Environmental Outcomes Report V5€¦ · Web viewThis is to provide inundation regularly enough to allow large clumps of lignum (approximately 60% cover) to thrive in the key bird

How does water recovery affect flows and environmental outcomes in the northern basin?

DRAFT 5 AUGUST 2016, WITHOUT PREJUDICE

Please note that the outcomes report is currently being drafted. The following information is a draft excerpt only. There will be additional explanatory sections before and after what is provided here.

The following results and content are draft and are provided for public information and to inform discussion.

5 August 2016

Page 2

Executive summary (in-part)The Basin Plan provides a framework for the management of the water resources of the Murray–Darling Basin. The objectives of the Basin Plan include to protect and restore water-dependent ecosystems and functions, with the aim of achieving a healthy working Murray–Darling Basin.

Prior to the making of the Basin Plan in 2012, the environmental water requirements of 24 large environmental assets (known as umbrella environmental assets) across the Murray–Darling Basin were assessed. These assessments, along with information from other disciplines, were used as part of the implementation of the peer reviewed Environmentally Sustainable Level of Take method to inform the setting of long-term average sustainable diversion limits in the Basin Plan.

At the time of the making of the Basin Plan, it was decided that there would be a review into aspects of the Basin Plan in the northern basin. The Northern Basin Review includes research and investigations in social and economic analysis, hydrological modelling, and environmental science, supported by stakeholder engagement. The review is re-applying the established Environmentally Sustainable Level of Take method. This review has gathered new data and knowledge from a range of disciplines including environmental science. The review may lead to the re-setting of the sustainable diversion limits for the northern basin.

This report describes how different water recovery scenarios, which may lead to the re-setting of the sustainable diversion limits for the northern basin, affect flows and environmental outcomes in the northern basin.

Putting flow indicators into contextThe environmental science steps of the Environmentally Sustainable Level of Take method require selection of umbrella environmental assets (UEA) within the catchment, identification of the hydrological characteristics and ecological values and targets for those assets, and selection of flow indicators that represent important flow-ecology relationships. Each flow indicator is made up of a number of hydrologic metrics (magnitude, duration, timing, frequency) that have ecological relevance within the UEA and, by inference, the broader catchment. Flow indicators and UEA/catchment targets are described in full in the relevant Environmental Water Requiems reports, which can be found on the MDBA website.

To put each of the flow indicators into context, we firstly work out how often the different types of flow would have happened under ‘baseline’ and ‘without development’ model settings.

‘Baseline’ applies 2009 levels of development (water harvesting) over the historical flow record to understand the implications of this level of development over the longer term. This typically shows that small and medium flows happen less often and dry spells are much more frequent and longer with development.

‘Without development’ (WOD) is an approximation of what the flow record would have been if there was no water harvesting across the entire flow record. This typically shows small and medium flows happening much more often and that dry spells were shorter.

Page 3

Figure 1: Historical flows at Bourke on the Darling River. Observed flows and estimated without development flows highlight changes to the flow regime arising from water resource development.

We then develop a target range for how often the different flows need to occur in order to support the lifecycle needs of plants and animals. Target ranges are generally somewhere between baseline and without development frequencies. Target ranges are typically based on specific pieces of evidence, such as the lifecycle needs of fish, waterbirds and floodplain plants, or the likely persistence times of refuge waterholes. These targets are based on the best available knowledge at the time.

Once this framework has been established, we then apply different water recovery settings in hydrological models. We can adjust the levels of water extraction in the model as well as assumptions about how the system is managed, and where and what type of water entitlement is recovered. The model lets us test out different scenarios over an extended time period to see if there are changes to how often different types of flow event happen at particular stream gauge locations. Changes in the frequency of different types of flow event are then used to help understand possible environmental consequences of the different scenarios.

Environmental consequences include things such as:

• more base-flows or small freshes reducing the length of no-flow periods and increasing the likelihood that a network of refuge waterholes will persist through dry times in order to support plants and animals

• more large freshes providing opportunities for fish to move up and down the river channel over large distances, and for some species a strong stimulus to breed

• more bank-full or overbank flows providing opportunities for successful waterbird breeding, boost to plant condition and growth, and for fish to access the floodplain as important nursery and feeding habitat.

We also use the model output to see if there are any phasing differences between recovery scenarios (time between particular types of events exceeding known ecological thresholds). Time between events can be particularly important for ecological risk. E.g. 4 and 10 years between large freshes (river connectivity opportunities), representing a moderate and high

Page 4

risk to the spawning and recruitment of golden perch, a flow dependent native fish species (DSITI 2015).

We have analysed the lengths of time between indicator flow events within the 114 year record of the hydrological modelling to see if there are ecologically meaningful differences in the length of the gaps. The thresholds analysis considers the maximum period between the flow indicator events occurring for each recovery scenario, and the number of spells that are longer than a known ecological threshold.

The ecological thresholds used are indicative of levels of ecological risk, not prescriptive of ecological outcomes. Thresholds are being used as a tool to differentiate between water recovery scenarios in a modelled environment. Given the complex nature of ecological relationships, actual ecosystem responses could be quite different. For example although golden perch populations are known to benefit from successful breeding on large freshes and overbank events, they are also a highly adaptable fish species with a flexible life-history. Breeding is still possible on smaller freshes, albeit likely at much lower levels.

We also look to see if there are any phasing differences between recovery scenarios (time between particular types of events exceeding known ecological thresholds). These can be particularly important in how they link to specific life cycle needs.

