Summary

This research sought to investigate and quantify the magnitude of the values and services provided by headwater streams in the Melbourne Water management region. The research aimed to facilitate the effective management and protection of headwater streams through a clear business case and appropriate policy and design guidelines. The work comprised two parts—1) a literature review and 2) a monitoring program featuring 5 headwater sites to the west of Melbourne.

The literature review defined headwater streams as the point in the landscape where catchment runoff first accumulates sufficiently to create overland flow paths. The review found that headwater streams are dominant and critical features of the landscape. They are the primary sources of streamflow, important sources of organic matter and invertebrates to downstream waters, and act as hot spots for retention and transformation of nutrients such as nitrogen and carbon. Their contribution to regional aquatic biodiversity is disproportionately large. For example, several studies have shown that headwater streams provide extensive habitat, with up to ~one-third of aquatic invertebrate species being unique to these running waters. Headwater streams are also the first source of aquatic life in the transition from hillslopes to the river network and thus can be an important source of colonists to lower reaches. The contribution of headwater streams to regional biodiversity and downstream ecological processes is not yet known in the Melbourne region.

The monitoring program collected data on hydrology, water quality, and ecological structure and function. In general, flow behavior at the sites was highly seasonal, with surface flow only occurring during the wetter months. Water draining from the sites was of a very high quality (e.g. filterable reactive phosphorus concentrations less than 0.01 mg/L). Leaf breakdown rates—an indicator of stream decomposition—were found to be lower in the headwater forested sites compared to nearby agricultural streams. The slow rates are indicative of healthy ecosystem function in Melbourne’s largely undistributed, forested headwaters. Bores in the stream substrate sampled for groundwater-dwelling aquatic animals (stygofauna) revealed few individuals.

Our results highlight the overarching and critical role that headwater streams play for maintaining downstream river and bay health. These systems however, are particularly vulnerable to degradation or loss in rapidly urbanizing cities such as Melbourne. Even small changes in land-use are likely to elicit severe consequences to hydrology, water quality and ultimately stream structure and function.

Recommendations

  • Consistent with regional Performance Objective 16 in the Healthy Waterway Strategy, headwater streams should be protected from urban development. The use of the new stream network layer—which now includes headwater stream extents—will increase the likelihood of this Performance Objective being met.

  • Quantify the loss of headwater streams to date and estimate the length of headwater streams that are vulnerable to urban development. In doing so, determine the implications for regional hydrology, water quality, biodiversity, and relevant targets set (for key values) within the Healthy Waterways Strategy.

  • In locations where development occurs, appropriately designed stormwater control measures (SCMs) must be implemented. The SCMs must be designed in ways that mimic natural flow and water quality regimes.

  • Headwater streams are particularly vulnerable to even small changes in land-use. Thus SCMs must drain all impervious surfaces in new urban developments.

  • Develop guidelines for the protection or restoration of headwater streams in urban developments based on project outcomes, along with data and knowledge from other related studies (e.g. Sunbury Monitoring Project).

Report outline

This technical report comprises four sections. In the first section, we provide a summary of three different pieces of literature review work related to the project. The second section describes the headwater sites and monitoring program. The third section displays and synthesizes the data collected to date. And the final section develops management implications based on the research.

Literature review

A summary of each piece of literature review related work is shown below. Refer to the citation for full details.

Monitoring sites and field work

Site description

In 2019, we commenced the monitoring of four headwater streams in the west of Melbourne (Table 1). These streams drain 100% forested catchments in 1) Mount Macedon, 2) Wombat State Forest, 3) Lerderderg State Park, and 4) Mount Charlie Reserve. An additional site in Mickleham was added to the monitoring program in 2022.

