4.2 Soil water pathways and generation of runoffs
During the intense drainage period, a saturated zone is observed 75% of the time, 30 cm below the soil surface or higher. The quasi-permanent saturation of the soil leads to a low infiltration capacity and a continuous water pressure above tile drains. Before event A, most the entire soil profile was already saturated but the depth of the saturated zone decreased during the event to reach less than 1.6 cm. The closer to the soil surface the saturated zone is, the higher the subsurface runoff flow is: it suggests subsurface discharge variations are caused by variations of the water pressure above the drain. Moreover, as shown by water tracing, subsurface runoff results from the mixing of both rainwater and soil water. Such a transfer had also been observed by (Coulomb & Dever, 1994) at the beginning of the intense drainage season but not in the middle of the drainage season. The authors explain this difference by the saturation of macropores during the drainage season. Our results suggest that macroporosity was still active during the studied event of the intense drainage period. Thus, a preferential flow through the macropores is possible even when the soil is saturated. Concerning the surface runoff, surface drainage may be caused either by a refusal of infiltration or by a transfer of water from the first soil centimeters to the SDRs. Isotope water tracing shows that surface runoff is a mix of both soil water and rainwater. The presence of soil water in the surface runoff may be due to the increase of the saturated zone level and lateral subsurface flows. Following isoproturon transfer, Haria et al. (1994) gave the same explanation and specified that a lateral flow circulated over an impeding layer. The presence of such a lateral flow is therefore also possible in our case. The lateral flow would then flow directly into the ditch at the edge of the field and could constitute a significant but unquantified part of the water balance.
During the low drainage period, soil moisture is constantly low. However, the studied event shows that a sufficiently intense event can generate surface and subsurface runoff. For such events, subsurface runoff occurs before surface runoff, proving that the process of rainwater transfer to subsurface drains is faster than the generation of surface runoff. Two reasons can be suggested: (i) the presence of the wheat cover that slows the surface runoff and (ii) a preferential vertical flow through macropores. Isotopic measurements support the hypothesis of preferential flow through the macroporosity but our results do not support any conclusion about the soil cover effect on the surface runoff. During the studied event, surface runoff is only composed of rainwater. The similarity of the isotopic signatures of surface and subsurface runoffs suggest the subsurface runoff is also only composed of rainwater. Moreover, cracks present at the soil surface before the event constitute a macroporosity that can support a direct water transfer from soil surface to tile drains. This result is consistent with the conclusions of the dry and isotopic tracers based studies conducted by (Øygarden et al., 1997) and (Klaus et al., 2013) in fields with similar soils. Moreover, these authors underlined the role of the macropores distribution and connectivity for the water transfer to tile drains.