Geology Reference
In-Depth Information
results for the two non-bare soil treatments at the
Kemaman site (Table 11.2), for which canopy
cover by the cocoa tree crop and companion shade
trees was similar. However, the grass and legume
contact cover in the improved practice reduced
average annual soil loss by a factor of five, which
is a greater reduction than that due to the accom-
panying reduction in runoff coefficient.
At both Los Banos and VISCA sites in the
Philippines, the improved practice for which
data are given in Table 11.2 included the intro-
duction of leguminous hedgerows whose trim-
mings were a component of mulch added to the
cropped alley between contour-planted hedge-
rows (Paningbatan
et al
., 1995). Although the
relative improvement in soil loss due to the
presence of the hedgerow or the mulch cannot
be dissected from the data, runoff coefficient
and sediment concentration are both reduced by
this combined practice (Table 11.2). Although
hedgerows can fail during extreme typhoon
events, causing a dramatic increase in soil loss
akin to contour bank failure, hedgerows do
reduce soil loss more than would be expected in
terms of reduced runoff. This reduction is evi-
dently due to net deposition from sediment-
laden water which is slowed down and infiltrates
as it moves through the half-metre or so of
hedgerow. The reasons for this net deposition
have been clarified in separate studies (Rose
et
al
., 2002, 2003; Hussein
et al
., 2007a,b).
Total runoff from plots with a conventional
farmer practice of up-and-down slope cultivation
was less than for bare soil except at the VISCA
site, where weed control by hand hoe in the bare
plot under wet conditions appeared to produce
large stable aggregates and enhance infiltration.
However, soil loss was reduced by conventional
practices compared with bare soil, sometimes
considerably, and this was mostly due to a reduc-
tion in sediment concentration (Table 11.2).
Results in Table 11.2 for the improved practice
(the last listed treatment for each site) show a
substantial reduction in runoff, notably at the
Kemaman site, although longer period data are
needed for the Nan site. For soils of light texture
(at Goomboorian in Australia, and Khon Kaen in
Thailand), the improved practice had a greater
effect on sediment concentration than on runoff.
The runoff coefficient,
R
c
, is shown in Table
11.2 to be very variable. However, the range of
R
c
(0.27 to 0.62) for soils of lighter texture is higher
than for the range for clay soils (0.02 to 0.19).
Clay content can aid better water-stable aggrega-
tion and associated biotic activity, perhaps lower-
ing
R
c
, although
R
c
is affected by rainfall amount
and distribution and other factors.
(ii) Soil erodibility
Khon Kaen, Thailand
: The
relative importance of rainfall-driven erosion ver-
sus erosion driven by overland flow was investi-
gated as follows. Soil from the runoff plots at each
site was placed in specially constructed detach-
ment trays as described in Rose (1993) designed to
yield the average sediment concentration during
an erosion event due to rainfall impact alone. The
sediment concentration from these small low-
slope trays was compared with the average sedi-
ment concentration measured in plot runoff. At
the low slope (4%, Table 11.1) plots at Khon Kaen,
these sediment concentrations were similar, with
that from the detachment tray sometimes exceed-
ing that from the plot, and a mean value of the
soil erodibility parameter
b
(defined in Equation
(11.11) ) of 1.05 was obtained, with very little
storm-to-storm variation (Sombatpanit
et al
.,
1995). Whilst this value of
b
indicates that over-
land flow was virtually at the transport limit, the
detachment tray data indicate that the sediment
concentration at this limit could all be provided
by the impact of rainfall. Note, however, that the
parameter
b
incorporates the effects of any proc-
ess that contributes to erosion. Detachment tray
data showed flow-driven erosion to be the domi-
nant erosion mechanism at other ACIAR project
sites with higher slopes than those at Khon Kaen.
The local concern with soil erosion at Khon Kaen
may have had more to do with loss in soil fertility
associated with the measured preferential loss of
fine soil components and associated nutrients
than with excessive soil losses. It is well known
that nutrient loss is enriched in eroded sediment
(Rose & Dalal, 1988), and this was measured in
these projects (Hashim
et al
., 1997).