Table 1: Catchment statistics for the study sites. Abbreviated site names are used throughout the report.
Site Abbrev.name Reach.code Catchment.area.ha Mapped.lengh.m
Tributary of Barringo Creek Barringo RJ3_4 4 15
Tributary of Jacksons Creek Jacksons JZ5_47 47 564
Tributary of Coopers Creek Coopers COA_17 17 522
Trib of Charlies Creek us Kent Rd Charlie CH4_88 88 5030
Aitken Creek (west branch) Aitken HTC_259 259 2839

Tributary of Barringo Creek (Mount Macedon)

This shallow, groundwater fed stream drains a small forested (wet eucalypt forest) hillslope (Figure 1). The channel itself is narrow, moderately steep, and is likely the farthest upstream point of surface flow. The riparian zone is dominated by native soils and is largely indistinguishable from the general landscape. Surface soils include forest litter, overlying dark organic loam. Deeper soils are clay.

Figure 1: Site map for the tributary of Barringo Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Figure 1: Site map for the tributary of Barringo Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Tributary of Jacksons Creek (Wombat State Forest)

A somewhat trapezoidal channel drains a small forested (dry eucalypt forest) catchment (Figure 2), which is dominated by a number of eucalypt species and a sparse understory of native grasses and shrubs. The head of the channel network is several hundred metres upstream of the monitoring site. There is scarce forest litter, with surface soils being light organic loam. The soil profile was mostly clay and appeared fairly uniform up to one metre in depth.

Figure 2: Site map for the tributary of Jacksons Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Figure 2: Site map for the tributary of Jacksons Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Tributary of Coopers Creek (Lerderderg State Park)

This very shallow channel drains a small forested (eucalypt forest) catchment (Figure 3), which is dominated by native trees. Channel banks are gentle, resulting in high floodplain connectivity. The first point of surface flow is likely several hundred metres upstream of the monitoring point. Catchment soils appear relatively dry and rather loamy.

Figure 3: Site map for the tributary of Coopers Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Figure 3: Site map for the tributary of Coopers Creek. Red dotted lines = catchment boundary. Mapped stream lines shown in blue.

Tributary of Charlies Creek (Mount Charlie Reserve)

A well defined and fairly deep channel drains a medium-sized forested (dry eucalypt forest) catchment (Figure 4). There are likely 5-6 active channel heads upstream of the monitoring site. Surface soils appeared rather grey in colour and were friable. The initial monitoring site in the Reserve (green lines; figure below) never showed evidence of recent surface flow and was thus moving to a larger catchment to the east (red lines; figure below).

Figure 4: Site map for the tributary of Charlies Creek (upstream of Kent road). Red dotted lines = catchment boundary. Mapped stream lines shown in blue. The abandoned monitoring site in the Mount Charlie Flora Reserve is shown to the west (green lines).

Figure 4: Site map for the tributary of Charlies Creek (upstream of Kent road). Red dotted lines = catchment boundary. Mapped stream lines shown in blue. The abandoned monitoring site in the Mount Charlie Flora Reserve is shown to the west (green lines).

Tributary of Aitken Creek

A somewhat poorly defined channel drains a small agricultural catchment (Figure 5), which is slated for urban development. Dense grass and rock surround the monitoring site. Drainage flow paths have been modified in the top half of the catchment. Catchment soils are fractured clays.

Figure 5: Site map for Aitken Creek (west branch). Red dotted lines = catchment boundary. Mapped stream lines shown in blue. The background is a grayscale representation of the terrain surface (hillshade).

Figure 5: Site map for Aitken Creek (west branch). Red dotted lines = catchment boundary. Mapped stream lines shown in blue. The background is a grayscale representation of the terrain surface (hillshade).

Field work

Hydrology

At each monitoring site, both surface and sub-surface water level has been observed every 6-mins. We measure surface water level by placing probes into stilling wells—40 mm polyvinyl chloride (PVC) slotted pipes. These pipes are placed 50 mm below the invert of the stream bed to allow the measurement of very shallow surface water and are fixed to metal star pickets. We measure sub-surface water level by placing probes into shallow bores (750 mm depth) installed in the invert of the stream bed. The bores comprise a 40 mm PVC pipe, slotted from 750-500 mm below the surface and surrounded by 7 mm screenings up to 400 mm below the surface. The upper void space surrounding the PVC pipes is replaced with compacted, excavated native soils. The 40 mm PVC pipe protrudes (unslotted) 1,200 mm above the invert of the stream bed. The sub-surface monitoring bore is fixed to the same metal star picket mentioned above.

