The direct measurement of changes in soil level is appropriate in the case of localized erosion where rates are high and the position of the erosion can be predicted, such as steepland which has been deforested, or cattle tracks on rangeland. It is usually not suitable for soil losses from arable land because the surface level is affected by cultivation and settlement, although short-term changes have been studied in potato furrows in Australia (McFarlane, Delroy and van Vreeswyk 1991). Changes can be measured in one dimension for surface level at a point, or in two dimensions to give a profile or cross-section, or in three dimensions for volumetric measurements of rills or gullies.
- Point measurements
Individual measurements of change in level at a single point will vary widely, but if it is an inexpensive and simple method, and a large number of points can be sampled, then a usable estimate can result.
- Erosion pins
This widely-used method consists of driving a pin into the soil so that the top of the pin gives a datum from which changes in the soil surface level can be measured. Alternatively called pegs, spikes, stakes or rods, the pins can be of wood, iron or any material which will not rot or decay and is readily and cheaply available. Off-cut lengths of round iron bars for reinforced concrete can usually be picked up at little or no cost from construction sites. In some developing countries, iron or steel pins or nails might be stolen, in which case bamboo or reed canes cut locally might be more suitable.
The pin should be a length which can be pushed or driven into the soil to give a firm stable datum: 300 mm is typical, less for a shallow soil, more for a loose soil. A small diameter of about 5 mm is preferable, as thicker stakes could interfere with the surface flow and cause scour. A rectangular or square grid layout will give a random distribution of points with a spacing appropriate to the area being studied.
An illustration of the point method comes from a study in Japan, where pegs were installed to a 2 m square grid on three 100 m² plots with a 30° slope recently cleared of forest. Measurements of the peg heights were made every month for ten years and showed that the annual rate of erosion from each plot was almost consistent at about 13 mm/year (Takei, Kobaski and Fukushima 1981).
In another example the pin method unexpectedly gave a quantitative measure of the effect of a single heavy storm in western Colorado. As part of a long-term hydrology study in a 5 ha basin, pins were installed at 1.5 m intervals on six selected profile lines. All runoff and sediment from the basin is captured in a reservoir at the basin outlet, so that estimates of loss as measured by the pins could be compared with measurements of sediment from surveys of the reservoir. A heavy storm, with an estimated return period of 25 years, occurred shortly after the installation of the pins and the first reservoir survey, enabling an assessment of the isolated effect of the storm. The mean soil loss calculated from the pin results was a depth of 2.7 mm, compared with the estimate of the sediment held in the reservoir which would correspond to a loss of depth of 2.3 mm (Hadley and Lusby 1967), a good measure of agreement.
Some researchers slip a metal washer over the pin to give a better base from which to measure to the top of the pin. If there are likely to be cycles of erosion and deposition such as in a gully floor, the washer method may give useful additional information by falling to the lowest erosion level and being covered by any later deposition which can also be measured. On the other hand, the presence of the washer may cause turbulence and scour, or it could reduce splash erosion and leave the washer sitting on a pedestal of soil. All these variations and possible causes of false readings have been reported in the literature on the use of the pin method which is reviewed by Haigh (1977).
- Paint collars
An indication of large changes in level, for example in a stream bed or gully floor, can be obtained by painting a collar just above soil level round rocks, boulders, tree roots, fence posts, or anything firm and stable. Erosion reveals an unpainted band below the paint line, indicating the depth of soil removed. When painting the collar it is advisable to mask the soil with old newspaper as paint accidently sprayed or brushed onto the soil might make it less erodible.
- Bottle tops
Another simple way to record the original level is to press bottle tops into the soil surface. The depth of subsequent erosion is shown by the height of the pedestals where the soil is protected by the bottle top. This leads to the use of naturally occurring indicators of changes in soil surface level.
- Pedestals
When an easily eroded soil is protected from splash erosion by a stone or tree root, isolated pedestals capped by the resistant material are left standing up from the surrounding ground (Plate 1). The erosion of the surrounding soil is shown to be mainly by splash rather than by surface flow if there is little or no undercutting at the base of the pedestal. Like the bottle top method, it is possible to deduce approximately what depth of soil has been eroded by measuring the height of the pedestals.
- Tree mounds and tree roots
In arid or semi-arid climates it is not unusual to find that the surface under trees is raised in a gently sloping dome. In a comprehensive project in Tanzania from 1968-1972 Rapp and colleagues suggested that the mounds are the result of the trees protecting the soil from splash erosion while the surrounding soil is eroded. By measuring the height of the mounds and the age of the trees from tree-ring counts, they estimated a soil lowering of about 10 mm/year (Rapp et al. 1972). However, based on later research in Botswana, Biot (1990) calculated that the rate of denudation as calculated by this method is ten to fifteen times greater than estimates by other methods. He offers the alternative suggestion that the tree mounds can be explained by a difference in bulk density between soils in the mounds and the surrounding flat soils. He concluded that the mound results from a raising of the local surface rather than erosion of the surrounding surface.
Exposed tree roots may offer a valid indication of change when the reason is obvious, such as erosion in a streambed below a paint collar, but exposed tree roots offered as evidence of sheetwash, or of wind erosion in dry climates, should be treated with caution for Biot's hypothesis may also apply. Very long-term rates of erosion (over several centuries) were estimated from tree root exposure in Colorado (Carrara and Carroll 1979).
Clumps of grass elevated above the surrounding soil surface should also be treated with caution for the change may be the result of the grass trapping soil particles splashed from the surrounding soil. This was conclusively shown in Zimbabwe where erosion was measured from runoff plots under various tobacco/grass rotations. After a few years the tufts of weeping lovegrass (Eragrostis curvula) were found to be several centimetres higher than the soil surface between, although the measured soil loss from the plot was negligible. Some simple tests with splash boards showed that there was no net soil loss from the plot, but consideration translocation of soil within the plot. Clearly it is necessary to be certain that changes in soil surface level are the result of erosion down rather than elevation upwards.
- Profile meters
To measure small changes in surface level along a cross section such as an area with a number of parallel cattle tracks, a profile meter may be suitable. (The case of larger changes as in rills and gullies is discussed in the next section on volumetric measurements.) The requirement for a profile meter is to be able to set up a datum from which changes in level can be measured along a straight line and which can be re-established at the same points later to measure changes in level. Usually this takes the form of a horizontal bar with rods which can be lowered down to the soil surface, and is the same principle as used to measure surface roughness in studies of tillage and tilth. Such a device to measure surface levels accurately on grazing land was developed by Hudson (1964). Metal pegs were set unobtrusively at ground level in concrete blocks at intervals of 2 m. A light aluminium girder could be fitted onto any two adjacent pegs and this gave a firm datum from which the level to the soil surface could be accurately measured at positions marked on the girder. Between readings the girder was removed so that there was no interference with cattle movements. Measurements were taken to the nearest millimetre, which allowed annual changes to be clearly recognized. A similar device was used in the Philippines by Ramirez (1988) and is shown in Figure 8. Another approach was used by McCool, Dossett and Yecha (1981). In this case the pins are lowered to the soil surface at the same time and the profile is recorded by camera for later evaluation.
Several other more sophisticated profile meters have been developed and details are given in the section Further reading.
Source : FAO.org
No comments:
Post a Comment