Environmental Engineering Reference
In-Depth Information
attempts to explain dramatic shifts in response to
anthropogenic or natural environmental changes.
These include semi-arid grasslands (Rietkerk & van de
Koppel 1997), coral reefs (Knowlton 1992), wet dune-
slacks (Adema
et al
. 2002) and rocky shores (Petraitis
& Latham 1999), where large and sudden changes in
ecosystem state have been observed. The success of
these attempts to apply the aforementioned models of
alternative stable states was limited (Bertness
et al
.
2002). Apart from a lack of experimental evidence, an
important reason for this is that the theory and models
developed for semi-enclosed, relatively homogeneous
bodies of shallow-lake water were not suffi cient to
explain the dynamics of open, spatially extended and
heterogeneous ecosystems such as those listed above.
In ecosystems such as those, the interplay between
local feedback relations between organisms and their
respective environments and feedback processes occur-
ring at large scales, generates complex dynamics that
can lead them to respond in unexpected ways. Thus, a
more complex theory is required to predict how these
ecosystems respond to changing environmental condi-
tions. Such a theory should explicitly take into account
feedback relations and interactions that cross spatial
scales, as discussed in the next section.
We elaborate here on the example of banding or
patchy vegetation patterns as found in arid and semi-
arid lands in the African Sahel, Australia and else-
where, which have been much studied (e.g. Valentin &
d ' Herbes 1999 ; Tongway
et al
. 2001). In this case, once
again, a recurrent research question has been whether
or not such patterns result from pre-existing environ-
mental heterogeneity, from spatial self-organization or
from both. Klausmeier (1999) analysed a series of spa-
tially explicit models and showed that the patterns can
be explained by spatial self-organization alone, that is
they are caused by one single mechanism (cf. Thiery
et al
. 1995 ; Valentin
et al
. 1999). Banded or patchy veg-
etation promotes the infi ltration of water into the soil,
a process which benefi ts vegetation growth under arid
or semi-arid conditions. In vegetated bands or patches,
more water infi ltrates than in bare patches. Overland
fl ow, in particular on hill slopes, then generates a
net fl ux of water into the vegetated patches, and
decreases water availability in the bare patches. This
interaction between vegetation, water infi ltration,
and overland fl ow of water fully explains the formation
of spatial patterns involving so-called
runoff
and
run - on
areas, and can thus be considered as spatial
self - organization.
Mathematical models indicate that spatial self-
organization can have important implications for the
functioning of ecosystems (Ludwig
et al
. 1999 ; Rietkerk
et al
. 2002), in line with indications from empirical
work (Valentin & d'Herbes 1999). Indeed, for arid eco-
systems, spatial patterns are predicted to compensate
for reduced and unpredictable rainfall, allowing for
plant growth under conditions that would not sustain
plant life if spread homogeneously. This implies that
spatial patterns generate
feedback mechanisms
that can
compensate for changed environmental conditions,
such as drought. Moreover, consistent and predictable
changes occur in spatial patterning before the buffer-
ing feedbacks are overwhelmed and the system shifts
to an alternative state characterized by a bare, degraded
landscape (Rietkerk
et al
. 2004b). This provides a basis
for the development of indicator systems that can
predict sudden shifts between alternative stable states
in complex dynamic systems (see below).
6.2.3
Spatial self-organization
When ecologists observe spatial patterns in ecosys-
tems, typically they seek the cause or mechanism in
environmental variations at the local or regional levels
of organization. Although this is valid in many cases,
a number of studies over recent decades revealed clear
and consistent spatial variation in the structure of eco-
systems in landscapes that exhibit no underlying vari-
ation in environmental conditions and can be explained
by so - called
spatial self - organization
, a process whereby
internal interaction between biotic and abiotic compo-
nents of an ecosystem generates complex but recogniz-
able and repeated spatial patterns (Rietkerk & van de
Koppel 2008; see Figure 6.1). Spatial self-organization
has been suggested as the mechanism in nature that
leads to the creation of regular patterns, such as those
found in the vegetation of some semi-arid land systems
(Klausmeier 1999) or peatlands (Rietkerk
et al
. 2004a ),
as well as the irregular patterns found in mussel
beds on wave-disturbed rocky shores (Guichard
et al
.
2003), and in some Mediterranean grasslands (Kefi
et al
. 2007 ).
6.2.4
Implications for restoration ecology
Insight in the interactions between both negative and
positive feedback relations is crucial to understanding