We have analysed the lengths of time between indicator flow events within the 114 year record of the hydrological modelling to see if there are ecologically meaningful differences in the length of the gaps. The thresholds analysis considers the maximum period between the flow indicator events occurring for each recovery scenario, and the number of spells that are longer than a known ecological threshold.

For example, ecological thresholds identified from the literature include:

• 4 and 10 years between large freshes (river connectivity opportunities), representing a moderate and high risk to the spawning and recruitment of golden perch, a flow dependent native fish species (DSITI 2013).

• 9 years between inundation events for low-level wetlands, representing the viability of wetland plant seeds based on Ribbon weed (Vallisneria australis), a common aquatic plant, having seeds that will remain viable in dry wetland sediments for up to nine years (Roberts and Marston 2011).

• 10 and 13 years between inundation events for river red gum (Eucalyptus camaldulensis) forests and woodlands, representing an indicative period to cause decline from good to critical condition for modelling state and transitions in floodplain vegetation (Casanova 2015).

• 14 years between inundation events for black box (Eucalyptus largiflorens), representing an indicative period to cause decline from good to critical condition for the purposes of modelling state and transitions in floodplain vegetation (Casanova 2015).

• 11 years between inundation events for lignum (Duma florulenta), representing an indicative period to cause decline from good to critical condition for the purpose of modelling state and transitions in floodplain vegetation (Casanova 2015).

Page 5

• 8 years between waterbird breeding events in the Narran Lakes Nature Reserve, representing the longest observed period between straw-necked ibis (Threskiornis spinicollis) breeding in the Narran Lakes between 1999 and 2008 (MDBA 2012).

The ecological thresholds used are indicative of levels of ecological risk, not prescriptive of ecological outcomes. Thresholds are being used as a tool to differentiate between water recovery scenarios in a modelled environment. Given the complex nature of ecological relationships, actual ecosystem responses could be quite different. For example although golden perch populations are known to benefit from successful breeding on large freshes and overbank events, they are also a highly adaptable fish species with a flexible life-history. Breeding is still possible on smaller freshes, albeit likely at much lower levels.

Even with ecological flexibility and opportunism, time between events is a useful tool to look at phasing differences between scenarios and potential differences in ecological risk.

What water recovery scenarios are being looked at?The Northern Basin Review is working out the possible social, economic and environmental outcomes for a range of water recovery limits below and above what is currently legislated in the Basin Plan.

Seven possible water recovery options are being looked at. The options differ in terms of how much water is to be recovered, as well as some key differences in modelling assumptions, such as how and where the water is to be recovered (Figure 1).

• 278 GL - water recovery at December 2015 represents the amount of water recovered in the Northern Basin as at December 2015, based on current planning assumptions and an estimate of 10 GL for expected future infrastructure-related recovery.

• 320 GL A - water recovery less than Basin Plan represents water recovery at December 2015 (278 GL) supplemented with a further 42 GL recovered from the Condamine–Balonne, Border Rivers and Namoi catchments. Other catchments remain unchanged as they have already met their default contribution towards the shared recovery volume or have less influence on flows in the Barwon–Darling.

• 320 GL B – default water recovery less than Basin Plan (pro rata) represents existing water recovery plus an additional 42 GL of recovery, but with a water portfolio re-balanced to ensure the NSW and Qld contributions follow a further 42 GL of water recovery to be shared across all Northern catchments using the default approach in the Basin Plan. Shared recovery apportionment is default and does not take into account where water has already been recovered.

• 350 GL - targeted recovery represents a targeted water recovery — within Qld, recovery of the shared component has been targeted in the Border Rivers (rather than the Condamine–Balonne) based on relative connectivity with the Barwon–Darling — furthermore, recovery within the Condamine–Balonne has been targeted for volume, location and entitlement types.

Page 6

• 390 GL - Basin Plan current river operations represents the Basin Plan settings without assuming being able to coordinate river flows to Bourke through active management.

• 390 GL - fully implemented Basin Plan represents the fully implemented Basin Plan as currently legislated. This scenario is the benchmark to compare with other scenarios tested as part of the Northern Basin Review.

• 415 GL - water recovery more than Basin Plan represents a similar water recovery approach to the fully implemented Basin Plan, but with an increase to the volume of water recovered.

Figure 2: Core modelling assumptions (explanatory text below)

Targeted recovery strategy – represents a targeted water recovery approach where the volume, location and entitlement types have been specified to optimise environmental outcomes for site-specific flow indicators in the Condamine-Balonne.

Existing recovery – represents water recovered in the northern basin as at December 2015 (based on current planning assumptions and an estimate of 10 GL for anticipated future infrastructure-related recovery).

Page 7

Default (pro rata) – represents the apportionment of shared recovery between catchments. The pro-rata is based on their baseline levels of diversion.

Default + No Condamine-Balonne shared – represents the default for all catchments except the Condamine-Balonne, where there is no shared recovery and local recovery is capped at 100 GL.

BP Current Ops – represents current river operations, as they currently are managed. Other scenarios include enhanced river operations that would result in more optimal delivery /management of water resources.

These scenarios also target different ratios of flow types and entitlements (low-mid flows and mid-high flows). These differences are of particular importance in the Condamine Balonne catchment. Figure 2 shows how each scenario targets different flow types.

Figure 3: Volumes and ratio differences between scenarios for the Condamine-Balonne

Page 8

Figure 4: Water recovery targets for different scenarios

Figure 5: Additional water to be recovered for different scenarios (based on December 2015 levels of recovery as represented by the 278 GL recovery scenario)

Page 9

How flow indicator results and recovery scenarios are assessedThere are 43 flow indicators in the northern basin, distributed across the catchments. Each of these flow indicators is assessed against the seven recovery scenarios using the hydrological modelling. The models work out how often each of the flow indicators happens over the modelled 114 years (the frequency) for each recovery scenario, and under baseline and without development (WOD) conditions. These are then compared to the target frequency ranges (the high and low uncertainty range) outlined in the Environmental Water Requirements Reports. Frequency results for each individual flow indicator are shown at Appendix 1.

