® .Geomorphology 32 2000 109­128
Channel instability in a forested catchment: a case study from
Jones Creek, East Gippsland, Australia
Tim J. Cohen )
, Gary J. Brierley 1
School of Earth Sciences, Macquarie Uniersity, North Ryde, NSW 2109, Australia
Received 22 October 1998; received in revised form 15 May 1999; accepted 12 July 1999
Abstract
Jones Creek, a forested sub-catchment of the Genoa River, Victoria, Australia has experienced channel metamorphosis
induced by a series of floods since 1971. In the flood of record, in 1971, the Genoa River widened by up to two times at the
confluence with Jones Creek. This effectively shortened the course of Jones Creek, resulting in a lagged tributary response.
Incision and bed steepening during floods in 1975 and 1978 triggered significant changes in channel form along Jones
Creek. Channel changes between 1972­1997 resulted in a four-fold increase in cross-sectional area. Channel depth increased
by up to 1.5 m, but has subsequently refilled by around 0.6 m. Initially, incision resulted in increased stream power as a
result of increased mean depth, while sinuosity was maintained. This was followed by channel widening, a reduction in
sinuosity and a continued increase in slope. Estimated stream power remains high as the channel continues to laterally
adjust. This study highlights the dynamic nature of tributary­trunk stream relationships in a cut-and-fill landscape,
demonstrating how trunk stream adjustments can induce profound tributary instability in a forested subcatchment. Channel
widening of the trunk stream primed the tributary for change. Bedlevel incision and increases in bedslope breached the
threshold of landscape stability in this steep alluvial tributary. Three periods of channel change can be identified, reflecting
the complex response of the system to channel incision. Rather than developing simple cutoffs, Jones Creek now exhibits a
range of lateral adjustment and realignment features. Responses to external disturbance provide insights into `natural'
recovery mechanisms in this forested setting. Lateral adjustment and associated sediment deposition have been accompanied
by rapid rates of vegetation colonisation and stabilisation of realignment features, effectively reducing the volume of
sediment that is available to be reworked through the channel network. q 2000 Elsevier Science B.V. All rights reserved.
Keywords: channel instability; bed-level incision; forested catchment; channel recovery; riparian vegetation; Australia
)
Corresponding author. School of Geosciences, University of
Wollongong, Wollongong, NSW, 2522, Australia. Fax: q61-02-
4221-4250.
® .E-mail addresses: tjc12@uow.edu.au T.J. Cohen ,
® .gbrierli@laurel.ocs.mq.edu.au G.J. Brierley .
1
Fax: q61-02-9850-8427.
1. Introduction
To date, nearly all studies of the geomorphology
of the rivers in southeastern Australia have been
framed in landscapes which have been dramatically
altered in the period following European settlement.
As such, our understanding of controls on the charac-
ter and behaviour of these rivers may be skewed by
0169-555Xr00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
® .PII: S0169-555X 99 00079-3
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128110
the significant alterations in boundary conditions that
these landscapes have experienced. For example, in
virtually all instances in which channel metamorpho-
sis has been documented, there has been extensive
clearance of riparian and floodplain vegetation, mod-
ifying the geomorphic effectiveness of flood events
® .Brooks and Brierley, 1997 . Although the Aus-
tralian landscape is characterised by fluvial systems
that have been destabilised since European settle-
ment, there is little documented evidence of the
mechanisms that induce the breaching of thresholds.
In this study, the term channel stability refers to a
change in channel pattern and form process associa-
tion as a result of an altered sediment load or dis-
® .charge regime Chorley et al., 1984 . This study
presents data on the metamorphosis of an alluvial
tributary in a relatively undisturbed forested setting
and isolates controls on landscape stability.
Our understanding of channel instability largely
stems from theoretical and experimental work and a
range of case studies carried out in the northern
®hemisphere Schumm and Lichty, 1963; Schumm et
al., 1984; Osterkamp and Costa, 1987; Simon, 1989,
.1992; Schumm, 1993 . Most of these studies have
been performed in landscapes that have been sub-
jected to a history of human impact. It is only
recently that literature on fluvial geomorphology has
highlighted the unique character of the Australian
®fluvial landscape cf., Tooth and Nanson, 1995;
.Brierley et al., 1996 . The theoretical and docu-
mented processes on channel instability exemplified
in the northern hemisphere need to be tested in an
Australian setting. This issue is compounded by the
fact that very few, if any, studies of channel instabil-
ity have demonstrated how fluvial systems in tem-
perate environments unaffected by human activity
respond to large `natural' disturbances. For example,
® . ® .while Prosser et al. 1994 , Brierley and Fryirs 1998
® .and Fryirs and Brierley 1999 have demonstrated
the distinct cut and fill nature of certain alluvial
landscape settings in southeastern Australia, it has
been difficult to ascertain the character of incisional
processes in forested settings. To our knowledge, no
studies have documented the character and mecha-
nisms of channel incision and landscape recovery in
catchments with an intact riparian forest. While a
significant volume of literature from the northern
hemisphere provides a basis with which to assess the
role of riparian vegetation and woody debris on
®channel dynamics in a forested setting cf., Keller
.and Swanson, 1979; Nakamura and Swanson, 1993 ,
much of this work has been in fluvial settings which
are fundamentally different to landscape settings in
southeastern Australia.
Jones Creek is a forested tributary of the Genoa
catchment, which straddles the New South Wales­
Victoria border in southeastern Australia. Since 1972
Jones Creek has undergone significant changes in
channel geometry and planform. In this period, the
catchment has been subjected to `natural' distur-
bances such as a series of floods and the 1983 Ash
Wednesday fire. This study documents how Jones
Creek has responded to these `natural' disturbances.
Insights are provided into recovery mechanisms that
follow large scale channel instability in settings with
an intact riparian zone. The role of within-channel
vegetation, large woody debris and sediment reten-
tion are examined as integral components for the
onset of channel recovery, thereby adding to the
theoretical notions on channel behaviour following
incision.
2. Regional setting
Jones Creek drains a 31 km2
subcatchment of the
® .Genoa River in far East Gippsland Fig. 1 . Up-
stream of Jones Creek confluence, the Genoa River
is bedrock-controlled, but the confluence zone itself
® .termed Wangarabell Reach is characterised by an
area of valley expansion resulting in discontinuous
floodplain formation. The geology of the upper part
of Jones Creek catchment comprises Devonian gran-
ites and Ordovician metasedimentary rocks. Much of
the middle part of the catchment comprises Tertiary
sediment consisting of unconsolidated coarse sand
and gravels with clay rich sands.
