Geology Reference
In-Depth Information
adopt new configurations and act as focal points for
landscape change. Relatively immobile, slowly respond-
ing parts (plateaux and interfluves, some soils and
weathering features) lie far from susceptible parts. This
differential susceptibility of landscapes to erosion would
permit fast-changing 'soft spots' to exist alongside stag-
nant areas. But it does not explain why some areas
are stagnant. Weathering should construct regolith, and
erosional processes should destroy it on all exposed
surfaces, though the balance between constructive and
destructive forces would vary in different environments.
The second mechanism helps to explain the occurrence
of stagnant areas. This is
the persistence and domi-
nating influence of rivers
(Twidale 1997). Rivers are
self-reinforcing systems: once established and dominant,
they tend to sustain and augment their dominance.
Thus, major rivers tend to persist in a landscape. In
Australia, some modern rivers are 60 million years old
and have been continuously active since their initia-
tion in the Eocene epoch. Other equally old or even
older rivers, but with slightly chequered chronologies,
also persist in the landscape (see Ollier 1991, 99-103).
Likewise, some landscapes reveal the ghosts of other very
old rivers. Rivers of similar antiquity occur in other
Gondwanan landscapes. Such long-running persistence
of rivers means that parts of landscapes remote from
river courses - interfluves and summits for example -
may remain virtually untouched by erosive processes
for vast spans of time and they are, in geomorphic
terms, stagnant areas. A third possible mechanism for
landscape stagnation comes from theoretical work. It
was found that landscape stability depends upon time-
lags between soil processes, which act at right-angles to
hillslopes, and geomorphic processes, which act tangen-
tially to hillslopes (Phillips 1995). When there is no
lag between debris production and its availability for
removal, regolith thickness at a point along a hillside
displays chaotic dynamics. On the other hand, when a
time-lag is present, regolith thickness is stable and non-
chaotic. The emergence of landscape stability at broad
scales may therefore result from time-lags in different
processes. Where regolith production is slow, and ero-
sion even slower, stagnation might occur. Even so, the
conditions necessary for the first two mechanisms to pro-
duce landscape stagnation would surely be required for a
landscape to maintain stability for hundreds of millions
of years.
If substantial portions of landscapes are indeed stag-
nant and hundreds of millions of years old, the implica-
tions for process geomorphology are not much short of
sensational. It would mean that cherished views on rates
of denudation and on the relation between denudation
rates and tectonics would require a radical revision, and
the connections between climate and landforms would
be even more difficult to establish.
EVOLVING LANDSCAPES
Landscape cycles
Several geomorphologists believe that landscape history
has been
cyclical
or
episodic
. The Davisian system of
landscape evolution combined periods dominated by
the gradual and gentle action of geomorphic processes
interrupted by brief episodes of sudden and violent tec-
tonic activity. A land mass would suffer repeated 'cycles
of erosion' involving an initial rapid uplift followed by
a slow wearing down. The Kingian model of repeated
pediplanation envisaged long-term cycles, too. Remnants
of erosion surfaces can be identified globally (King 1983).
They correspond to pediplanation during the Jurassic,
early to middle Cretaceous, late Cretaceous, Miocene,
Pliocene, and Quaternary times (Table 15.2). However,
King's views are not widely accepted, and have been
challenged (e.g. Summerfield 1984; Ollier 1991, 93).
A popular theme, with several variations, is that the
landscape alternates between stages of relative stability
and stages of relative instability. An early version of
this idea, which still has considerable currency, is the
theory of biostasy and rhexistasy (Erhart 1938, 1956;
cf. Butler 1959, 1967). According to this model, land-
scape change involves long periods of
biostasy
(biological
equilibrium), associated with stability and soil develop-
ment, broken in upon by short periods of
rhexistasy
(disequilibrium), marked by instability and soil erosion.
During biostasy, which is the 'normal' state, streams
carry small loads of suspended sediments but large loads
of dissolved materials: silica and calcium are removed
to the oceans, where they form limestones and chert,
