resistant to stimulated puddling intensity than the OPF soil because
the water aggregate stability (data not shown) of the YPF soil was
lower than the OPF soil. Deng et al. (2014) reported that paddy soil
with higher SOC is higher in resistance to water slaking and intensive
puddling. Although Fe/Al oxides are binding agents of microaggregates
or even smaller in the YPF soil, they are sensitive to disruption under
slaking and puddling (Zhou et al., 2012). Hence, the clay-sized aggregates
in the YPF soil were much higher than the OPF soil under the
same intensity of stimulated puddling, which contributed more cracking
in the YPF soil (Fig. 7). The difference in cracking patterns between the
two soils under simulated puddling may also be related to the difference
in swelling vermiculite, which was not taken into consideration in this
study. However, compared with SOC and Fe/Al oxides, the effect of
swelling vermiculite on cracking patterns seemed to be minor (Table 5).
Overall, the removal of SOC or Feo/Alo significantly reduced the
difference in crack pattern between the two paddy soils, suggesting
that they were the determinant factors that caused the difference of
cracking patterns. Although the oxalate treatment did not reduce the
difference in clay-sized aggregates between the two soils, it reduced
the difference in Feo/Alo, which contributed a large part of Dc and AW,
especially for the OPF soil (Table 6). The clay-sized aggregates were
quite similar in the two soils after the oxidization of H2O2 (Fig. 2), and
the chemical effect for SOC removal was relative small (Table 6). Thus,
the cracking patterns in the two soils were similar after the removal of
SOC. However, the difference in cracking patterns between the two
soils became larger after the DCB treatment. The reason was that the
removal efficiency of Fed/Ald was lower in the OPF soil than in the YPF
soil (Fig. 3). Moreover, the higher SOC in the OPF soil, to some extent,
could offset the effect of the DCB treatment.