The lower 3.5 km of Jones Creek has an alluvial
channel with a sinuosity of 1.58, mean bankfull
width of approximately 33 m, and mean depth of 2
m. Although the forested floodplain increases in
width downstream from 50 m to 500 m towards the
® .Genoa River confluence Fig. 2 , there are no sys-
tematic downstream changes in channel width, mean
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 111
Fig. 1. Location map showing Jones Creek within Wangarabell Reach in the Genoa River catchment.
depth, cross sectional area, sinuosity or bed material
size. The average floodplain slope in the lower sec-
tion of the catchment is 0.0075 mrm, while overall
channel slope within the study area is 0.0045 mrm.
Jones Creek is an emphemeral system and is charac-
terised by widespread bank erosion, with numerous
examples of bank collapse and tree fall. Various
abandoned channels are evident near the channel and
on the floodplain.
Jones Creek is a mixed load system, transporting
silt, sand and gravel. At present, Jones Creek trans-
ports significant volumes of sediment forming sand
sheets, point bars and bank attached bars. The bed
® ® .material size assessed using the Wolman 1954
.technique averages 40 mm in the upper 2.5 km of
the alluvial reach, but increases in size up to 320 mm
adjacent to the Genoa River confluence, where older
Genoa River gravels are reworked. The presence of
these larger clasts indicates that the Genoa River has
adjusted its position through the confluence zone in
the past, with these coarser materials extending up to
400 m upstream from the confluence.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128112
Bank materials are extremely variable, ranging
from clay-rich uniform bank exposures to medium
sands and gravels in channel fill sequences. Bankside
vegetation in the study area often has exposed root
zones which are well above present-day bedlevels.
®Large amounts of large woody debris LWD --
defined as wood )10 cm in diameter and )1 m in
.length and log jams are evident within the channel.
®Based on rainfall records from Wroxham 10 km
. ®NE of Jones Creek and Timbillica 25 km NE of
.Jones Creek , the mean annual rainfall at Jones
Creek is approximately 900 mm. Continuous flood
discharge records for the Genoa River downstream
of Wangarabell are available from mid 1971 on-
® .wards the `gorge' -- station 221210 . Based on the
annual series from 1972­1996, the Genoa River has
a steep flood frequency curve as indicated by a flash
flood magnitude index of 0.74, which is high by
®world standards Erskine 1992, 1993; Erskine and
.I.D.&A., 1997 . Based on monthly and maximum
48-h rainfall at Wroxham since 1971, periods of
above average rainfall can be discerned for 1971­
1978, 1983­1988 and 1991­1995. These periods of
high rainfall have been related to anecdotal and
historical evidence to establish a flood record for
Jones Creek.
Approximately 90% of Jones Creek catchment
falls within Coopracambra National Park. There are
no records of extensive timber extraction from the
catchment in the last 25 years. Prior to this, the
catchment has had a history of small scale timber
extraction in the form of spot milling. The catchment
is dominated by regrowth dry sclerophyll vegetation
on the hillslopes and floodplain with more mesic
communities in the gullies and headwaters. Species
composition across the floodplain shows consider-
able spatial variability with a mosaic of vegetation
patches. The tall open forest is dominated by Euca-
® .lyptus iminalis subsp. iminalis Manna Gum , E.
® . ®elata River Peppermint and E. baueriana Blue
.Box . The mid-storey in the study area is dominated
® .by Acacia mearnsii Black Wattle , A. melanoxylon
® . ®Blackwood and Pomaderris aspera Hazel Po-
Fig. 2. Study area within Jones Creek, highlighting the contempo-
rary planform and position of cross sections along with the 1996
longitudinal profile.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 113
.maderris . Ground cover is dominated by Poa ensi-
® .formis Sword Tussock-grass , Lomandra longifolia
® .Spiny-headed Matrush and Blechnum nudum
® .Fishbone Water fern .
®Based on tree ages dendrochronological evidence
.on A. melanoxylon , rainfall records from the 1800s,
® .and historical information, Chesterfield et al. 1990
estimated that significant fires occurred in Jones
Creek catchment in the early 1800s, late 1800s or
early 1900s, 1952 and 1983. The intensity and per-
centage of the catchment burnt in fires prior to 1983,
however, is unknown. The 1983 Ash Wednesday fire
was the most intense fire on record in East Gipps-
land, and brought about profound changes to vegeta-
®tion composition in many areas Chesterfield et al.,
.1990; Loyn et al., 1992 . All of Jones Creek catch-
ment was intensely burnt in 1983, including warm
®temperate rainforest in the headwaters D. Cameron,
.pers. comm. . After the 1983 fires the density of
Acacia sp. and Eucalyptus sp. increased dramati-
cally in the headwater areas, with densities ranging
y1 y1 ®from 250 ha to 100,000 ha Chesterfield et al.,
.1990 . Prior to the 1983 fire, vegetation in the study
®area was of a more mesic association D. Cameron,
.pers. com. . Riparian communities were significantly
affected by the 1983 fires, with a shift towards a
more sclerophyll based association. This resulted in a
floodplain vegetation community that is now charac-
terised by areas with high stem density of dry sclero-
phyll species. There are no quantifiable data on the
direct impacts of the 1983 fire on channel processes
such as sediment supply or changes to the hydrologi-
cal regime.
3. Methods
®Aerial photographic interpretation at a scale of
.1:4000­1:5000 , along with field mapping of geo-
®morphic and vegetation attributes including den-
.drochronology of A. mearnsii growing in cutoffs ,
provide the basis for documenting the nature and
timing of channel changes in Jones Creek since 1967
® .see Cohen, 1997 for further details . Many of the
channel cutoffs in the study area are not fully pre-
served. Rather, they are infilled, inset features at the
margins of the wide channel. These features are not
classic channel cutoffs and have been termed re-
alignment features. Realignment features reflect lat-
eral channel adjustment, in contrast to channel cut-
offs which represent bends or sections of the channel
that have been separated from the contemporary
® .channel Fig. 3 .
Based on their morphology and position, two
categories of palaeochannels were identified in the
field. These are termed the 1983 and 1972 channels.
Identification of 1983 cutoff channels and realign-
ment features was initially made from the 1983
® .aerial photo taken after the fire . These were con-
firmed in the field by their dimensions and eroded
asymmetric cross-sectional morphology, their prox-
imity to the contemporary channel, and the lack of
vegetation older than 14 years. The 1972 channels
are characterised by smaller hydraulic dimensions
with symmetrical cross-sections. These channels are
not clearly shown on any of the photographs, al-
though the 1967 photograph indicates fragments of
their presence and rough dimensions. The 1972
channels are considered to be representative of chan-
nel form in Jones Creek prior to the latest period of
channel change. Cross-sections of 1972 and 1983
palaeochannels were surveyed to measure their cross
sectional area.
