between aggregates and particles. Upon soil drying, cracking easily
occurred. As a result, the Dc and AW of cracks for the two paddy soils
increased significantly with increasing puddling intensity for both the
soils (Fig. 5, Table 4). This increased cracking mostly resulted from the
increasing clay-sized aggregates induced by puddling, and good regression
models were established between cracking patterns and clay-sized
aggregates (Fig. 7). Therefore, the crack patterns (Dc and AW) that resulted
fromaggregate disruption could be estimatedwith the content of claysized
aggregates in laboratory simulation experiment for the two paddy
soils. However, these models may not be available in the field because
of more complicated environmental conditions. Our results demonstrate
that the aggregate size distribution after puddling is a key factor influencing
the cracking patterns in paddy fields, regardless of the soil cultivation
history (YPF or OPF soils).
Although the simulated puddling was carried out in the laboratory
with a small amount of soil, the results are in accordance with some
field studies. Bandyopadhyay et al. (2003) and Mohanty et al. (2004)
conducted field experiment in a Vertisol and found that higher puddling
intensity induced lower aggregate mean weight diameter and larger
length, width and surface area of cracks in paddy field. Yoshida and
Adachi (2001) observed that repeated puddling resulted in wider and
simpler cracking patterns in an over 20 year paddy field. This observation
is the same as the simulated puddling experiment in the YPF soil.
Therefore, the results from simulated puddling experiment in the laboratory,
to a certain extent, illustrated that puddling practice in the field
could greatly influence cracking patterns. Furthermore, the laboratory
simulation experiment revealed that Dc and AWof cracks exponentially
increased with increasing puddling intensity (Fig. 6).