From this individual assessment the following questions can be answered Did the frequency identified in the flow indicator meet the target range under a

specific recovery scenario (binary assessment)? Did the frequency identified in the flow indicator improve toward the target range

under a specific recovery scenario (degree of improvement)?

Binary Assessment The binary assessment simply tells us if the target range was met, or not met. Using the example above, the binary assessment would look like this.

Improvement assessmentThe degree to which a flow indicator’s frequency improves toward the target range is very important. While specific ecological responses to only an improvement are complex, it can be reasonably expected that there will be some level of environmental benefit.

To undertake the improvement assessment each individual flow indicator frequency result is converted to an indicator score based on the degree to which the flow indicator frequency

Page 10

Example flow indicator for the Lower Balonne (frequency target is underlined).

38,000 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of six days, any time of the year. The frequency target is to have, on average, no longer than 10-20 years between events.

result improves, achieves, or surpasses the high uncertainty target frequency (as shown in the table below). This process is done for each indicator, and the scores are summed either to the aggregated flow size group (i.e. river channel connectivity group and floodplain and wetland connectivity group) or for each catchment.

Based on the example given above (38,000 ML/d at Culgoa), the following scores would be applied.

The essence of flow indicators is that they work as a group, and contribute broadly to environmental outcomes. Hence, looking at how indicators improve in particular groupings is also undertaken. Two board groupings used in this report are river channel connectivity (baseflows and freshes) or floodplain and wetland connectivity (overbank floodplain and wetland filling flows) (see figure below). These groupings are restricted to within-catchment (i.e. they are not assessed over multiple catchment together).

Page 11

Figure 6: River channel connectivity and wetland/floodplain connectivity groupings and their sub-groups

Once the river channel and wetland/floodplain groupings have been established for each catchment, the group of individual scores are then converted to a percentage of the highest possible score (number of flow indicators for the aggregated group multiplied by 4), as shown by the example for floodplain and wetland connectivity in the table below.

Aggregated group

278 GL 320 GL A

320 GL B

350 GL 390 GL A

390 GL B

415 GL

Score for floodplain and

wetland connectivity group

in the Lower Balonne

(4 flow indicators - highest possible

score is 4 x 4=16)

1+0+3+0

= 41+0+3+0

= 42+0+2+

2= 62+1+2+

2= 72+1+3+

3= 92+1+3+

3= 92+1+3+

3= 9

percent of highest possible score

25% 25% 38% 44% 56% 56% 56%

Using this aggregation method, the higher the percentage score, the greater the improvement in frequency for that aggregated group, within the specific catchment. Once the aggregated score reaches 75% or above, the lower end of the grouped target range frequency (high uncertainty) has therefore been achieved. This process minimises the potential for large increases in frequency for one indicator to mask poor levels of frequency increases in another indicator.

In this report, aggregated results are shown mostly in graphic form (as below for the Lower Balonne UEA floodplain and wetland connectivity aggregated group). Where the exact

Page 12

percentage of improvement in frequency is not required the associated symbols below may also be used.

Figure 7: Aggregated improvement in wetland/floodplain connectivity for the Lower Balonne

Slight improvement in aggregated frequency from baseline to high uncertainty target, 0 to 24% aggregated score

Minor improvement in aggregated frequency from baseline to high uncertainty target, 25 to 49% aggregated score

Major improvement in aggregated frequency from baseline to high uncertainty target, 50 to 74% aggregated score

Aggregated high uncertainty target frequency met (i.e. more than 75%)

The individual and aggregate results method described here underpins the reporting of flow indicators against the various recovery scenarios, at a range of spatial scales.

Page 13

Results: Macquarie-Castlereagh catchmentThe baseline diversion limit for the Macquarie-Castlereagh (excluding interceptions), or how much water was available for consumptive use as at 2009, is 424.3 GL. Current Basin Plan legislation has a local water recovery target of 65 GL for the Macquarie-Castlereagh, which is a 15% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon-Darling system. This includes a contribution from the Macquarie-Castlereagh.

An estimate of 82.5 GL has been recovered as at December 2015, which was used in the modelling as a standard point-in-time estimate. This is not necessarily where water recovery is at now.

The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Macquarie-Castlereagh, up to 5 GL of water recovery, and one scenario that has return of 6 GL back to the consumptive pool.

Macquarie-Castlereagh flow indicatorsThere are four wetland/floodplain flow indicators identified for the Macquarie-Castlereagh aimed at:

Providing large enough flows for long enough to reach key flood dependent vegetation on the floodplain to maintain its character and condition

Providing a large enough flow to reach key waterbird breeding and foraging sites for long enough to enable waterbirds to fledge their young

Providing flows often enough so that waterbirds have more than one opportunity to breed during their lives (some ducks only live for 3-4 years while some of the bigger birds such as ibis can live up to 8 years).

Table 1: Macquarie-Castlereagh flow indicators

Page 14

What are the Macquarie-Castlereagh environmental results?Water recovery for the Macquarie has largely been completed. Only one recovery scenario looks at recovering further water (5 GL under the 415 GL scenario), and one scenario looks at returning water for consumptive use (6 GL under the 320 GL pro rata scenario).