Prior bedlevels of Jones Creek were determined
using a range of indicators. Most evidence is pro-
vided by the depth of abandoned sections of the
1972 channel and 1983 realignment features and
channel cutoffs. This has been defined by the marked
boundary between fine grained channel fill and coarse
® .bedload cf., Erskine et al., 1992 . A further indica-
tor of prior bedlevel is provided by a series of gravel
lag outcrops perched above the contemporary chan-
nel bed. Finally, prior water tables evidenced by
gleying and blocky fabric on bank exposures are
®indicative of bed degradation see Cohen, 1997 for a
.full description of prior bedlevel assessment . Each
of these indicators was surveyed in to the present
thalweg bedlevel to assess bedlevel adjustments
throughout the study area.
Eighteen channel cross sections were surveyed
along the study reach. Twelve of these sections were
resurveys of cross-sections emplaced by the then
East Gippsland River Management Board in 1993.
The longitudinal profile of Jones Creek was also
resurveyed at 30 m increments along the thalweg
® .Fig. 2 .
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128114
Fig. 3. Schematic example of a lateral realignment feature in contrast to a section of the 1983 channel that has been separated from the
channel. Realignment features reflect lateral adjustment of the channel planform and shift in the focus of bend migration.
Sedimentological attributes of bed and bank mate-
rials were analysed at each of the surveyed cross
sections. Samples were wet sieved at 63 mm to
determine the percent siltrclay. Sediment storage on
the channel bed was assessed at three positions at
each cross section using a bed-level probe, with the
depth to a coarse gravel lag used to assess volume of
sediment stored on the channel bed. Sediment stor-
age in all realignment features and channel cutoffs
was also assessed.
4. Channel changes on the Genoa River and Jones
Creek, 1967­1998
Aerial photographs from 1967, 1972, 1986, 1992
and 1994 indicate that the Genoa River and its
tributaries in Wangarabell Reach have undergone
considerable change in form since 1971. These
changes were initiated by a sequence of floods in the
1970s.
The enlarged 1967 aerial photograph of Genoa
River­Jones Creek confluence shows no evidence
for active erosion along the Genoa River. At this
time, the Genoa River was around 35­40 m wide,
and was characterised by regularly spaced, large
® .pools see Fig. 4A . The tributary confluence was
dissected, and there were a series of vegetated bars
up and downstream of the confluence.
In 1967, Jones Creek was a small, indistinct
tributary with dense riparian and floodplain vegeta-
tion. Bankfull widths measured from the 1967 air
photo range between 12­15 m. Field evidence con-
®firms this, with measured palaeochannels termed
.1972 channels in this study having bankfull widths
of 10­18 m.
In 1971, the Genoa River experienced the largest
flood on record since European settlement. This
flood had an estimated peak instantaneous discharge
of 2611 m3
sy1
, 12.4 times greater than the mean
® .annual flood Erskine, 1992, 1993 . The 1971 flood
resulted in floodplain inundation, sand splay deposi-
tion on the floodplain, extensive channel margin
()T.J.Cohen,G.J.BrierleyrGeomorphology322000109­128115
Fig. 4. Genoa River at Wangarabell Reach showing the changes in channel morphology from 1967 to 1979. Arrow denotes position of Jones Creek confluence. Note the
destabilisation of Jones Creek by 1979. Flow is from left to right.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128116
stripping and changes in channel planform in the
lower reaches of the Genoa River. Around 5 ha. of
floodplain and channel margin land were removed
between Jones Creek confluence and Wangarabell
® . ®Creek confluence Erskine, 1992, 1993 see Fig.
.4B . Channel widening of the right bank occurred at
Jones Creek confluence, resulting in bankfull dimen-
sions of 106­132 m. This indicates a 230% increase
in channel width since 1967. Vegetated and non-
vegetated bars at Jones Creek confluence were com-
pletely removed, while vegetation was stripped from
bars upstream of, and on the bank opposite to, the
confluence of Jones Creek.
Although the 1971 flood resulted in extensive
changes in channel form on the Genoa River, the
®tributaries within Wangarabell Reach Jones Creek
.and Wangarabell Creek remained stable. However,
the geomorphic responses of the Genoa River and
associated tributaries in Wangarabell Reach to a
series of closely spaced, moderate to high magnitude
floods in 1971, 1975 and 1978 were spatially and
temporally variable. The 1975 and 1978 floods on
the Genoa River, which had estimated recurrence
intervals of 6.5 and 13 years, respectively, main-
tained erosion along the Genoa River, and resulted in
the destabilisation of the tributaries in Wangarabell
Reach, including Jones Creek and Wangarabell
Creek. Fig. 4C shows the Wangarabell Reach in
1979, highlighting the widened character of the
Genoa River with continued bank erosion at the
confluence with Jones Creek. It was not until the
subsequent floods of 1975 and 1978 that tributary
instability occurred
It is difficult to ascertain a definite planform for
Jones Creek prior to it becoming destabilised. How-
Table 2
1972 bedlevel indicators
® ® .1972 Palaeochannel 1972 Channel bedlevel q m :
® .location above present day thalweg; y m :
.below present day thalweg
® .Urs of XS 10 Auger q0.964
® .Urs of XS 10 Auger q0.795
® .Adjacent to XS 10 Pit q0.721, q0.764
® .Drs of XS 10 Auger q0.886
® .Drs of XS 10 Auger q0.807
® .Drs of XS 10 Auger q0.788
® .Drs of XS 10 Pit q0.722
® .Urs of XS 4 Auger q1.004
® .Urs of XS 4 Auger q1.448
® .Drs of XS 2 Auger q0.611
a
Gravel lag urs of XS 2 q0.821
a
Gravel lag urs of XS 2 q0.628
Mean 1972 bedlevel q0.843
a
Outcropping gravel lag in channel.
Point of bedlevel is taken as the bedload channelrfill contact
boundary.
ever, insights can be gained from a series of narrow
and symmetrical palaeochannels, which have a sinu-
osity of approximately 2.1­1.71, mean bankfull
®width of 15.1 m and a mean depth of 1.15 m see
.Table 1 . Supporting evidence for these channel
dimensions is provided from anecdotal sources. For
® .example, D. Cameron pers. comm. identified Jones
Creek as having a channel that was small and indis-
tinct in the mid 1970s, with permanent pools and
dense mesic riparian vegetation. All bedlevel indica-
tors from 1972 paleochannels are above the bedlevel
® .of the contemporary channel Table 2 , ranging from
0.61­1.45 m above present day bedlevel, with a
mean of 0.84 m.