Capacitance water level probes (Odyssey brand) were initially used in the monitoring program. They were subsequently replaced with custom-made submersible pressure sensors—technology which is superior for measuring groundwater level. Water level has been downloaded every 1-2 months and processed to remove data spikes and check against manual measurements taken on site. On site measurements not only verified the data being recorded, but also allowed water level to be accurately compared to the stream bed even if the stream bed changed. We found that while some of the early data recorded using the Odyssey probes was trustworthy, other periods of data were not reliable. Use of our custom-made loggers has greatly increased the reliability and accuracy of the observations.

Time-series of the water level data collected to date are presented below.

Water quality

Water quality has been sampled from the sites during both dry- and wet weather. When possible, water is sampled from four locations—1) the surface, 2) the shallow bore, 3) hyporheic soils 250 mm below the stream bed, and 4) hyporheic soils 625 mm below the stream bed. Most of the samples collected to date have been from below the surface. We use a YSI 6820 V2 Multi Parameter Water Quality Sonde to measure the following properties of water: dissolved oxygen, electrical conductivity, pH, redox potential, and temperature. Water samples are also sent to accredited laboratories for the measurement of: filterable reactive phosphorus, total phosphorus, ammonia, nitrate/nitrite, total nitrogen, total suspended solids, and the stable isotopes of water. We present trends of the various water quality parameters over time, along with differences between the locations sampled within each site. We also compare the quality of surface waters against appropriate management objectives.

Stream ecology

A fundamental ecological process, rates of organic matter decomposition, has been quantified over time in both surface (riparian and benthic zone) and sub-surface (bore and hyporheic zone) environments. We have used the cotton-strip approach to measure the overall decomposition potential of the sites (Tiegs et al. 2013). For this method, cotton strips and temperature loggers are installed at each of the sites in 4 locations—1) riparian zone, 2) stream bed surface, 3) shallow bore (250 mm from the surface), and 4) buried 250 mm below the stream bed surface. The strips are collected from the field after 5 weeks and their strength tested and compared to controls at the end of this period. The relative loss in strength correlates with microbial degradation processes at each location. At the time of collection, a 5 mm sample of cotton strip from each location is also collected for DNA analysis. There are sent to AGRF for DNA analysis to determine microbial diversity present at each location.

We have also sampled the bores for invertebrates inhabiting the sub-surface (stygofauna). A 6L sample is collected from the shallow bore at each site, filtered through a 63 micron sieve and preserved in ethanol. These are sorted and identified to the lowest possible taxa (mostly order or family) in our laboratory. Samples collected in 2019 were sent onto Stygoecologia for identification to lower taxonomic levels (mostly genus or species).

Results and Discussion

Hydrology

The presence of surface water at the forested sites has been highly seasonal (Figure 6). In general, the streams begin to flow in winter and quickly recede sometime during summer. Such flow behavior is very typical of headwater streams located in regions with a temperate climate (REF). When the streams are flowing, the rainfall-runoff response is rather dampened for Barringo. In contrast, the surface water level in Jacksons and Coopers can rise fairly sharply. Surface water has been rare in Charlies and has only been observed following very wet conditions (e.g. the winter of 2022). Aitken has been flowing over most of the monitoring period, but did cease to flow in the summer of 2022/2023 (Figure 7).

Ground water behavior appears fairly complex. The sub-surface is generally dry when the streams are not flowing. Sharp rises of groundwater do occur during dry periods in response to rainfall. In the wetter months, groundwater levels can be higher than surface levels—e.g. clearly illustrated at Barringo (Figure 8). This likely occurs when the piezometer (the bore) intersects a line of equal hydraulic head, and signals a saturated aquifer with lateral flow paths. The saturated aquifer is recharged from the vadose (unsaturated) zone above. Contrasting behavior has been observed at Jacksons (Figure 8) and Aitkens (Figure 9), with surface and sub-surface water levels converging. This likely occurs when the entire catchment is saturated, and any rain causes runoff to the stream.

