Geoscience Reference
In-Depth Information
surface of low cohesive strength, such as a relatively
coarse or sandy layer. It has been shown that the
mineralogy-dependent flocculation of clays can result in
depositional seals having varying densities and hydraulic
conductivities (Lado, Ben-Hur and Shainberg, 2007), and
given that electrochemical forces are involved, the com-
position of the transporting water, including its salinity,
are certain to affect the properties of depositional seals.
associated with soil surface seals. There is a great diver-
sity in the methods and approaches used to study these
phenomena, and we will consider only some examples of
what has been learned.
The effects of seals on water erosion in drylands remain
only partially understood. Singer and Shainberg (2004) ar-
gued that if there is abundant erodible sediment and ero-
sion is transport limited (i.e. there is insufficient capacity
in the overland flow to remove all the available material),
then sealing by promoting more runoff may increase sur-
face erosion. On the other hand, they argued, if erosion
is detachment-limited (i.e. the surface is somewhat ero-
sion resistant and provides a smaller load of particles than
could be carried away by the overland flow), then seal
development is only likely to increase the extent of de-
tachment limitation, and hence reduce the erosion rate. In
the field, conditions are more complex than these analyses
suggest. It has to be remembered that there may be runon
water arriving from upslope, perhaps with an associated
sediment load, so that erosion over a sealed surface may
depend upon its landscape context. Moreover, the soil sur-
face in the field may have a variable cover of organic litter
and stones, and the interaction of these materials with rain
and overland flow cannot readily be isolated from those of
seals on the exposed soil. Such issues were explored at a
field site in New Mexico under simulated rainfall (Neave
and Rayburg, 2007). The soils were mechanically broken
up prior to the start of experiments, and it was found that
penetration resistance increased progressively as seals de-
veloped through successive applications of rain, separated
by time for soil drying. Soil surfaces protected from rain-
drop impact by mesh screens nevertheless showed similar
behaviour, perhaps because of the occurrence of deposi-
tional seals. However, the results showed complex pat-
terns of runoff coefficients and sediment yields as a func-
tion of other surface properties such as stone and organic
litter cover. The highest sediment yields were recorded
by Neave and Rayburg (2007) from partially sealed, par-
tially stone covered plots, rather than from bare, sealed
ones. This may reflect the formation of concentrated flows
passing downslope between the stones and forming more
efficient transport pathways for sediment. A related ar-
gument has been raised in relation to surface roughness
and microrelief. Where sealed soils are flat, overland flow
may occur as a relatively uniform sheetflow, with shal-
low depth and low flow speed. This would allow greater
time for water absorption. On the other hand, a site with
an undulating surface would carry overland flow in faster-
moving flow threads, tracing out the lower-lying pathways
across the surface. These faster flows would drain water
more quickly from the soil surface, and hence tend to in-
7.5.4
Effects of seals on infiltration and erosion
An example of the effect of soil seals on infiltrability
was provided by Ries and Hirt (2008) from studies of
abandoned fields in the Spanish drylands. Using cylin-
der infiltrometer tests, they reported a final infiltrabil-
ity of 4.6 mm/h for crusted silty soils, compared with
10.2 mm/h for uncrusted soils. Thus, the presence of a seal
reduced the infiltrability in this area by more than 50%.
For small plots under simulated rain (40 mm/h) Ries and
Hirt (2008) also demonstrated runoff ratios on sealed soils
of up to 80%, and averaging 61.3%, more than double the
value on vegetated plots. Similarly low infiltrability val-
ues were reported from sealed silty soils at an arid site in
Jordan, where the mean infiltrability was just 3.2 mm/h
(Al-Qinna and Abu-Awwad, 1998). A final example can
be drawn from work done in Israel (Carmi and Berliner,
2008) in which micropermeameter measurements were
collected on sealed and nonsealed (tilled) soils. The sealed
soils exhibited a mean infiltrability of 5.05 mm/h, while
for nonsealed soils the value was about 20 mm/h. Thus,
the soil seal in this study reduced the infiltrability by
about 75%. Other data on infiltrability of various kinds of
soil seals from west Africa were presented by Casenave
and Valentin (1992). Experimental studies have confirmed
that the conductivity of soil seals can be sufficiently low
that the soil beneath conducts only unsaturated flow (e.g.
Morin, Benyamini and Michaeli, 1981). In this situation,
the hydraulic conductivity that is manifested by the soil
below the crust is dramatically lower than the saturated hy-
draulic conductivity, since strong soil suction forces arise
during unsaturated conditions, as well as surface tension
effects at the air-water interface in partially filled void
spaces.
It is clear from studies such as those just mentioned
that seals act to throttle infiltration to much lower rates
than would occur in the absence of the seal. Therefore, it
follows that a larger fraction of the rain falling on a sealed
surface must be partitioned into surface ponding or over-
land flow. The prospect of increased overland flow brings
with it the risk of increased erosion and transport of soil
Search WWH ::




Custom Search