Table 1
Channel dimensions for Jones Creek 1972­1997
Mean 1972 Channel 1983 Channel 1993 Channel 1997 Channel % Change 1972­1997
2® .Cross sectional area m 17.2 35.9 60.1 66.1 q385
® .Bankfull width m 15.1 16 32.3 33.3 q120
® .Mean bankfull depth m 1.15 2.23 1.86 1.99 q73
® .Max. bankfull depth m 3 4.1 3.8 3.5 q16
WrD 5 3.9 8.5 9.5 q78
Sinuosity 2.1­1.71 1.73 1.56 1.58 y7.6
WrD ratio calculated using maximum channel depth.
Increase in mean depth between 1993 and 1997 is a function of largeincreases in cross sectional area at cross section 10 and 8, rather than
increases in absolute depth.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 117
Channel expansion of the Genoa River in the
1971 flood effectively shortened the length of Jones
Creek by 75­100 m. On the 1979 aerial photograph
® .Fig. 4B Jones Creek had estimated bankfull dimen-
sions that range from 15­31 m, with a sinuosity of
1.71. The study area was actively eroding, with some
reaches characterised by within-channel bars and
debris accumulations. It is clear, however, that chan-
nel changes between 1972­1979 were not spatially
uniform.
®On the 1983 aerial photograph taken after the
.fire the channel was actively eroding its banks, with
a sinuosity of 1.73 and bankfull dimensions that
ranged from 18­32 m. Channel width was highly
variable, with areas undergoing cutoffs having wider
bankfull conditions. Measured mean bankfull width
was 16 m, while mean bankfull depth was 2.23 m
® .Table 1 . Detailed field analysis of channel cutoffs
and realignment features indicate that the active
® .1983 channel was asymmetrical with near vertical
upper banks, in stark contrast to the symmetrical
1972 channel.
The period of tributary instability between 1972­
1983, as constrained by the aerial photograph record,
has been termed the first phase of channel change in
Jones Creek. Following a series of inferred floods on
Jones Creek between 1972­1979, the mean depth of
Jones Creek increased by 93% as a result of bedlevel
incision, while cross sectional area increased by
108%. Although erosion was widespread in this first
phase of channel change, sinuosity was maintained,
and the width:depth ratio decreased as the channel
deepened.
Between 1983­1992, cutoffs and channel realign-
ment occurred throughout the study area with a
reduction of sinuosity from 1.73 in 1983 to 1.56 in
1992, most likely during a series of floods between
1983­1988. Dendrochronological evidence confirms
that colonisation of 1983 channel cutoffs and re-
alignment features occurred from late 1984 to early
1985. Bedlevel indicators from 1983 realignment
features in the middle of the study area indicate that
1983 bedlevels ranged from 0.236­1.515 m below
present day thalweg bedlevel, with a mean of 0.66 m
® .Table 3 . Mean channel width in this period in-
creased by approximately 100% to 32.3 m and cross
® .sectional area increased by a further 67% Table 1 .
Mean depth decreased by approximately 17% as the
Table 3
Bedlevel indicators for 1983 channels
® ® .1983 Realignment featurer 1983 Channel bedlevel q m :
® .cutoff location above present day thalweg; y m :
.below present day thalweg
® .Urs of XS 10 Auger y0.368
® .Urs of XS 10 Pit y0.27
® .Drs of XS 10 Auger y0.76
® .Drs of XS 10 Auger y1.515
® .Urs of XS 9 Auger y0.236
® .Drs of XS 9 Pit y0.673
® .Urs of XS 7 Auger y0.687
Adjacent to XS 7 y0.8
Mean 1983 bedlevel y0.66
Point of bedlevel is taken as the bedload channelrfill contact
boundary.
channel became wider and shallower along with
large increases in the width:depth ratio. These
changes are referred to as the second phase of chan-
nel change in Jones Creek. Fig. 5 highlights the
character of planform changes in a segment of the
study area between 1972 and 1997.
Aerial photographs from 1992­1994 indicate lit-
tle change in channel planform of Jones Creek. The
last three months of 1995 were characterised by high
maximum monthly rainfall producing a bankfull
event on Jones Creek on the 5th and 6th of Decem-
ber with an estimated discharge of around 130 m3
sy1
. The 1995 flood induced little change in plan-
form in the study reach. A flood in June 1998, which
was approximately three quarters bankfull, resulted
in a cutoff in the upper part of the study area. The
period from 1992 to the present has been charac-
terised by continued channel expansion, bar deposi-
tion and lateral adjustment and is identified as the
third phase of channel change in Jones Creek. Bank-
full widths and cross sectional area increased slightly
® .with small fluctuations in mean depth see Table 1 .
Straightening is still operative with 1­2 cutoffs in
the process of occurring. The contemporary be-
haviour of Jones Creek is characterised by localised
increases in cross sectional area, with shifts in the
erosional foci being the dominant process as exem-
plified by Fig. 6.
Data from Tables 2 and 3, summarised in Fig. 7,
indicate that Jones Creek was characterised by
bedlevel incision in the first phase of channel change
()T.J.Cohen,G.J.BrierleyrGeomorphology322000109­128118
Fig. 5. Planform changes in the middle of the study area of Jones Creek between 1972 and 1997, showing nature of lateral adjustment as Jones Creek progressively straightened.
Symbols indicate location of bedlevel indicators.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 119
Fig. 6. Cross sectional changes in the middle section of Jones
Creek between 1993 and 1997 highlighting localised increases in
cross sectional area and lateral adjustment of channel geometry.
® .1972­1983 , but that aggradation occurred from
1983­1997. These data on bedlevel adjustments are
seemingly confirmed by analysis of prior water ta-
® .bles Table 4 , which range from 0.48 to 1.83 m
above present day bedlevel, with a mean of 1.29 m.
This is approximately 45 cm above the 1972 mean
bedlevel, indicating the range of water level fluctua-
tions in the 1972 channel. While these data display a
range of elevations above the 1972 mean bed eleva-
tion, they still confirm that the mean watertable was
above the mean 1972 bedlevel, indicating watertable
lowering in the incision process. Based on these
bedlevel assessments, it is estimated that channel
changes between 1972­1983 resulted in up to 1.5 m
of bedlevel incision, whereas an average of 0.66 m
of bed aggradation occurred between 1983 and 1997.
The assessment of bedlevel indicators highlighting
channel incision is supported by aerial photographic
evidence of channel straightening and channel short-
ening on Jones Creek between 1972 and 1983.
While periods of channel change can be identified
for Jones Creek, changes on the Genoa River since
1979 have involved extensive deposition of bank-at-
tached bars opposite the confluence of Jones Creek.