In all water recovery scenarios 4 out of the 4 flow indicators target ranges are met. All scenarios reach the target frequency for high, mid and low-level floodplain, and wetlands and near-channel vegetation flow indicators (including supporting colonial waterbird breeding).

Figure 8: Macquarie-Castlereagh aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

These results provide confidence that, across the range of scenarios, good environmental outcomes are likely for the Macquarie-Castlereagh.

Results: Namoi River CatchmentThe baseline diversion limit for the Namoi (excluding interceptions), or how much water was available for consumptive use as at 2009, is 343.3 GL. Current Basin Plan legislation has a local water recovery target of 10 GL for the Namoi, which is a 3% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon-Darling system. This includes a contribution from the Namoi.

An estimate of 13 GL has been recovered as at December 2015, which was used in the modelling as a standard point-in-time estimate. This is not necessarily where water recovery is at now.

Page 15

The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Namoi up to 15 GL of water recovery, including shared recovery.

Namoi River flow indicatorsThere are two river channel indicators identified for the Gwydir River:

• One base-flow to connect habitats along the river. Base flows are important for maintaining refuge waterhole habitats during dry times. If too many refuge waterholes dry out, local extinction of fish populations could occur with much slower re-colonization upon the return of wetter conditions.

• One fresh to connect habitats along the river and stimulate fish to breed and move. Freshes help provide more habitat, and different types of habitat for fish and other aquatic animals. Support a greater diversity of species as well as more fish.

There is one wetland/floodplain flow indicator identified for the Gwydir River:

• The indicator aims to provide large enough flows for long enough to reach key flood dependent vegetation and wetlands on the floodplain to maintain their character and condition. Overbank flow are important for nutrient exchange between the river and its floodplain, supporting both aquatic and terrestrial food webs.

Table 2: Namoi River flow indicators

What are the Namoi environmental results?Water recovery in the Namoi is well underway. In most water recovery scenarios 3 out of the 3 flow indicator target ranges are met. The exception is the 278 GL scenario, where only the fresh indicator is met (1,800 ML/d). The 278 GL scenario foes not meet the target range for the base-flow and wetland/floodplain indicators.

Page 16

Figure 9: Namoi River aggregated river channel flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Figure 10: Namoi River aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Continuing with water recovery will have benefits in getting water into wetlands and anabranches, and connecting the river with its floodplain more regularly. It will also assist with flows to link and freshen waterhole habitats during dry times to protect refuge pools and fish that depend on them.

The results to date show that continuing with some water recovery will likely result in good environmental outcomes for the Namoi.

Page 17

Results: Gwydir River catchmentThe baseline diversion limit for the Gwydir (excluding interceptions), or how much water was available for consumptive use as at 2009, is 352.2 GL. Current Basin Plan legislation has a local water recovery target of 42 GL in the Gwydir, which is a 13% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon–Darling system. This includes a contribution from the Gwydir.


The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Gwydir, up to 11 GL of water recovery that includes shared recovery.

Gwydir River flow indicatorsThere are two river channel indicators identified for the Gwydir River:

• One base-flow to connect habitats along the river. Base flows important for maintaining refuge waterhole habitats during dry times. If too many refuge waterholes dry out, local extinction of fish populations could occur with much slower re-colonization upon the return of wetter conditions. The Gingham Waterhole is notable as an important refuge for fish during dry times.

• One fresh to connect habitats along the river and stimulate fish to breed and move. Freshes help provide more habitat, and different types of habitat for fish and other aquatic animals. Support a greater diversity of species as well as more fish.

There are seven wetland/floodplain flow indicators for the Gwydir River catchment aimed at:

• Providing large enough flows for long enough to reach key flood dependent vegetation on the floodplain to maintain its character and condition

• Providing a large enough flow to reach key waterbird breeding and foraging sites for long enough to enable waterbirds to fledge their young

• Providing flows often enough so that waterbirds have more than one opportunity to breed during their lives (some ducks only live for 3-4 years while some of the bigger birds such as ibis can live up to 8 years).

Page 18

Table 3: Gwydir flow indicators

What are the Gwydir environmental results?A large amount of the water needed from the Gwydir has already been recovered. In all water recovery scenarios 5 out of the 9 flow indicators meet the target range. The flow indicators that are met are for the high, mid and low-level floodplain, and wetlands and near-channel vegetation flow indicators (including supporting colonial waterbird breeding).

All scenarios just fall short of the target frequency for the base flow, in-channel fresh (connecting habitats and fish movement and breeding) or low-lying wetland indicators (more water-dependent vegetation types). However the degree of improvement toward the target range is significant, and consequently represents a low level of ecological risk.

Page 19

Figure 11: Gwydir River aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

The range of water recovery scenarios being looked will likely result in good environmental outcomes for the Gwydir.

Results: Border Rivers catchmentThe baseline diversion limit for the Border Rivers (excluding interceptions), or how much water was available for consumptive use as at 2009, is 242.1 GL for the Queensland Border Rivers and 207.6 GL for the New South Wales Border Rivers. Current Basin Plan legislation has a local water recovery target of 8 GL in the Queensland Border Rivers and 7 GL for the New South Wales Border Rivers This is a 3% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon–Darling system. This includes a contribution from the Queensland and New South Wales Border Rivers.

An estimate of 15 GL (QLD) and 3 GL (NSW) has been recovered as at December 2015, which was used in the modelling as a standard point-in-time estimate. This is not necessarily where water recovery is at now.

The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Border Rivers, up to 10 GL in QLD and 15 GL of additional contribution to the local and shared recovery.

