Two drains were installed in a poorly drained field using a drainage-mole. Drain sections were either (1) filled with surface soil (method F) to recreate a textural continuity from drain to surface, or (2) let opened as growers usually do (method C) allowing the soil to almost reinstate its profile. A complementary treatment was to break the hardpan at selected places. Water-level loggers were launched in 24 piezometers installed at two depths over 12 points. All sampling points (24) were equipped with two tensiometers (two depths). The set-up was balanced to collect data close to the drain, at ¼-¾ drain spacing, and at ½ drain spacing. Infiltration rates and hydraulic gradients were compared using R. Then HYDRUS 2D/3D was used to provide additional information on involved mechanisms.
Results show that drainage occurred. Infiltration velocities calculated from surface-piezometer hydrographs did not show significant differences between methods. The complementary treatment did not add any significant effect on infiltration velocities. However, variations in deep-piezometer hydrographs suggested a water table refill or an increase in the soil water level that appeared higher for method C. This may indicate that method F would be more efficient to maintain a low water level between drains. The relative drained volumes must be calculated from deep-piezometer hydrographs to statistically compare between methods. Attempting to improve drainage efficiency in organic soils appeared to involve complex mechanisms specific to these soils, where challenges remain for the current and future researches.