The other dominant process in this period has been
the recovery of instream and bank vegetation follow-
®ing the 1983 fires cf., Schumm and Lichty, 1963;
.Friedman et al., 1996 . This has resulted in a nar-
rower channel of similar character to that which was
evident in 1967.
Finally, detailed field survey of channel dimen-
sions, along with augering and probing of sediment
thickness on the channel bed and in realignment
features and cutoffs, has enabled a summary assess-
ment of the alluvial sediment budget of the study
® .reach to be performed see Table 5 . From the three
periods of channel change identified in Jones Creek,
between 1972 and 1997, it is estimated that a total of
171,100 m3
of material have been eroded. Sediment
analysis of bank material suggests an estimated mean
ratio of sand to mud of 58:42 and it is likely that the
mud fraction has been predominantly flushed through
® 3 .the system representing 71,862 m of sediment . Of
the remaining 99,200 m3
of sediment, an estimated
29,950 m3
of sand are stored in point bars and
realignment features, with another 3170 m3
of sand
stored on the bed of Jones Creek. The total of 33,120
m3
of material stored within the channel zone repre-
sents 33.4% of the sand fraction eroded during the
three periods of channel change. Of this volume,
65% is consolidated in vegetated realignment fea-
()T.J.Cohen,G.J.BrierleyrGeomorphology322000109­128120
Fig. 7. Jones Creek 1996 longitudinal profile showing location of prior bedlevels highlighting that prior to destabilisation bedlevels were on average 0.84 m above present day
thalweg elevation, while 1983 bedlevels were on average 0.66 m below present day bedlevel.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 121
Table 4
Prior watertable bedlevel indicators
® ® .Prior watertable Prior watertable q m :
® .location above present day thalweg; y m :
.below present day thalweg
Urs of XS 10 q1.814
Urs of XS 10 q1.83
Urs of XS 10 q1.381
Urs of XS 10 q1.467
Urs of XS 9 q1.13
Urs of XS 9 q0.771
Drs of XS 9 q0.478
Urs of XS 2 q1.436
tures. Following incision, an estimated 66,000 m3
of
sand have been flushed through Jones Creek into the
Genoa River, resulting in increased rates of aggrada-
tion in the trunk stream. Perhaps the most significant
aspect of this alluvial sediment budget is that re-
alignment features and cutoffs within the enlarged
channel of Jones Creek have already trapped up to
33% of sands released upon channel expansion.
Trapping of these deposits is considered to form the
primary platform for subsequent channel recovery.
5. Channel instability in a forested catchment
The three periods of channel change identified
along Jones Creek since 1972 are shown schemati-
cally in Fig. 8. It must be noted that the processes
documented within these periods have not occurred
in isolation, but have been part of a continuum of
change as a result of alterations to variables within
the fluvial system. Prior to destabilisation Jones
Creek had a symmetrical cross-sectional form with a
mean width of 15.1 m and cross sectional area of
2 ® .17.2 m Fig. 9 . By 1983, bedlevel lowering had
increased mean channel depth and mean channel
cross-sectional area around 100%. This resulted in an
estimated increase in bedslope from 0.0033­0.0037
®mrm in 1972 to 0.0044 mrm in 1983 around
. ® .30% . Estimated specific stream power v values
for Jones Creek prior to destabilisation ranged from
y2 ® .30­35 Wm Table 6 . This falls within approxi-
® .mations by Nanson and Croke 1992 for medium
energy floodplains and is on the threshold of channel
® .`stability' presented by Brookes 1990 . During the
first phase of instability, estimated specific stream
power increased by up to five times, as bedlevel
lowering increased mean depth, greatly increasing
®the capacity of the system to transport sediment cf.,
.Brizga and Finlayson, 1990 .
Between 1983­1992 the mean channel width of
Jones Creek increased by a further 70%, while mean
depth was reduced by around 16%. As the bedslope
® .increased during channel incision Phase 1 , in-
creased energy within Jones Creek eroded the bed
and banks, resulting in the input of large volumes of
sediment. The channel was unable to maintain a high
sinuosity given the high stream powers and the
increased sediment load. Hence, the channel straight-
Table 5
Volumes of sediment excavated in the periods of channel change in Jones Creek 1972­1997
Period of Total volume Proportion of Dominant process Volume Volume Location of Proportion
channel of sediment sand size of siltrmud of sand sediment of stored
3 3 ) 3® . ® . ® .change eroded m sediment flushed m stored m storage sediment
3® .eroded m available to
be reworked
3® .m
1972­1983 65,400 37,932 bedlevel incision 27,468 ? ? ?
1983­1993 84,700 49,126 channel widening 35,574 ? ? ?
1993­1997 21,000 12,180 lateral activity 8820 29,950 and 3170 realignment 11,592
features and
bed storage
1972­1997 171,100 99,238 NA 71,862 33,120 NA 11,592
? refers to an unknown volume of sediment or unknown location of sediment storage.
® .Sediment budget is based on the assumption that the mud fraction has been flushed through the system Sandrmud ratios58:42 .
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128122
Fig. 8. Schematic diagram of the 3 phases of channel change in Jones Creek from 1972 to 1997.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 123
Fig. 9. Downstream view of a 1972 channel in Jones Creek, prior to destabilisation. Note the narrow symmetrical channel.
ened its course, reducing its sinuosity from 1.71 in
1983 to 1.56 in 1992, thereby further increasing
bedslope. This period of channel adjustment is termed
Phase 2. Channel changes in this phase can be
related to known floods in 1983, 1985, 1988 and
1992. As a result of these planform changes, the
width:depth ratio of Jones Creek in 1992 had in-
creased by 74% since 1972. The dimensions of the
enlarged channel virtually preclude floodplain inun-
dation, effectively concentrating the geomorphic im-
pacts of high magnitude events within the channel
®zone Brizga and Finlayson, 1990; Nanson and
.Croke, 1992 .
® .Phase 3 1992 to present has been a period of
ongoing channel expansion and straightening, with a
® .reduction in sinuosity to 1.5 Fig. 10 . Estimated
specific stream powers are approximately four times
higher than in 1972. There has been continued aggra-
Table 6
Changes in estimated bedslope, discharge and streampower in Jones Creek 1972­1997
Mean 1972 Channel 1983 Channel 1993 Channel 1997 Channel % Change 1972­1997
Slope 0.0033­0.0037 0.0044 0.0047 0.0046 q24­39.4
® . ® .Lc estimated length of channel m 4000­3500 3330 3070 3000 y14­25
® .Estimated drop in elevation m 13.06 14.56 14.4 13.9 y6.5
3 y1® .Estimated bankfull discharge m s 15.1 70 118.4 132.7 q778
Estimated Manning's n 0.07 0.05 0.05 0.05 y28.6
y2® .Bagnold's stream power W m 29.9­35.4 185.7 171.8 174.8 q394­485
Discharge estimates used for Bagnold's stream power are based on the Manning's equation. q indicates an increase in a channel attribute.