Border Rivers flow indicatorsThere are three river channel flow indicators identified for the Border Rivers:

• Three fresh flows to connect habitats along the river and stimulate fish to breed and move. Freshes help provide more habitat, by inundating different parts of the river channel, and different types of habitat, for fish and other aquatic animals. These

Page 20

flows increase river channel connectivity, allowing movement up and down the catchment of fish as well as transport of sediment, nutrients and carbon.

There has been analysis of larger flows to inundate anabranches and wetlands. This suggested that these types of flows are not significantly modified by water resource development and MDBA has not set flow indicators for wetland and floodplain connectivity in the Border Rivers.

Table 4: Border Rivers flow indicators

What are the Border Rivers environmental results?Only two scenarios (320 GL pro rata and 350 GL scenarios) meet the 4,000 ML/d (Oct-Dec) flow indicator target range. This indicator provides a flow, between October and December, during the main fish spawning season. All scenarios other than current recovery are within 1 year of meeting the target. The differences between these scenarios is not considered significant, and all scenarios provide improved flow conditions for fish breeding. The remaining two flow indicators are not met by any of the scenarios.

If there was no more water recovery in the Border Rivers, none of the indicators would be met and the level of environmental improvement would be expected to be minor. This would limit improvements in outcomes for fish, and would reduce downstream environmental benefits.

Page 21

Figure 12: Border Rivers aggregated river channel flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Continuing with water recovery will improve in-channel environmental outcomes by increasing the frequency of flows to provide more opportunities for fish to breed and move. It would also assist with wetting different habitats and increase the productivity of the river.

Continuing with some water recovery beyond the current recovery is likely to provide good environmental outcomes for the Border Rivers and for the downstream Barwon–Darling river system.

Results: Condamine-Balonne CatchmentThe baseline diversion limit for the Condamine–Balonne (excluding interceptions), or how much water was available for consumptive use as at 2009, is 713.3 GL. Current Basin Plan legislation has a target of 100 GL of water recovery in the Condamine–Balonne. This is around a 14% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon–Darling system. This includes a contribution from the Condamine–Balonne.


The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Condamine–Balonne, up to 85 GL of additional contribution to the local and shared recovery.

Lower Balonne flow indicatorsThere are five river channel flow indicators identified for the lower Balonne:

• Two indicators for the replenishment of waterhole refuges along the Culgoa and Narran rivers. Periods of no-flow are normal but water resource development has

Page 22

resulted in these periods being much longer for some sections of the river system. If too many refuge waterholes dry out, local extinction of fish populations could occur or much slower re-colonisation upon the return of wetter conditions.

• Three fresh flows to connect habitats along the river and stimulate fish to breed and move. Freshes help provide more habitat, and different types of habitat for fish and other aquatic animals. These flows support a greater diversity of species as well as more fish.

There are four wetland/floodplain flow indicators identified for the lower Balonne:

• The four indicators aim to provide flows to connect different parts of the floodplain and reach key flood dependent vegetation to maintain their character and condition. Overbank flows are important for nutrient exchange between the river and its floodplain and for supporting both aquatic and terrestrial food webs.

Table 5: Lower Balonne flow indicators

Narran Lakes Flow indicatorsThere are four wetland/floodplain flow indicators identified for the Narran Lakes:

• The four indicators aim to provide a range of different flow volumes to reach key parts of the Narran Lakes ecosystem, including waterbird breeding and foraging sites. Provide flows to support large scale waterbird breeding (some ducks only live for 3-4 years while some of the bigger birds such as ibis can live up to 8 years). The four indicators are of different magnitudes to water different parts of the Narran Lakes system and ensure vegetation and wetlands are maintained in character and condition.

Page 23

Table 6: Narran Lakes flow indicators

What are the environmental results?Water recovery in the Condamine–Balonne is well underway. The water recovery scenarios achieve between 2-4 of the 9 lower Balonne flow indicator targets and 2-3 of the 4 Narran Lakes flow indicator targets. Under all scenarios, the watering needs of the Narran Lakes rookery habitat and the large in-channel fresh in the Culgoa River are achieved.

Continuing water recovery will have benefits for connecting the floodplain more regularly, providing flows to link the network of waterholes and provide opportunities for fish and waterbirds to breed. There are still risks, especially for the more water-dependent riparian and inner floodplain in the lower Balonne, and large-scale waterbird breeding in the Narran Lakes.

There are also differences in the sequencing of the flow indicators (period between events). Increasing recovery can break some of the longer dry spells for small to medium floods and reduce high risk spells between bankfull events for fish movement.

Page 24

Figure 13: Lower Balonne aggregated river channel flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Figure 14: Lower Balonne aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Page 25

Figure 15: Narran Lakes aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

The results suggest that targeted water recovery is an efficient way of improving ecological outcomes. Narran Lakes outcomes are driven by the Narran River recovery: 20 GL appears to be a threshold to achieve balanced outcomes (current recovery is at around 10 GL). Lower Balonne outcomes can be achieved from recovery in the lower Balonne and from upstream of Beardmore Dam: 80 GL appears to be a threshold to achieve balanced outcomes (current recovery is at around 55 GL).

Page 26

Results: Barwon–Darling catchmentThe baseline diversion limit for the Barwon–Darling (excluding interceptions), or how much water was available for consumptive use as at 2009, is 198 GL. Current Basin Plan legislation has a local water recovery target of 6 GL in the Barwon–Darling. This is 3% reduction in the consumptive pool. Current Basin Plan legislation also has a shared water recovery target of 143 GL across catchments of the north to meet the needs of the Barwon–Darling system. This includes a small contribution from the Barwon–Darling itself.

