ภาพ
Fig. 4. Schematic of a pan lysimeter designed to electronically record the rate of percolate accumulation.
during the summer that produced percolate from no-till columns often did not produce percolate from tilled columns. Hall et al. (1989) also noted that percolate accumulated faster and more frequently in pan lysimeters in a no-till Typic Hapludalf (FAO – Haplic Luvisol) than when this soil was tilled. Since transmission of water during rainfall is characteristic of macropore flow (Thomas and Phillips, 1979), the implication of these results is that disruption of macropore continuity by tillage reduced the contribution of macropores to the total flow. To confirm these observations, Shipitalo and Edwards (1993b) devised a method to electronically record the rate of percolate accumulation in pan lysimeters installed 50 cm deep in the field (Fig. 4). The response observed for a storm on 11 July 1993 that produced 27.4 mm of rain was typical of that observed for high intensity thunderstorms (Fig. 5).
Percolate accumulation began 1 h after the start of the storm and shortly after the rainfall intensity began to increase rapidly. After the storm was over percolate accumulation was negligible. In total, the pan lysimeter captured 17% of the rainfall. Based on antecedent soil water content and water holding capacity,
however, the soil above the pan should have been able to retain all the rainfall. Thus, the water that was
collected in the pan lysimeter must have bypassed the matrix.
3.2.2. Flow in earthworm burrows
To determine if earthworm-formed macropores could contribute to bypass flow,Edwards et al. (1989) developed a method to monitor the flow in individual ≥5 mm diameter burrows by inserting a tight-fitting tube into the base of holes intercepted from the sidewall of an open pit. The tube led to a