Search

Laboratory experiments were used to investigate the influence of simulated cracks and roots on soil water repellency (SWR) dynamics with and without basal drainage impedance in wetting–drying cycles. Observations and measurements were taken following water application equivalent to 9.2-mm rainfall and then periodically during 80 h of drying. In total, 180 experiments were carried out using 60 samples of three homogeneous, reconstituted soils with different organic matter contents and textures, but of similar initial severity of SWR [18% molarity of an ethanol droplet (MED)]. Water flowing down the cracks and roots left the soil matrix largely dry and water repellent except for vertical zones adjacent to them and a shallow surface layer. A hydrophilic shallow basal layer was produced in experiments where basal drainage was impeded. During drying, changes in SWR were largely confined to the zones that had been wetted. Soil that had remained dry retained the initial severity of SWR, while wetted soil re-established either the same or slightly lower severity of SWR. In organic-rich soil, the scale of recovery to pre-wetting MED levels was much higher, perhaps associated with temporarily raised levels (up to 36% MED) of SWR recorded during drying of these soils. With all three soils, the re-establishment of the original SWR level was less widespread for surface than subsurface soil and with impeded than unimpeded basal drainage. Key findings are that as follows: (1) with unimpeded basal drainage, the soils remained at pre-wetting repellency levels except for a wettable thin surface layer and zones close to roots and cracks, (2) basal drainage impedance produced hydrophilic basal and surface layers, (3) thorough wetting delayed a return to water-repellent conditions on drying, and (4) temporarily enhanced SWR occurred in organic-rich soils at intermediate moisture levels during drying. Hydrological implications are discussed, and the roles of cracks and roots are placed into context with other influences on preferential flow and SWR under field conditions. Copyright © 2014 John Wiley & Sons, Ltd.

The conventional view of soil water repellency is that it promotes overland flow and soil erosion, but this is not always borne out by observations. This study aimed to isolate the effects of repellency on long-unburnt and recently burnt terrain on infiltration, overland flow and erosion at the small plot scale (0·36 m2). Rainfall simulations (30 min; intensity 100 mm h−1), using untreated water, and water treated with surfactants to eliminate repellency, were conducted on in situ repellent soils in fire-prone Eucalyptus globulus plantations, north-central Portugal at (i) a long-unburnt site with and without litter, and (ii) a recently burnt site. On long-unburnt terrain, the mean overland flow coefficient (33%) was 16 times higher and mean slopewash was 23 times higher under repellent compared with wettable conditions. On recently burnt terrain, no overland flow was recorded under wettable conditions, while under repellent conditions the mean coefficient was 70%. The water storage capacity of the litter layer under 10-year-old eucalyptus stands for dry antecedent conditions was at least 3 mm water depth per cm litter depth, implying at least a delay to the onset of overland flow. Severe repellency (36% ethanol) was found to persist through a 30-min storm (100 mm h−1) when a litter layer was present. A continuous wetting front was observed in the upper ∼1 cm of exposed soil, indicating a breakdown in repellency at the time of observation. Below ∼1 cm, repellent, dry soil conditions generally persisted through the simulated storm event. A major implication is that prediction of hydrological impacts of repellency must also take into account the infiltration characteristics of any litter layer and any non-repellent soils, if present. Copyright © 2007 John Wiley & Sons, Ltd.

event, yielding clockwise and anticlockwise hysteretic loops respectively. In contrast, metal concentrations, although high throughout the winter storm event, were diluted somewhat during the storm peak itself. The Pb and Cu appear to be supplied by quickflow processes and transported in close association with fine sediment, whereas Zn and Cd are delivered to the channel and lake by slower subsurface seepage in dissolved form. In the winter 2001 event, antecedent soil moisture and shallow groundwater levels were anomalously high and seepage sources of dissolved metals dominated. Downstream of the lake, Pb and Cu levels and suspended sediment were high in the summer storm, but low in the winter storm, suggesting retention with deposition of fine sediment in the lake during the latter. In the winter storm, Zn and Cd levels were higher downstream than upstream of the lake, perhaps because of additional seepage inputs from the surrounding slopes, which failed to have an impact during summer. An understanding of the complex interplay of antecedent soil moisture and the dynamics of subsurface seepage pathways in relation to the three-dimensional distribution of sources is important in modelling heavy metal fluxes and levels in contaminated urban catchments. Copyright  2003 John Wiley & Sons, Ltd.

Small bounded plots were used to investigate dissolved nutrient losses in overland flow in eucalyptus and pine forests in northern Portugal following understorey fires. Losses of NON, POCa2+, Mg2+ and K+ were measured over 19 months during the first 2-3 years after fire. Solute losses in overland flow increased after fire in eucalyptus and pine terrain due to the mineralization of litter and vegetation and increased overland flow. The data suggest that (i) elevated losses persist for at least 2 and 3 years at the pine and eucalyptus sites respectively and (ii) soluble losses of POK+ in a post-fire cycle may exceed those adsorbed to eroded material. Losses of POalthough relatively small, are potentially the most detrimental to soil fertility and forest productivity because of the limited opportunity for P replenishment and correspondingly high P losses adsorbed to eroded sediment. Copyright © 2000 John Wiley & Sons, Ltd.