Figure 6: Surface water level measured in the primary headwater stream sites (black lines). The red dotted vertical lines denote the start of winter in each year. A value of y = 0 m is the base of the streambed. Note the y-axis limits are different for each site.

Figure 6: Surface water level measured in the primary headwater stream sites (black lines). The red dotted vertical lines denote the start of winter in each year. A value of y = 0 m is the base of the streambed. Note the y-axis limits are different for each site.

Figure 7: Surface water level measured in the western tributary of Aitken Creek (black line). The red dotted vertical line denotes the start of winter in 2022.A value of y = 0 m is the base of the streambed.

Figure 7: Surface water level measured in the western tributary of Aitken Creek (black line). The red dotted vertical line denotes the start of winter in 2022.A value of y = 0 m is the base of the streambed.

Figure 8: Groundwater level measured in the subsurface of the primary headwater stream sites (black lines). The red dotted vertical lines denote the start of winter in each year. Level data is presented relative to the base of the streambed (y = 0 m). Bore depth is approximately -0.8 m from the surface.

Figure 8: Groundwater level measured in the subsurface of the primary headwater stream sites (black lines). The red dotted vertical lines denote the start of winter in each year. Level data is presented relative to the base of the streambed (y = 0 m). Bore depth is approximately -0.8 m from the surface.

Figure 9: Groundwater level measured in the subsurface of the western tributary of Aitken Creek (black line). The red dotted vertical line denotes the start of winter in 2022. Level data is presented relative to the base of the streambed (y = 0 m). Bore depth is approximately -0.8 m from the surface.

Figure 9: Groundwater level measured in the subsurface of the western tributary of Aitken Creek (black line). The red dotted vertical line denotes the start of winter in 2022. Level data is presented relative to the base of the streambed (y = 0 m). Bore depth is approximately -0.8 m from the surface.

Water quality

We have observed consistently high quality surface water at the sites (Figures 10 and 11). Nutrient concentrations of surface waters have been low and meet the management objectives for streams in the region (Table 2). Suspended solid concentrations are also low and indicative of clean water (i.e. values less than 20 mg/L). Salinity and pH of surface waters have decreased over time and have contracted to similar levels at the sites (Figure 11). Dissolved oxygen and salinity levels are well below target management objectives (Figure 12). In contrast, pH levels have been quite acidic, but this is typical for headwater streams in the Melbourne region (REF).

Dissolved phosphorus and nitrogen levels have been much higher in the sub-surface compared to surface waters (Figure 13), perhaps due to build-up of soil organic matter.

Over time, oxygen and hydrogen isotopic compositions have converged to similar levels for both surface and sub-surface waters (see Figure 14 for oxygen, similar patterns for hydrogen). This could signal well-mixed (homogeneous) catchments, caused by consecutive La Nina wet weather systems.

Figure 10: Trend of nutrient concentrations from surface waters at the headwater sites. No samples for Charlies or Aitkens have been collected yet.

Figure 10: Trend of nutrient concentrations from surface waters at the headwater sites. No samples for Charlies or Aitkens have been collected yet.

Table 2: 75th percentile nutrient and suspended solid concentrations for surface waters at the headwater sites (calculated from n samples). Environmental Reference Standards for the Uplands of Maribyrnong basin are shown in the last column.
siteName wqVariable n.samples p75_obs_mg_L ers_target_mg_L
Trib of Barringo TN mg/L 11 0.7100 1.050
Trib of Barringo TP mg/L 11 0.0250 0.055
Trib of Barringo TSS mg/L 10 17.2500 NA
Trib of Coopers TN mg/L 10 0.3950 1.050
Trib of Coopers TP mg/L 10 0.0275 0.055
Trib of Coopers TSS mg/L 9 8.6000 NA
Trib of Jacksons TN mg/L 6 1.0750 1.050
Trib of Jacksons TP mg/L 6 0.0475 0.055
Trib of Jacksons TSS mg/L 6 6.5250 NA
Figure 11: Trend of physical parameters from surface waters at the headwater sites. No samples for Charlies or Aitkens have been collected yet.