Slope estimates for 1972­1983 based on changes to elevation associated with bedlevel indicator data.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128124
Fig. 10. Upstream view of the contemporary channel in Jones Creek highlighting the widened nature of the current channel.
dation of realignment features and bed-level aggrada-
tion, with localised bed-level degradation. These
changes in channel form are a result of known floods
in 1995 and 1998.
The change in channel pattern of Jones Creek
conforms to classical notions of complex response
® .within incised channels cf., Schumm et al., 1984 .
Initial incision increased channel depth, bedslope and
® .unit stream power Phase 1 . As degradation slowed,
channel widening occurred to disperse available en-
®ergy and achieve a degree of stability Jackson and
.Beschta, 1984; Schumm et al., 1984; Simon, 1992 ,
resulting in a reduction in sinuosity. This supplied
large volumes of sediment to the channel as part of a
negative feedback system, where the channel ag-
®grades as it adjusts to new energy gradients Phase
.2 . The development of alternate bars and point bars
enhances the widening process by promoting lateral
adjustment. This is the initial step in the eventual
®development of a sinuous inner channel Schumm et
.al., 1984 . In Jones Creek, the present behaviour is
characterised by continued expansion, lateral adjust-
ment, downstream shifts of the erosional foci and the
® .deposition of alternate bars Phase 3 .
The channel-floodplain relationship along lower
Jones Creek has been fundamentally transformed
since 1972. Prior to incision, channel changes in
Jones Creek resulted in cutoffs forming and then
infilling via overbank and backfilling processes. Sed-
iment composition of these channel fills is highly
variable, ranging from 25­78% mud. However,
channel fill samples from 1983 realignment features
and cutoffs show mud fractions that range from
® .3­5% cf., Erskine et al., 1992 . The confinement of
flood flows within the enlarged channel, and associ-
ated increases in stream power, have resulted in
realignment features and cutoffs which are com-
prised of medium-coarse sands. This is a likely
consequence of an increase in the grain size that is
transported as suspended load.
It is not possible to accurately assess rates of
channel fill in Jones Creek prior to channel incision.
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 125
However, field observations of shallow depths of fill
® .in 1972 channels 1­1.5 m contrast starkly with the
maximum depth of aggradation in 1983 realignment
® .features and cutoffs 2.7­3.0 m . This indicates that
there has been a maximum annual aggradation rate
in realignment features of 190­215 mm yry1
since
1983. However, aggradation rates in Jones Creek are
episodic, due the ephemeral nature of the system.
This suggests that the realignment features have
formed rapidly in a small number of events since
1983. Rapid deposition at the channel margin has
formed an incipient floodplain which is now sta-
bilised by A. mearnsii and P. ensiformis. This has
resulted in coarser grained facies proximal to the
channel throughout the channel network.
Channel adjustments along Jones Creek reflect the
sensitive balance between a forested subcatchment
® .Jones Creek and a trunk stream prone to channel
® .expansion Genoa River . It is uncertain how charac-
teristic the style of trunk stream expansion of the
Genoa River in 1971 was, or whether it represents
accelerated rates of channel expansion as a result of
human impacts in Wangarabell Reach. Erosional
scars on the Genoa floodplain suggest that periodic
reworking of discontinuous floodplains is a recurrent
component of floodplain processes within this mixed
load system. However, given the presence of numer-
ous palaeochannels along Jones Creek, with dimen-
sions akin to the 1972 channels, it is reasonable to
assume that this style of tributaryrtrunk stream in-
stability occurs infrequently.
Between 1972­1983 approximately 1.5 m of bed
degradation occurred along lower Jones Creek as a
secondary response to trunk stream controls. The
destabilisation of the fully vegetated riparian zone of
Jones Creek was preconditioned by the 1971 flood,
which removed the tributary confluence bar and
shortened the length of Jones Creek. Catastrophic
®widening andror bedlevel lowering i.e., changes to
.bed sediment storage in the Genoa River initiated
dramatic changes in channel morphology of the allu-
vial Jones Creek tributary. As the bed-level of Jones
Creek adjusted to new energy gradients, the system
became destabilised. Such a threshold response has
been documented on tributaries of the Mississippi
® .River Schumm et al., 1984 , and elsewhere in Aus-
®tralia e.g. the Avon River, Victoria; Erskine et al.,
.1990 .
A discernible lag of up to 7 years was evident
between widening of the trunk stream and destabili-
sation of the tributary, conditioned by the timing of
®subsequent flood events on the tributary cf., Schumm
.et al., 1984; Kochel, 1988; Erskine et al., 1990 .
Based on rainfall records, localised flood intensity in
Jones Creek is inferred for 1975 and 1978. Indeed,
the 1978 flood event along Genoa River reflected the
highest 48 h rainfall over the 25 year record. It is
inferred that upstream-progressing degradation in
Jones Creek was the dominant control in initiating
bedlevel incision and is confirmed by tributaries
within the catchment displaying knickpoints. As
® .Galay 1983 noted, however, upstream-progressing
degradation may often be accompanied by down-
stream-progressing degradation. In Jones Creek this
may have occurred as a result of the inferred lo-
calised flood intensities between 1975 and 1978.
Changes in the channel slope to valley slope
relationship have been documented as an important
®control on channel form and behaviour Erskine et
al., 1990; Schumm, 1993; Schumm et al., 1972,
.1984, 1996 . Steep valleys with meandering channels
carrying a mixed load that undergo bed degradation
experience rapid changes to channel morphology as
bedslope adjusts to valley slope. It is suggested that
the valley slope to channel slope relationship in
® .Jones Creek was near a threshold Schumm, 1979 ,
®whereby the increase in channel slope approxi-
.mately 30% associated with bedlevel lowering sig-
nificantly altered the channel pattern. In Jones Creek,
®changes in mean depth were considerable exceeding
the average depth of the root zone of riparian vegeta-
.tion , resulting in widespread bank instability and
channel metamorphosis. From these data, changes to
channel depth and associated increases in bedslope
have been the dominant controls on subsequent ad-
justments to channel morphology in Jones Creek.