There has been local and shared water recovery in the Barwon–Darling with an estimate of 31 GL recovered as at December 2015 (including 3 GL of expected infrastructure recovery). This recovery number has been used in the modelling as a standard point-in-time estimate. (Noting this is not necessarily where water recovery is at now).

The Northern Basin Review is looking at a range of water recovery scenarios. These range from no further water recovery in the Barwon–Darling, to a potential return of 20 GL to the consumptive pool.

The majority of flow in the Barwon–Darling originates from the upstream tributaries. The shared recovery volume acknowledges the importance of these catchments to improving environmental outcomes in the Barwon–Darling. However, recovery within the Barwon–Darling is also important as it is the most direct and efficient way to improve the flows within the Barwon–Darling. This is because, across the scenarios investigated, around 43% of water recovered in the upstream catchments makes it down to the Barwon–Darling as inflows.

Barwon–Darling flow indicatorsThere are elven river channel flow indicators identified for the Lower–Balonne. The flow indicators target outcomes throughout the broader Barwon–Darling river system. The river channel freshes are specified at specific locations along the river but will flow through the system to provide benefits to the water dependent ecosystem along the way. The differences in the indicators relate to the size, duration and seasonal timing of the flow events with each indicator aiming to achieve a different outcome for the environment. The larger overbank flow events are also specified at a location but mapping of floodplain inundation gives us confidence about the level of floodplain connectivity that these flows provide for the whole river system. The indicators are nested which means that larger flows can achieve more than one indicator.

River channel indicators: • Seven small and large freshes at Bourke, Louth and Wilcannia to connect habitats

along the river, provide water for vegetation on the banks of the river and stimulate fish to breed and move. Freshes help provide more habitat, and different types of habitat for fish and other aquatic animals. These flows support a greater diversity of species as well as more fish.

Page 27

Wetland/Floodplain indicators:• Four overbank flow indicators to provide flows to connect different parts of the

floodplain and reach key flood dependent vegetation to maintain their character and condition. Overbank flows are important for nutrient exchange between the river and its floodplain and for supporting both aquatic and terrestrial food webs.

Table 7: Barwon Darling flow indicators

What are the environmental results?Water recovery in the Barwon–Darling is largely complete and is over and above the recovery target. Most scenarios look at no further water recovery in the Barwon–Darling, while one scenario looks at potentially returning some of the water for consumptive use (20 GL). However water recovery is potentially not yet complete in upstream catchments and this will influence inflows into the Barwon–Darling under the range of possible scenarios being looked at.

In all water recovery scenarios we achieve 4 out of the 11 flow indicators. These are the small fresh at Bourke, the small and large fresh at Wilcannia, and the large, rare outer floodplain flow at Wilcannia.

The level of water resource development in the north has had a profound influence on the flows of the Barwon–Darling river system. This makes it relatively difficult for the levels of water recovery being considered to significantly influence (top-up) larger bankfull and overbank flows and increase how often they occur.

Page 28

Regarding the overbank flow indicators that are not met, there is little to no improvement (from baseline). These indicators are related to the flooding of the riparian, inner and mid floodplain. Over the long term this could result in changes to the extent and character of the Barwon–Darling floodplain vegetation. Vegetation stress with increasing dry-spell length is likely to lead to changes in floodplain vegetation composition and condition, moving towards less water dependent species.

Longer duration freshes are also difficult to achieve. Flow indicators with shorter duration targets are typically met (e.g. Wilcannia freshes for 7 days), whereas those with longer durations show little improvement in frequency (e.g. Louth freshes that last for 20 days).

Figure 16: Barwon–Darling aggregated river channel flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Page 29

Figure 17: Barwon–Darling aggregated wetland/floodplain flow indicator results showing the level of frequency improvement from baseline (0%) toward the target range (75%)

Continuing water recovery will have benefits for increasing how often freshes down the river occur, providing flows to link the network of waterholes and provide opportunities for fish and other animals to feed, breed and move. This is especially important for the Darling River which has been identified as a key breeding site for golden perch across the basin.

With continued water recovery alone, it is likely that environmental risks will remain for the Barwon–Darling floodplain and its vegetation condition.

Additional opportunities Other actions that further improve environmental outcomes will be important in the management of the northern basin aquatic ecosystems (e.g. helping restore smaller flows in the lower Balonne that are hard to achieve through water recovery alone). The Northern Basin Review ecological results to date have been focused on outcomes as a result of water recovery, which focuses on different volumes of water recovered within the catchments. It is considered that improvements to some parts of the flow regime can be improved through complementary management actions, such as varying flow rules, temporary trade and event based management. Any variation in access and/or operating rules would need to be part of a co-operative approach between stakeholders, jurisdictions and the Australian Government. The Commonwealth Environmental Water Holder is already investigating options for buying water on a temporary basis through an expression of interest process.

Page 30

Appendix 1. Individual flow indicator rationales and resultsOutlined below are the 43 flow indicators and how the modelled water recovery scenarios increase (or decrease) how often the flow event occurs compared to the modelled '2009 baseline' setting and the 'without-development' settings.

For the 24 flow indicators specified for the lower Balonne River floodplain, Narran Lakes and the Barwon–Darling, the rationale underpinning the flow indicators (magnitude, duration, timing and frequency), and an explanation of the results, is also included.

The rationales for the 19 flow indicators for the other catchments are not shown below, as they have remained unchanged and full explanations can be viewed on line at the relevant Environmental Water Requirements reports available on the MBDA website,

The individual flow indicator results are reported using the graphic below. The graphic (which is an example only) shows the following key points:

Flow indicator: the description of the flow event being assessed in terms of its magnitude (flow rate or volume), duration and timing.

