Geology Reference
In-Depth Information
to yield a sediment concentration thought to be
typical for the container. It has been shown by
Ciesiolka
et al
. (2006) that this type of technique
can lead to serious underestimation of soil loss.
The reasons for this underestimation include the
effect of rapid settling of larger soil particles
which escape sampling, and the difficulty in
bringing the collected sediment to a spatially uni-
form concentration by stirring.
The technique for measuring soil loss from
runoff plots which was developed in association
with GUEST is described in Ciesiolka
et al
.
(1995), Coughlan (1997), and Coughlan and Rose
(1997b). Sediment leaving the hydrologically-
defined runoff plot is collected in a shallow trough
of low slope (
varied in length at different sites, from 30 m at low
slopes to 12 m at higher slopes (reflecting farmer
practice), with experimental plots being essen-
tially planar and hydrologically defined by bound-
aries at the top and sides.
11.6
Results of Some Field Projects Using
GUEST Technology
11.6.1
ACIAR projects in southeast Asia
and Australia
This section summarizes the results of Projects
8551 and 9201 supported by the Australian Centre
for International Research (ACIAR), which used
the GUEST program and experimental methodol-
ogy as described by Rose
et al
. (1997) and Rose
and Yu (1998). Figure 11.3 shows the location of
the seven ACIAR project sites, together with
sites for the IBSRAM-ASIALAND projects to be
described later. The two sites in Australia were
located on mechanized commercial pineapple
farms in a subtropical zone, whilst the tropical
southeast Asian plots were cultivated by hand or
oxen plough. Some details of the sites, including
soil type, and the size and slope of plots are given
in Table 11.1. At Kemaman the main treatment
plot of cocoa and bananas was accompanied by a
small bare plot. Tropical crops at other sites
included rozelle at Khon Kaen, maize at Nan,
maize and mungbean at Los Banos, and maize
and peanuts at VISCA. At VISCA there were both
low-slope plots at 10% as well as the quite steep
slopes typical of hand-hoe cultivation on the
island of Leyte. Plots of different lengths (and
therefore area) were investigated in the Australian
pineapple plantation sites.
Some information on plot treatment and exper-
imentation at the different sites is given later.
1%). This low slope encourages
deposition of the coarser and more rapidly set-
tling fraction of the eroded sediment. This depos-
ited sediment is collected after the erosion event,
weighed and subsampled to determine its water
content, which is then used to convert the mass
of deposited sediment to an oven-dry basis.
The flow of suspended sediment leaving the
trough is then passed through a flow measuring
device. For larger plots this can be a flume, or
for plots of more modest size a 'tipping bucket'
device is convenient (Ciesiolka
et al
. 1995;
Ciesiolka & Rose, 1998). Using either tech-
nique, an average concentration of this sus-
pended sediment can be obtained, yielding a
total loss of suspended sediment which is then
added to the deposited sediment load collected
in the trough to yield the total soil loss for the
erosion event. Use of either of these techniques
also directly yields information on runoff rate
for use in GUEST.
The separate measurement of the fine sus-
pended sediment fraction also allows information
on the loss of nutrients and carbon to be obtained,
which are commonly enriched ingredients in the
suspended load. It is the loss of such components
that can subvert the long-term sustainability of
productive land uses.
Plate 7 illustrates the described experimental
arrangements for the very steep experimental
plots at the Visayas College of Agriculture (VISCA),
Leyte, the Philippines. The hand-cultivated plots
≈
(i) Soil and water loss
Table 11.2 provides a sum-
mary of average annual runoff, runoff coefficient
(defined as the ratio of average annual runoff to
average annual rainfall), soil loss and sediment con-
centration for the five sites involved in ACIAR
Project PN9201. Results for each site are given,
firstly for a reference plot kept essentially bare by