6. Channel recovery in a forested catchment
System recovery in Jones Creek is dependent on
within-channel aggradation and subsequent stabilisa-
tion of depositional surfaces. The deposition and
colonisation of channel marginal features or incipient
floodplains are part of this process. Long term chan-
nel recovery is dependent on the reduction of stream
® .power, increase in sinuosity within a wider channel ,
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128126
decrease in channel width, and subsequent reconnec-
tion of the channel to its floodplain, so that flood-
® .plain inundation results Schumm et al., 1984 . Based
on volumetric analysis undertaken in this study 33%
of the sand sized material eroded in the periods of
channel change is presently stored within the channel
in realignment features, cutoffs, point bars and bed
storage. The storage of this material is part of the
® .recovery process, as noted by Williams 1978 ,
® . ® .Nadler and Schumm 1981 and Eschner et al., 1983
who have highlighted the role of vegetation in sedi-
ment storage and the reduction in channel capacity.
Further analysis of this 33% shows that 65% of this
stored material is `locked up' in vegetated realign-
ment features and unavailable to be reworked.
Rates of vegetative colonisation of depositional
surfaces are extremely high in Jones Creek with A.
mearnsii being the dominant early successional
species. Each year this species sets seed prolifically,
covering the bed of the channel. Greater vegetation
cover increases the roughness and trapping capacity
of the channel zone, thereby reducing flow velocity,
stream power and long term sediment flux, and
promoting further channel recovery. The rate at which
colonising species grow and withstand burial is an
important aspect to the stabilisation of realignment
features. Trees aged between 15 and 13 years have
heights of 10­15 m, with diameters up to 25 cm.
While A. mearnsii is an important successional
species, secondary canopy species with similar
growth rates, such as E. elata and E. iminalis,
provide the longer term stabilisation of these inset
floodplains. Other secondary species, such as Po-
maderris apera, L. longifolia and P. ensiformis,
result in these surfaces becoming densely vegetated,
further enhancing their stability.
While not quantified within this study, LWD, log
jams and within-channel vegetation provide impor-
tant roughness elements that reduce flow velocity
®and stream power in Jones Creek cf., Hupp and
.Osterkamp, 1985, 1996; Friedman et al., 1996 . Four
rush and sedge species tend to dominate and grow in
the sand bed, on bars and within LWD. These are
Juncus usitatus, Cyperus eragrostis, Phragmites aus-
tralis and L. longifolia. Other successional species
occur within the channel as a result of bank failure
which introduces floodplain alluvium into the chan-
nel.
Channel recovery in Jones Creek is seemingly
being achieved by restabilisation on a reach by reach
basis, as certain reaches experience bed aggradation
and reduced rates of bank erosion. This has allowed
within-channel and channel marginal species to es-
tablish. While recovering reaches remain susceptible
to sediment transfer, variations in within-reach sedi-
ment storage will not result in significant changes in
channel geometry or channel position. Once parts of
the channel network have been stabilised, and as
sediment flux is reduced, a sinuous inner channel
will develop, allowing the channel margin to become
vegetated. This process is co-dependent on sediment
depletion rates from upstream and the stabilisation
by vegetation of the inset channel zone.
7. Conclusion
This study has provided an insight into channel
response in a forested catchment to flood distur-
bances induced at the trunk stream confluence. In
Jones Creek, channel metamorphosis was a lagged
response to flood activity on the Genoa River. Inci-
sion and subsequent channel expansion were not a
direct response to the 1971 flood. Rather, channel
widening of the Genoa River at Jones Creek conflu-
ence effectively shortened the course of the creek in
this forested tributary catchment and primed the
tributary for change. Subsequent floods, most likely
in 1975 and 1978, induced dramatic channel adjust-
ments along the tributary river. The over-riding con-
trol on the pattern and extent of channel adjustments
was probably induced by changes to the channel
® .bedslope associated with increases in depth , as
exceedance of a stability threshold induced a series
of secondary channel adjustments. The sensitivity in
the relationship between the Genoa River and Jones
Creek highlights that landscapes in southeastern Aus-
tralia can be on the threshold of instability, only
requiring a suitable trigger.
The style of response documented within this
forested setting conforms to the notion of a complex
response scenario for incised channels. However, the
character of lateral adjustment and the formation of
®realignment features forming an incipient flood-
.plain do not necessarily conform to simple
cutoffrstraightening models. Indeed the spatial vari-
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128 127
ability of channel morphology seen within the study
area highlights the non-uniformity in response to
breaches in thresholds. Feedback and recovery mech-
anisms within the incised, `natural' system have
been documented. Sediment storage associated with
within-channel vegetation and large woody debris
has initiated the early stages of channel recovery and
is an aspect of incised channel behaviour that has not
been previously been documented in the Australian
landscape.
Further studies of alluvial rivers in forested sys-
tems in Australia will improve our understanding of
the `natural' range of river behaviour in southeastern
Australia and thus provide an assessment of river
character and the variability of processes that oc-
curred prior to European settlement.
Acknowledgements
We thank the Logan family for their hospitality
and their insight into Jones Creek and Sarah McGirr,
Andrew Brooks, John Jansen, Kirstie Fryirs, Mal-
colm Tomkins for assistance in the field. David
® .Cameron DNRE provided botanical advice on Jones
Creek. Support for this project was provided by the
® .then East Gippsland River Management Board,
River Basin Management Society and I.D.&A. Criti-
cal review by Edward Keller and Ellen Wohl greatly
improved the presentation of this manuscript.
References
Brierley, G.J., Fryirs, K., 1998. A fluvial sediment budget for
Upper Wolumla Creek, South Coast, New South Wales, Aus-
® .tralia. Aust. Geog. 29 1 , 107­124.
Brierley, G.J., Fryirs, K., Cohen, T.J., 1996. Geomorphology and
river ecology in southeastern Australia: an approach to catch-
ment characterisation. Part one, Working Paper 9603, pp.
1­54.
Brizga, S.O., Finlayson, B.L., 1990. Channel avulsion and river
metamorphosis: the case of the Thomson River, Victoria,
Australia. Earth Surf. Process. Landforms 15, 391­404.
Brookes, A., 1990. Restoration and enhancement of engineered
river channels: some European experiences. Reg. Rivers Res.
and Manag. 5, 45­56.
Brooks, A.P., Brierley, G.J., 1997. Geomorphic responses of
lower Bega River to catchment disturbance, 1851­1926. Geo-
morphology 18, 291­304.
Chesterfield, E.A., Taylor, S.J., Molnar, C.D., 1990. Recovery
after wildfire: warm temperate rainforest at Jones Creek. Arthur
Rylah Institute for Environmental Research, Technical Report
Series, Melbourne, Victoria, 101: 1­27.