Frequency metric and scale line: details how frequency is measured and shows the scale. The metric can change for each indicator and may include for example percentage of years with an event or the number of years between events.

Baseline (B): shows the frequency under the baseline 2009 conditions of development, and should be read against the scale line immediately below (37% of years had an event).

Without development (W): shows the frequency under without development conditions, and should be read against the scale line immediately below (91% of years had an event).

Target range: shows the frequency target range that, if met, gives a greater degree of certainty around achieving the desired ecological outcomes for the flow indicator (70-80% of years with an event).

Water recovery scenarios: shows the frequency range achieved by the modelled water recovery scenarios as a group. (e.g. 42-68% of years with an event).

A table is also included that presents an overview of the main assumptions and confidence in the SFIs (Appendix 2).

Page 31

Figure 18: Explanatory graphic showing individual flow indicator results range; the target range; the baseline frequency and the 'without development frequency.

Page 32

3 9

70-8042-68

Lower-Balonne flow indicators resultsThe section of the Balonne downstream of St. George is principally an unregulated system, and in these types of systems, the flow indicators represent a broader range of flows than just the event specified. Therefore, the flow indicators specified help to measure changes across the flow regime. The nine lower Balonne floodplain flow indicators include five river channel connectivity indicators and four floodplain and wetland connectivity indicators as shown in Figure 6. These groups are further divided into sub-groups to describe flow components that the flow indicators represent, with the aim to have a range of flow indicators that collectively represent the ecologically important elements of the flow regime.

Figure 19: Breakdown of lower Balonne flow indicators into flow component groups and sub-groups

River channel connectivity results (lower Balonne)

Refuge Waterhole IndicatorsThe watercourses of the lower Balonne contract back to become a series of waterholes during periods of no-flow (dry spells). Waterholes act as refuges and provide critical habitat to allow many water-dependent plants and animals to survive through dry times. Once the rivers start to flow again, aquatic animals can move and spread out to re-populate habitats across the broader river system.

Although there are upstream refuges in the system that continue to hold water during dry spells, the importance of maintaining a network of waterholes in the lower part of the system should not be underestimated. In addition Jack Taylor Weir and Beardmore dams are recognised as upstream refuges that provide some aquatic habitat value during drought. However, the weir and dam wall structures often act as barriers to fish movement, limiting connectivity between upstream and downstream river reaches. Essentially this means it is

Page 33

Lower Balonne UEA

9 flow indicators

River channel connectivity

5 flow indicators

Waterhole refuges

2 flow indicators

Small in-channel freshes

1 flow indicator

Large in-channel bankfull flows

2 flow indicators

Floodplain & wetland connectivity

4 flow indicators

Overbank flows

4 flow indicators

much harder for aquatic animals to move up and down between habitats and to recolonise areas after dry spells.

Research by DSITI (2015) into the persistence of the waterholes identified no-flow periods longer than a year as a threshold for waterhole stress, and no-flow spells of over 18 months as being a threshold where parts the system is approaching complete failure. A finding of the research was that periods of no-flow is variable across the lower Balonne with the longest no flow periods identified in the mid Culgoa and mid to lower reaches of the Narran River. It was considered that maintaining a network of waterholes with at least 0.5 metres of water is needed to maintain aquatic communities during extended dry periods.

Without a network of refuge waterholes in the lower reaches of the river system of reasonable water quality, the ecological consequences are likely to be severe. Consequences such as reduced population resilience (numbers, distribution, and genetic diversity), or even the prospect of local extinction for many aquatic species. Dry times in the lower Balonne are a normal aspect of the hydrology, especially for the lower reaches of the river system, with even the without development modelling showing times of increased stress.

Water resource development has significantly increased the duration of no-flow events in both the Culgoa and Narran rivers. The number of no-flow periods in the 114 year model scenario that exceed the persistence thresholds of waterhole refuges has significantly increased under baseline conditions as compared to without development (see CB1 and CB2 below). This affects how often waterholes are connected and re-filled, with ecological risk increasing as no-flow periods become longer, to the point that these periods surpass the persistence time of the waterholes.

There are two flow indicators specified for maintaining the network of waterhole refuges in the lower Balonne. One described at Weilmoringle gauge on the Culgoa River, and the other described at Narran Park gauge on the Narran River. The intention of the flow indicators is to use a flow indicator site that is downstream of the refuge waterholes and can therefore observe the length of no-flow periods and assess dry periods against the persistence thresholds of refuge waterholes. The logic is that any flow (taken to be 2 ML/d) downstream of the refuge waterholes will have filled the river channel and therefore topped up the network of waterholes. The timing is not constrained to reflect the intention to reduce longer no-flow periods (longer than a year) whenever they occur.

The frequency that flow is needed is based on the research by DSITI (2015). The upper end of the target range for both flow indicators is 350 days between flow events to represent the network of waterholes under stress. The lower end of the target range provides for at least two refuge waterholes in the lower sections of each river. This is represented by a period between flow events of 430 days for the Culgoa River and 470 days for the Narran River.

The frequency requirement of the flow indicators considers the longest 10% of the no-flow periods across the 114 year modelled time period. Considering a range of the longest no-flow periods rather than the single longest no-flow spell was considered to ensure results are not determined by a single event. The distribution of the no-flow events under without development, baseline conditions are presented in the figures below, demonstrating the increase in severity of no-flow periods under the 2009 baseline conditions.

Page 34

The distribution of longer no-flow spells (longer than 250 days) under the without development and baseline scenarios is shown below. The top graph shows Weilmoringle on the Culgoa River and the bottom is Narran Park on the Narran River. The thresholds identify different levels of stress for refuge waterholes and are were informed by waterhole modelling done for one of the environmental science projects done by the Queensland government to inform the Northern Basin Review.

