following: (i) different methods have been used in separate studies and have not been compared to date; (ii) they rely on different theoretical backgrounds; and (iii) they are technically different. The objective of this paper was to compare three methods that allow measurements of low levels of water repellency on soil aggregates. The chosen techniques include water drop penetration time (WDPT), capillary rise method (CRM) and the water repellency index (R) tests. Soil macroaggregates (3–5 mm) from silty soil were used in the study after addition of several rates of maize residue and different incubation times in order to achieve a wide range of organic matter contents, microbial abundances and water repellency levels. The water repellency results from all used techniques were in general agreement. Using the well established concepts we developed the theoretical relationships existing between WDPT, R index and CRM, which are in agreement with the experimental data. The methods do not determine the same soil property, however, as WDPT estimates the breakdown of repellency with time, whereas CRM and R index estimate initial soil hydrophobicity upon wetting. We propose mechanistic and practical arguments to select the appropriate method to measure soil repellency.

Research Highlights►Three independent methods, WDPT, R and CRM, showed the same tendency in characterizing water repellency. They use the same physical principle, but they measure different related soil properties. In soils with small levels of water repellency, WDPT determines the rate of wetting by water, whereas the R index and CRM measure hydrophobicity. ►Using a silty soil sample set characterized by a wide range of samples from completely wettable to subcritical water repellency, the results of three methods were in general agreement. Based on the concepts of Philip (1957) and Tillman et al. (1989), we developed theoretical relationships between WDPT, R index and CRM that are coherent with experimental results. We suggest using WDPT for field survey applications since it is simple, cheap and discriminates subcritical water repellency above thresholds that have practical relevance. However, it is not appropriate from a mechanistic point of view since it does not separate between changes in the pore system and changes in hydrophobicity between treatments. ►R index and CRM are robust, sensitive and repetitive techniques that may be used to measure hydrophobicity independently of the impact of soil pores. CRM and R techniques have some similarities as both use non-polar liquids to remove the effect of porosity and are based on early-time transport properties. Using water as one of the liquids, both are restricted to advancing contact angles < 90°. Meanwhile, using different liquids, CRM can measure the solid surface free energy (Owens and Wendt, 1969) and can thereby calculate contact angles > 90°, extending its applicability. As CRM measures a bed of aggregates, it provides good repeatability, although the coefficient of variation was greater than tests on individual aggregates using the R index approach. Beds of aggregates present problems through surface abrasion and contacts that may influence results. Moreover, CRM is the most time consuming and expensive of the techniques tested. It also uses dangerous non-polar liquids (e.g. n-hexane). In comparison, the R index identified small levels of water repellency well, without problems of aggregate packing or abrasion. We conclude that CRM provides the best information and resolution for measuring water repellency, but that R index may be more suited to tests on individual aggregates. A modification of the two approaches, where CRM is applied to individual soil aggregates (Ramírez-Flores et al., 2008) was not developed at the time of this study, but should also provide a sensitive measurement technique that avoids problems with packed beds of aggregates.