® .Chorley, R.J., Schumm, S.A., Sugden, D.E. Eds. , Geomorphol-
ogy Methuen & Co., London, pp. 1­607.
Cohen, T.J., 1997. Channel instability in a forested catchment and
the role of coarse woody debris in channel adjustments: Jones
Creek, East Gippsland, Victoria. Unpublished B.Sc. Honours
Thesis, Macquarie University, School of Earth Sciences, pp.
1­156.
Erskine, W.D., 1992. An investigation of sediment sources in the
Genoa River catchment. In: East Gippsland River Manage-
ment Board, Mallacoota, Victoria, pp. 1­88.
Erskine, W.D., 1993. Erosion and deposition produced by a
catastrophic flood on the Genoa River, Victoria. Aust. J. Soil.
® .Water. Cons. 6 4 , 35­43.
Erskine, W.D., I.D.&A., 1997. Cann River: geomorphic assess-
ment and implications for stream management. East Gippsland
River Management Board and Department of Natural Re-
sources and Environment, pp. 1­100.
Erskine, W.D., Rutherford, I.D., Tilleard, J.W., 1990. Fluvial
geomorphology of tributaries to the Gippsland Lakes. Ian
Drummond and Associates, Wangaratta, Victoria, for Depart-
ment of Conservation and Environment, Victoria, pp. 1­186.
Erskine, W.D., McFadden, C., Bishop, P., 1992. Alluvial cutoffs
as indicators of former channel conditions. Earth. Surf. Pro-
cess. Landforms 17, 23­37.
Eschner, T.R., Hadley, R.F., Crowley, K.D., 1983. Hydrologic
and morphological changes in channels of the Platte River
basin in Colorado, Wyoming and Nebraska: a historical per-
spective. US Geol. Surv. Prof. Paper, 1277-A, pp. 1­39.
Friedman, J.M., Osterkamp, W.R., Lewis, W.M., 1996. The role
of vegetation and bed-level fluctuations in the process of
channel narrowing. Geomorphology 14, 341­351.
Fryirs, K., Brierley, G.J., 1999. Slope-channel decoupling in
Wolumla catchment, NSW, Australia: the changing nature of
sediment sources following European settlement. Catena 35,
41­63.
Galay, V.J., 1983. Causes of river bed degradation. Water Resour.
® .Res. 19 5 , 1057­1090.
Hupp, C.R., Osterkamp, W.R., 1985. Bottomland vegetation dis-
tribution along Passage Creek, Virginia, in relation to fluvial
® .landforms. Ecology 66 3 , 670­681.
Hupp, C.R., Osterkamp, W.R., 1996. Riparian vegetation and
fluvial geomorphic processes. Geomorphology 14, 277­295.
Jackson, W.L., Beschta, R.L., 1984. Influences of increased sand
delivery on the morphology of sand and gravel channels.
Water Resour. Bull. 4, 527­533.
Keller, E.A., Swanson, F.J., 1979. Effects of large organic mate-
rial on channel form and fluvial processes. Earth Surf. Process.
4, 361­380.
Kochel, R.C., 1988. Geomorphic impact of large floods: review
and new perspectives on magnitude and frequency. In: Baker,
® .V.R., Kochel, R.C., Patton, P.C. Eds. , Flood Geomorphol-
ogy. Wiley, Toronto, Canada, pp. 169­187.
Loyn, R.H., Cameron, D.G., Traill, B.J., Sloan, S.F., Malone,
B.S., Schulz, M., Earl, G.E., Triggs, B.E., 1992. Flora and
fauna of the Cooaggalah Forest Block, East Gippsland, Victo-
( )T.J. Cohen, G.J. BrierleyrGeomorphology 32 2000 109­128128
ria. Department of Conservation and Environment. Ecological
survey report, 20, pp. 1­129.
Nadler, C.T., Schumm, S.A., 1981. Metamorphosis of South
Platte and Arkansas Rivers, eastern Colorado. Phys. Geog. 2,
95­115.
Nakamura, F., Swanson, F.J., 1993. Effects of coarse woody
debris on morphology and sediment storage of a mountain
stream in western Oregon. Earth Surf. Process. Landforms 18,
43­61.
Nanson, G.C., Croke, J.C., 1992. A genetic classification of
floodplains. Geomorphology 4, 459­486.
Osterkamp, W.R., Costa, J.E., 1987. Changes accompaning an
extraordinary flood on a sand-bed stream. In: Mayer, L., Nash,
® .D. Eds. , Catastrophic flooding. Allen & Unwin, pp. 201­224.
Prosser, I.P., Chappell, J.M.A., Gillespie, R., 1994. Holocence
valley aggradation and gully erosion in headwater catchments,
southeastern highlands of Australia. Earth Surf. Process. Land-
forms 19, 465­480.
Schumm, S.A., 1979. Geomorphic thresholds: the concept and its
applications. Transactions of the Institute of British Geogra-
phers NS4, 485­515.
Schumm, S.A., 1993. River response to baselevel change: implica-
tions for sequence stratigraphy. J. Geol. 101, 279­294.
Schumm, S.A., Lichty, R.W., 1963. Channel widening and flood-
plain construction along Cimarron River in southwestern
Kansas. US Geol. Surv. Prof. Paper, 352-D, pp. 71-88.
Schumm, S.A., Khan, H.R., Winkley, B.R., Robbins, L.G., 1972.
® .Variability of river patterns. Nature Physical Science 237,
75­76.
Schumm, S.A., Harvey, M.D., Watson, C.C., 1984. In: Incised
Channels -- Morphology, Dynamics and Control. Water Re-
sources Publications, Littleton, Colorado, pp. 1­200.
Schumm, S.A., Erskine, W.D., Tilleard, J.W., 1996. Morphology,
hydrology, and evolution of the anastomosing Ovens and King
® .Rivers, Victoria, Australia. GSA Bull. 108 10 , 1212­1224.
Simon, A., 1989. A model for channel response in disturbed
alluvial channels. Earth Surf. Process. Landforms 14, 11­26.
Simon, A., 1992. Energy, time, and channel evolution in catas-
trophically disturbed fluvial systems. Geomorphology 5, 345­
372.
Tooth, S., Nanson, G.C., 1995. The geomorphology of Australia's
fluvial systems: retrospect, perspect and prospect. Progr. Phys.
® .Georg. 19 1 , 35­60.
Williams, G.P., 1978. The case of shrinking channels -- the
North Platte and Platte Rivers in Nebraska. US Geol. Surv.
Circular 781, 1­48.
Wolman, W.G., 1954. A method of sampling coarse river-bed
material. Am. Geophys. Union Trans. 36, 951­956.