3.2. Effect of slope steepness on wash loss
Table 1 shows the average wash loss and standard deviation for 4
replicates. The wash loss was also found to vary with slope.
Specifically, the wash loss increased and then decreased after the
peak wash loss was reached (Fig. 3). The maximum value occurred
when the slope gradient was equal to 58%. The wash loss was affected
by both runoff flow and slope gradient. When rainfall was kept
constant, runoff increased to a peak value and then decreased with
slope (Fig. 4). The decrease in runoff played a much more important
role in decreasing wash loss when the slope gradient was greater than
58%.
Eq. (4) derived by Foster from analysis of the data generated by
Lattanzi et al. (1974) has been recommended in the RUSLE (Renard et
al., 1997) for slopes shorter than 4.6 m (McCool et al., 1987):
S = 3:0ðsinαÞ0:8 + 0:56 ð4Þ
where S was the slope factor and α was the slope gradient (degree).
Lattanzi et al. (1974) measured the wash loss at slopes with lengths of
0.61 m under simulated rainfall. Fig. 5 provides a comparison of the
data generated in the present study to Eq. (4). Eq. (4) matched the
data from this study very well when the slope steepness was less than
or equal to 58%.
When the slope gradient was greater than 58%, the wash loss
decreased with slope gradient (Fig. 3). Regression analysis revealed
that the relationship between the wash loss normalized to 9% and the
sine of slope can be described by Eq. (5), which was significant at the
confidence level of 0.05.
S = −5:23 sinα + 4:97 R2 = 0:85 ð5Þ
Most studies of short slopes have found a relationship between
wash loss and the slope that was nearly linear. This may have been
because the experimental slope was not steep enough in these
studies. The maximum experimental slope gradient in previously
conducted studies was 50%. However, Chaplot and Le Bissonnais
(2003) applied low rainfall intensities of 1.5 to 30 mm h−1 to low
slope gradients of 4% and 8% and found that soil loss was not
correlated with slope gradient. Therefore, rainfall intensities may have
an effect on the relationship between wash loss and slope steepness.
3.3. Partitioning downslope transport into wash and splash components
Fig. 6 shows the variation in the ratio of splash loss to wash loss
with slope steepness. Slope had an effect on the ratio of splash loss to
wash loss. Specifically, the ratio of the net downslope splash loss
(downslope splash loss minus upslope splash loss) to the wash loss
increased from 0.21 to 1.33 as the slope gradient increased from 9% to
100%. These findings indicate that the wash loss was the dominant
Fig. 3. Effect of slope on wash loss.
32 S. Fu et al. / Catena 84 (2011) 29–34transport form of erosion when the slope angle was less than 70%,
close to the net downslope splash at slopes of 70% and 84%, and less
than net downslope splash at a slope of 100%. However, the ratio of
the net downslope splash loss to wash loss was greater than 0.5 at all
slopes except at 9% slope. Thus the result of this study indicates that
splash transport was also a very important part of interrill sediment
delivery at short steep slope settings, which are typical of many
interrill areas in agricultural fields