Limited data are available in the literature on the partitioning of in
terrill erosion into splash and wash components. Laboratory rainfall s
imulation experiments were conducted on an Oxisol at 5, 10 and 20 degr
ees slopes at a constant rainfall intensity. Rainwash was separated fr
om front splash and lateral splash using a novel experimental design.
Results indicate that output of sediment from the 0.18 m(2) erosion pl
ots was dominated by splash with a geometric mean aggregate diameter (
GMAD) similar to that of the original soil. The most easily detached a
ggregate size by splash was 500 to 1000 mu m, i.e., coarse sand-sized
aggregates. Wash was dominated by aggregates with a GMAD significantly
finer than the soil matrix, and this reflected the limited energy ava
ilable to transport coarse aggregates detached by splash. Wash and spl
ash sediment flux, and the runoff coefficient increased significantly
with slope angle. Peak front splash was associated with a different di
mensionless water depth (DWD) for each slope angle. As slope angle inc
reased the DWD at peak output increased from 0 to 0.4 at 5 degrees, an
d from 1.4 to 1.7 at 20 degrees. Time trends of front splash output fo
r 5 and 10 degrees slopes were similar to those presented in the liter
ature. However, data for the 20 degrees slopes were more complicated,
with a series of pulses related to periodic migrating waves of wash se
diment superimposed on the general pattern. These migrating bedload wa
ves are interesting in that they seem analogous to those documented in
the fluvial environment at steady or even decreasing discharge. Data
indicate that front splash flux exceeded wash transport on all slopes
tested. These data reflect the boundary conditions of the experiment,
but may be applicable under certain field situations, such as on short
, steep, furrow side-slopes.