Fig 1.
Bryan Creek catchment in western Victoria, Australia.
The Wannon River junction and major tributaries are shown in blue and the study area is outlined in black.
Fig 2.
The sand pulse in Bryan Creek (A) looking downstream into the reach adjacent to the Coleraine Township in 1950, (B) looking upstream toward the reach immediately above Coleraine in 1944, substantial in-stream extraction was undertaken in this reach between 1954 and 1990. Photo source: Coleraine Historical Society.
Fig 3.
Erosion control structures, constructed in the late 1960s and early 1970s, in small headwater streams in the Bryan Creek catchment.
(A) ‘Drop structures’–small grade control structures which span the channel and trap sediment, (B) concrete chutes which slow erosion by preventing nickpoints from migrating upstream. Photos courtesy of Glenelg Hopkins Catchment Management Authority.
Fig 4.
(A) Looking upstream into a reach not subject to any intervention, in 2017, the white arrow points to the sharp contact between the pre-pulse riverbank comprised of dark, organic rich sediments, and the upper levee of coarse sand deposited on the floodplain when the bedload pulse formed, (B) looking across towards the left bank in the lower part of Bryan Creek. Dashed white line marks the contact between the former stream bank and the overlying post-European settlement alluvium (PESA), white arrows mark exhumed tree trunks in the channel bed.
Fig 5.
Study reach of Bryan Creek in western Victoria with treatment reaches numbered.
Coloured bars indicate the type of intervention undertaken in the corresponding reach and the approximate date of intervention. The location of extraction sites (1954 to 1994) and road crossings are also shown. Flow is from right to left.
Table 1.
The combination of intervention used, the corresponding number of reaches treated, and the cumulative length of river treated by each combination of intervention.
Fig 6.
Looking downstream into the clay bed reach at the upstream end of the study reach (reaches 1 and 2 in Fig 5) immediately prior to intervention in the late 1980s, twenty years after erosion control measures and improved pasture practices in the catchment reduced the supply of sediment to Bryan Creek.
Note incision into the clay bed. Photo source: Glenelg Hopkins Catchment Management Authority.
Fig 7.
Changes in the catchment of Bryan Creek, the response of Bryan Creek to catchment changes and the timing of intervention between 1950 and 1984.
Table 2.
Variability metrics used in this study.
Fig 8.
Changes in the coverage and compassion of vegetation in the 11 reaches of Bryan Creek between 1991 and 2017.
Fig 9.
De-trended, baseline and post-intervention longitudinal thalweg profiles for Bryan Creek.
Black line is the 1984 profile and orange line is the 2019 profile. The eleven intervention reaches, the combination of interventions undertaken in each reach and channel substrate at the time of intervention are annotated. The location of historical extraction sites is marked with black stars (each star is a single extraction pit). The position of the sand-clay transition in 1984 and 2017 is also marked.
Fig 10.
Post-intervention thalweg complexity values measured using the std dev depths, vector dispersion and sum height diff squared metrics for each of the 11 reaches of Bryan Creek.
Fig 11.
Cross-section complexity values for the std deviation of depths, sum of depth differences squared and vector dispersion metrics for the 11 reaches of Bryan Creek.
Colours and labels at the base of the plot denote the intervention applied to each reach. Note that max/min values in box plot are defined as the 25th/75th percentile values ± 1.5 x the inter-quartile range.
Table 3.
Width-depth ratio, thalweg variability and cross-section variability grouped by intervention type, for each of the eleven reaches in Bryan Creek.
Fig 12.
(A) View looking upstream into reach 10 beside the Coleraine township in 1940, showing wide, flat bed of sand devoid of vegetation, (B) looking downstream from the upstream end of reach 8 in 2017 showing emergent macrophytes on the margin of a continuous channel, (C) pools in roughly the same location as (A) but looking downstream in reach 10 in 2017, (D) a 2017 aerial image of reach 11 showing the sequence of pools. Photo A is from the Coleraine Historical Society.
Fig 13.
(A) Looking north across reach 7 above Coleraine in 1944, significant in-stream extraction began here in 1954, (B) aerial view of the same section in 2017 showing dissected morphology, white arrow indicates the position photo (A) was taken from, (C) representative section of reach 11 in 1994, immediately after extraction ceased, note the absence of Juncus on the channel bed, and (D) representative section of reach 11 in 2017, showing Juncus colonising the bed, the dissected morphology, and bar-and-bench erosion. Photo source: A, Coleraine Historical Society, B, Glenelg Hopkins Catchment Management Authority.
Fig 14.
(A) Reach below original Douglas-Gaze-McMahon Road bridge, looking upstream, in 1940, (B) reach 1 looking upstream towards the current Douglas Road bridge, which is 300 m upstream from the original bridge, in the late 1980s, incising clay bed and bank-collapse clearly visible, (C) reach 1 in 2017, looking across from the right bank showing revegetated banks, small stands of Phragmites and exposed clay bed, (D) oblique aerial view of reach 1 looking upstream, showing formation of a series of shallow runs, emergent macrophytes and bank vegetation. Photo source: A, Coleraine Historical Society, B, Glenelg Hopkins Catchment Management Authority.
Fig 15.
The transformation of reach 8, above Coleraine, from (A) a continuous low-flow channel in 1991 to (B) a sequence of pools by 2017. Phragmites spread inwards from the channel margins, pinching-off the low-flow channel.
Fig 16.
(A) The upstream boundary of reach 7 and 8 and the abrupt changes from Juncus (upstream, top of photo) to Phragmites (downstream, bottom of photo), (B) the downstream boundary of reaches 8 and 9 and the abrupt change from Phragmites (upstream, top of photo) back to Juncus (downstream, bottom of photo). White lines and arrows indicate the location of stock exclusion fences.
Fig 17.
A revised conceptual model of the recovery of sand bed streams at the tail of a bedload pulse which have (or have not) been treated with stock exclusion fencing and bank revegetation at the local scale, following the earlier control on sediment supply from the catchment.