Geoscience Reference
In-Depth Information
to modify the absolute rates of infiltration based on other
soil characteristics, rather than providing the principal
control. Martinez-Mena, Castillo and Albaladejo (2002)
demonstrated that crusts also formed rapidly on marly
soils in southern Spain, so that they are by no means just
limited to sandy soils.
There is an increasing body of work which suggests that
biological processes are also highly significant in crust
formation. Belnap (2006) provides an excellent review of
biological crusts and their effects in drylands. Biologi-
cal crusts may be formed by cyanobacteria and lichens,
but green algae, microfungi, bacteria and bryophytes may
also live in the top few millimetres to centimetres of the
soil. They bind the soil together both directly, e.g. in the
form of filaments that can be seen with the naked eye if a
section of crust is held to the light, and chemically, by the
production of polysaccharides secreted on the outer sur-
face of the organisms. Belnap notes that many previous
studies on the effects of biological crusts have been incon-
clusive because they fail to account for the different types
of crust and confounding variables that are modified in
standard experimental treatments. She aims to overcome
these limitations by defining four types of crust character-
istic of different environments (Figure 11.5). Typically, as
conditions go from hyper-arid to semi-arid then cool and
cool-cold semi-arid, cyanobacteria in the crust decrease
and lichens increase, leading to smooth, rugose, pinnacled
and rolling crusts, respectively. However, cyanobacteria
will be dominant in very sandy or saline soils, or in loca-
tions where there is a high proportion of swelling clays;
lichens favour carbonate-, gypsum- or silt-rich soils, lead-
ing to crusts of the different types in other climate regimes.
Pinnacled crusts have the biggest effect on promoting in-
filtration by producing hollows in which ponded water in-
filtrates without running off. Rolling crusts may increase
infiltration to a lesser extent, while the smooth and rugose
crusts tend to decrease infiltration by blocking the pores
at the surface. Belnap
et al
. (2005) note that biological
crusts may make up as much as 70 % of the surface of
dryland soils. They also suggest that behaviour of biolog-
ical crusts is likely to be dynamic with relation to the size,
duration and temporal spacing of rainfall events. After ex-
tended dry periods, the first wetting may kill 30 to 50 %
of the bacterial biomass, so that subsequent events will be
less affected until regrowth occurs. Regrowth can be rapid
during a monsoon season, for example, but depends on the
exact timing and magnitude of subsequent rainfall pulses
(Cable and Huxman, 2004). For this reason, infiltration
and surface-roughness characteristics may be expected to
be highly dynamic at different times in the year. Biologi-
cal crusts may lead to other spatial feedbacks in that they
RAIN
4
2
7
3
4
1
4
5
3
3
6
7
6
5
8
1
Seal
6
6
6
8
1
8
Figure 11.3
Hydrological processes occurring on and in soils
with rock fragments at or near the surface: 1, water absorp-
tion; 2, interception and depression storage; 3, rockflow; 4,
evaporation; 5, infiltration; 6, percolation; 7, overland flow; 8,
capillary rise (from Poesen and Lavee, 1994).
Valentin and Bresson's second type of structural crust
occurs by slaking, and is perhaps better considered as a
surface seal (see below). Erosional and depositional crusts
are the result of selective entrainment and deposition of
soils, resulting in the formation of smooth surfaces where
surface pores are infilled with silt- and clay-sized parti-
cles. Depositional crusts can also be in the form of saline
deposition, which are most closely associated with playas
(Chapter 15). Graef and Stahr (2000) show that crusts are
almost ubiquitous in Niger but vary as a complex function
of soil, slope and land-use conditions, so that very differ-
ent crust types can be found over spatial scales of a few
square metres.
Valentin (1993) gives an example of an erosional crust
in Burkino Faso, where the final infiltration rate increased
from 1.6 to 6 mm/h when the crust was destroyed by
tillage. However, over an entire rainfall season, the dif-
ference was less marked (285 mm compared to 241 mm
total runoff), which is likely to be due to the reformation
of the crust after the initial storm events. Puigdefabregas
et al
. (1999) noted the presence of sieving crusts based on
soil micromorphological evidence at the Rambla Honda
in southern Spain, but noted that the infiltration rates as-
sociated were much higher than those measured in West