Figure 11: Trend of physical parameters from surface waters at the headwater sites. No samples for Charlies or Aitkens have been collected yet.

Figure 12: Boxplots for dissolved oxygen (A), salinity (B) and pH (C) for surface waters at the headwater sites. The red lines are the Environmental Reference Standard objectives for the region.

Figure 12: Boxplots for dissolved oxygen (A), salinity (B) and pH (C) for surface waters at the headwater sites. The red lines are the Environmental Reference Standard objectives for the region.

Figure 13: Boxplots for dissolved nutrients, sampled from 4 locations at the headwater sites. Y-axis has been log10 transformed.

Figure 13: Boxplots for dissolved nutrients, sampled from 4 locations at the headwater sites. Y-axis has been log10 transformed.

Figure 14: Trend of oxgyen isotope enrichment, sampled from 4 locations at the headwater sites.

Figure 14: Trend of oxgyen isotope enrichment, sampled from 4 locations at the headwater sites.

Stream ecology

Cotton strip breakdown

Rates of cotton decomposition were most variable in groundwater (hyporheic bore), with a high degree of within-site variation. Rates of cotton decomposition were, on average, slowest in the riparian zone. The fasted rates of cotton decomposition were observed within stream hyporheic sediment or on the benthic surface. Further investigation is required to unravel the key factors driving the observed variation; this includes the role of water temperature, nutrients, and water availability, which are all known to play important roles in regulating rates or organic matter decomposition.

Figure 15: % tensile loss/degree day from four locations within the headwater sites.

Figure 15: % tensile loss/degree day from four locations within the headwater sites.

Microbial data

Stygofauna

While work on the stygofauna data is currently in progress, a sample of the data collected is shown in Table 3. Most of the individuals found have been mites, which is not surprising (REF).

Table 3: Stygofauna found in the shallow bores at two headwater sites (identified by XXX) in 2019.
Site No.of.animals Phylum Subphylum Class Order Family Species
Barringo 8 Annelida Clitellata Oligochaeta Tubificida Enchytraeidae Cognettia antipodum c.f.
Barringo 1 Arthropoda Chelicerata Arachnida Mesostigmata Undetermined Undetermined
Barringo 1 Arthropoda Chelicerata Arachnida Sarcoptiformes Sarcoptidae Sarcoptes scabieic.f.
Barringo 1 Arthropoda Chelicerata Arachnida Sarcoptiformes Sarcoptidae Sarcoptes scabieic.f.
Barringo 1 Arthropoda Hexapoda Insecta Coleoptera Scarabidae Terrestrial/undetermined
Barringo 1 Arthropoda Hexapoda Insecta Coleoptera Staphylinidae Terrestrial/undetermined
Barringo 1 Arthropoda Hexapoda Insecta Diptera Tabanidae Terrestrial/undetermined
Barringo 1 Arthropoda Hexapoda Insecta Hemiptera Pentatomidae Terrestrial/undetermined
Barringo 1 Arthropoda Clitellata Maxillopoda Entomobryomorpha Isotomidae Australotomurus sp.
Barringo 1 Arthropoda Crustacea Maxillopoda Harpacticoida Canthocampidae Canthocamptus sp.
Coopers 1 Arthropoda Clitellata Maxillopoda Entomobryomorpha Isotomidae Australotomurus sp.

Management implications

dafasdfasdf Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.

References

Brown, S. Decomposition in Headwater Streams. Master’s thesis, The University of Melbourne.
Hatt, B.E. Monitoring Methods for Headwater Sites.
Imberger, M., Hatt, B.E., Brown, S., Burns, M.J., Burrows, R. & Walsh, C. Headwater streams in an urbanizing world. submitted to Freshwater Science.
Tiegs, S.D., Clapcott, J.E., Griffiths, N.A. & Boulton, A.J. (2013) A standardized cotton-strip assay for measuring organic-matter decomposition in streams. Ecological Indicators, 32, 131–139.