Geoscience Reference
In-Depth Information
amount of water available at the ground surface for infil-
tration. Abrahams, Parsons and Wainwright (2003) mea-
sured the maximum canopy storage in the desert shrub
creosotebush, which was found to be 5 mm of rain, and
a similar amount was found for juniper (Owens, Lyons
and Alejandro, 2006). In small events, this effect might
be significant, so that over a rainfall season in New Mex-
ico, Martinez-Meza and Whitford (1996) estimated about
44 % of rainfall was intercepted by creosotebush, about
45 % for tarbush and 37 % for mesquite. Dunne, Zhang
and Aubry (1991) note that the interception of rainfall also
increases infiltration by the reduction of rainfall energy,
which in turn reduces the likelihood of crust formation.
Even for sparse shrub canopies, Wainwright, Parsons and
Abrahams (1999) demonstrated a 30 % reduction in en-
ergy in this way. Stemflow generated by plants can, on
the other hand, increase the point source of precipitation
at the ground surface by concentrating it from the larger
canopy area. Proportions of 5-10 % of precipitation have
been commonly reported for stemflow (Martinez-Meza
and Whitford, 1996; Owens, Lyons and Alejandro, 2006),
although values as high as 42 % have been recorded
for acacia and eucalyptus species in Australia (Press-
land, 1973; Nulsen
et al.
, 1986). Abrahams, Parsons and
Wainwright (2003) note that it is important to look at
shrub understoreys to evaluate the impact of stemflow.
Where no understorey was present, they found that the
high point source can lead to some shrubs producing small
amounts of runoff in a storm event before surrounding bare
surfaces start to pond. However, when a grass understorey
was present, all of the stemflow was usually found to in-
filtrate - with an important feedback to water availability
to maintain plant growth. Spatial patterns that emerge as
a result of the feedbacks between increased infiltration
under and near vegetation are recognized as an impor-
tant characteristic of the ecogeomorphology of drylands
in many parts of the world (Cerda, 1997; Dunkerley, 2002;
Janeau, Mauchamp and Tarin, 1999; Gutierrez and Her-
nandez, 1996; Pariente, 2002; Wainwright, Parsons and
Abrahams, 2002; Ludwig
et al.
, 2005; Van Der Kamp,
Hayashi, and Gallen, 1967; Smit and Rethman, 2000).
Disturbance of the surface through digging by small
mammals has been noted to have different effects on infil-
tration. Moorhead, Fisher and Whitford (1988) and Mun
and Whitford (1990) have suggested that such digging in-
creases porosity and thus infiltration in the Chihuahuan
Desert, although Neave and Abrahams (2001) were un-
able to demonstrate a significant different in water yield
between areas with digging and those without. Larger
mammals can also have an impact on infiltration, most
notably by compaction, although Hiernaux
et al
. (1999)
high stocking densities. At moderate rates of animal pres-
ence, infiltration rates were observed to increase slightly
due to the effect of the animals on breaking up existing
crusts. There are also likely to be increases in such condi-
tions due to enhanced aggregate stability in the presence
of higher levels of organic matter.
Fire is an important process in many drylands (Naveh,
1975; Wainwright and Thornes, 2003; Snyman, 2003).
The effect of burning is dependent on temperature con-
ditions reached during a fire, which is a function of fuel
availability as well as temperature, wind and other lo-
cal conditions. Giovannini, Lucchesi and Giachetti (1988)
found experimentally that sand content changed little in
burnt soils for temperatures up to 170
◦
C, but between 220
and 460
◦
C there was a rapid rise so that sand became the
dominant particle size due to aggregation and fusing of
particles, but higher temperatures produced little further
effects. Loss of organic matter, and thus aggregate stabil-
ity is important, especially at temperatures above 170
◦
C
(Giovannini, Lucchesi and Giachetti, 1990). Water repel-
lency can be developed as organic matter is vapourized
at or below the surface (as a function of the tempera-
ture reached and its transmission through the soil profile;
see Neary, Ryan and DeBano, 2008), although Cerda and
Doerr (2007) have suggested that calcic soils may reduce
the likelihood of water repellency occurring. Certain veg-
etation types - such as pines in the Mediterranean and
creosotebush in the US - are more likely to release water-
repellent chemicals into the soil. Thus, there is likely to be
a significant spatial variability of the effects of fire, with
some locations having reduced infiltration and others in-
creased (Imeson
et al.
, 1992; Lavee
et al.
, 1995).
Water repellency can also develop in many dryland
soils as a result of the breakdown of phenolic, turpene and
related compounds from plants at ambient temperatures.
There have been suggestions that these substances com-
monly evolve in dryland plants as a way of competing
with neighbours (but see the discussion in Fitter and Hay,
1987) or as a means of reducing herbivory (Hyder
et al.
,
2002). Beyond the reduction of infiltration rates that wa-
ter repellency produces, Ritsema and Dekker (2000) have
demonstrated that in sandy soils it can lead to the develop-
ment of significant amounts of preferential - sometimes
known as 'fingered' - flow, which is a further reason why
the Darcy-Richards approximation is problematic in such
conditions.
11.3.2
Subsurface controls
