Geology Reference
In-Depth Information
Vanoni & Brooks, 1975: 99). For bare, cohesive
top soils with soil shear strength values at satu-
ration up to 10 kPa, laboratory experiments indi-
cate that
width and t
c
for ephemeral gully development in
these study areas is observed (Poesen
et al
., 2002).
The significant difference in
τ
c
between both study
areas cannot be explained by differences in land
use, as in both cases ephemeral gullies developed
in tilled cropland, but are attributed to different
soil types. Whereas no rock fragments are present
in the Belgian loess-derived soils, the rock frag-
ment content of topsoils in southern Portugal
amounts to 30% by mass on average. Poesen
et al
.
(1999) and Rieke-Zapp
et al
. (2007) demonstrated
experimentally that rock fragment content in top-
soils significantly reduces their susceptibility to
concentrated flow erosion. For non-cultivated land
in Australian valley floors, Prosser (1996) reported
τ
c
values for gully initiation of 21 Pa for a bare clay
soil, but for vegetated soils the values were 70 Pa
for heavily degraded aquatic plants or tussock and
sedge, >105 Pa for undisturbed aquatic plants,
>180 Pa for lightly degraded tussock and sedge
and, >240 Pa for undisturbed tussock and sedge.
Grassed irrigation canals have also been found to
resist flow shear stresses of up to 260 Pa before
showing signs of scour (Reid, 1989, cited by
Prosser 1996).
For prediction purposes, more data are needed
on critical hydraulic conditions leading to gully
initiation, development and infilling in a range
of environments, as well as for different land
management practices. Very few studies have
attempted to measure critical hydraulic condi-
tions for incipient gullying in field conditions,
mainly because of logistical constraints. Therefore,
several studies have instead attempted to assess
critical environmental conditions for gullying
that were more easily quantifiable in field condi-
tions, such as rainfall, topography, soils (or lithol-
ogy) and land use, as these factors control either
the runoff hydraulics, the resistance of the soil
surface to incision, or both.
τ
c
values can go up to 4 Pa (e.g. Rauws &
Govers, 1988; Brunori
et al
., 1989; Crouch &
Novruzi, 1989). These
τ
c
values are of the same
order as those reported for rill incision in bare
topsoils in the field under drainage conditions,
that is, 1.8-10.6 Pa depending on soil properties
(texture, soil water content, content of calcium,
iron, organic carbon and potassium; Gilley
et al
.,
1993). Soil shear strength values at saturation
appear to be a good indicator for the value of
τ
c
(Poesen
et al
., 1998; Knapen & Poesen, 2010).
Experimental data collected by Huang and Laflen
(1996) indicate that critical flow conditions for
rilling under seepage conditions may be signifi-
cantly less than those for drainage conditions.
Land management practices may affect the criti-
cal flow shear stress values for concentrated flow
erosion, as Franti
et al
. (1999) reported that
τ
c
val-
ues for no-till were about twice that for tilled
soil. Along the same lines, Laflen and Beasly
(1960) clearly demonstrated that compaction of
the topsoil increased
τ
c
values. Living plant roots
may increase critical flow conditions for rill
channel development (Li, 1995; Sidorchuk &
Grigorév, 1998; De Baets
et al
., 2006). Knapen
et al
. (2007) recently reviewed published
τ
c
data
and discussed the soil and environmental proper-
ties affecting the soil resistance to concentrated
flow erosion.
In contrast to the number of publications on
critical flow conditions for incipient rilling, very
few studies report critical flow conditions for
incipient gullying. During a rain event, many rills
may develop, but only a few may grow into a gully
provided that local flow intensities exceed those
needed for the erosion of a gully channel. For
cropland, Poesen
et al
. (2003) reported critical flow
shear stresses during peak flow ranging between
3.3 and 32.2 Pa (mean
14 Pa) for ephemeral gul-
lies eroded in silt loam (loess-derived) topsoils in
Belgium, whereas
=
(ii) Rainfall thresholds
Threshold rain depths
(
P
, mm) needed to initiate ephemeral gullies in
cropland (i.e. 14.5 mm <
P
< 22 mm) are only
slightly larger compared with those needed to ini-
tiate rills (i.e. 7.6 mm <
P
< 25 mm) (Poesen
et al
.,
2003). The range of observed threshold values for
τ
c
ranged between 16.8 and
74.4 Pa (mean
44 Pa) for ephemeral gullies formed
in stony sandy loams in Portugal. In general, an
inverse relationship between concentrated flow
=