These graphs show the significant impact of water resource development with the large increase in longer spells from without development to baseline conditions.

Figure 20: Distribution of longer no-flow spells under without development and baseline conditions on the Culgoa River (at Weilmoringle)

Figure 21: Distribution of longer no-flow spells under without development and baseline conditions on the Narran River (at Narran Park)

Page 35

CB1. Waterhole refuge - Any flow (taken as a minimum of 2 ML/d) on the Culgoa River, measured at the Weilmoringle gauge, for a minimum of one day, any time of the year. The frequency target is to have a period between no-flow events of no longer than 350-430 days for the average of the top 10% of no-flow periods (longest 23 to 25 no-flow periods depending on scenario).

• Under baseline conditions (2009 levels of development before water recovery), the average periods of no-flow (for the top 10% of no-flow spells) at Weilmoringle gauge on the Culgoa River is much longer compared to without development conditions (from 451 days compared to 247 days).

• Based on waterhole persistence research (DSITI 2015). The 350 day target is to maintain at least seven waterholes along the Culgoa River, with at least four of these to a depth of more than 0.5 m. The 430 days target provides at least two refuge waterholes to at least 0.5 m.

• The modelled water recovery scenarios result in little improvement in frequency from the 2009 baseline in the frequency of the Culgoa River refuge waterhole flow indicator; and all scenarios are well short of the target range. The 320 GL scenario is marginally better (1-2 days) than the other scenarios; however, this is not considered to represent an improvement to the maintenance of waterholes.

• During prolonged dry periods, there is likely to be a lack of a refuge waterhole network in the lower reaches of the river system. Without management intervention this creates a considerable risk of reduced population resilience (numbers, distribution, and genetic diversity), or even the prospect of increased local extinctions. Much larger flows are anticipated to be needed to enable repopulation.

Page 36

CB2. Waterhole refuge - Any flow (taken as a minimum of 2 ML/d) on the Narran River, measured at the Narran Park gauge, for a minimum of one day, any time of the year. The frequency target is to have a period between no-flow events of no longer than 350-470 days for the average of the top 10% of no-flow periods (longest 24 to 29 no-flow periods depending on scenario).

• Under baseline conditions (2009 levels of development before water recovery), the periods of no-flows (top 10%) for the Narran River waterhole refuge flow indicator are much longer compared to without development conditions (from 542 days compared to 349 days).

• The modelled water recovery scenarios result in little, to nil, improvement in frequency from the 2009 baseline of the Narran River refuge waterhole flow indicator; and all scenarios are well short of the target range.

• The 320 GL B pro rata scenario is marginally better, however this is not considered to represent an improvement to the maintenance of waterholes.

• During prolonged dry periods, there is likely to be a lack of a refuge waterhole network in the lower reaches of the river system. Without management intervention this creates a considerable risk of reduced population resilience (numbers, distribution, and genetic diversity), or even the prospect of increased local extinctions. Much larger flows are anticipated to be needed to enable repopulation.

The modelling shows no significant improvements to the no-flow/low flow part of the flow regime, which is likely because of the impact of water regulating structures such as Beardmore Dam and there being no demand placed in the modelling to actively deliver low flows. However, it is considered that improvements to this part of flow regime can be achieved by varying low flow rules or modifying existing access rules. However, any variation in access and operating rules needing to be part of a co‐operative approach between stakeholders, jurisdictions and the Australian Government.

Small in-channel freshesSmall in-channel freshes occur regularly, close to annually with it common for years to have multiple freshes. These flows are important for aquatic habitat maintenance as they regularly top up waterholes and improve water quality, provide a diversity of foods, and support some diversity in habitat types. Frequent opportunities for fish and other aquatic species to feed,

Page 37

breed and move between habitats helps to support healthy and resilient populations, better able to respond to extreme conditions like floods and droughts.

There is one flow indicator specified to represent small in-channel freshes in the lower Balonne. This flow event is likely to inundate the lower sections of the main river channels over hundreds of kilometres. This flow event was assessed as the minimum flow to be confident about providing system scale connectivity. It is highly likely to pass down the entire length of the Culgoa and Narran rivers, and also down at least part of the Bokhara River. This is based on analysis of observed flows between 1965 and 1975 (before the majority of the water resource development) and 2005 and 2015 (after water resource development), and on modelled flows under the baseline 2009 condition of development scenario.

The timing of the flow indicator has not been constrained as it is considered that a small fresh would provide benefits to the environment at any time of the year. Responses are likely to be greater during warmer months but this generally corresponds with the typical summer high flow season in the lower Balonne. The target frequency range aims to provide a small fresh close to annually to improve habitat conditions for instream vegetation, fish, frogs and other aquatic biota. This is supported by advice from fish ecologists (NSW DPI 2015), and is consistent with the without development hydrology of the lower Balonne where 98% of years had an event of this magnitude.

CB3. Small in-channel fresh - 1,000 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of seven days, any time of the year. The frequency target is to have the flow event occur between 80-90% of years.

• Under baseline conditions (2009 levels of development before water recovery), the small in-channel flow happens much less often than without development conditions (from 98 out of 100 years, to 74 out of 100 years)

• The modelled water recovery scenarios all respond with little improvement in frequency from the 2009 baseline; and all scenarios are short of the target.

• Based on the frequency performance alone (as shown above) there is no difference between the scenarios. This is likely due to hydrological model settings, as there are a limited number of water harvesting licences that pump water at lower flow thresholds and these licences were not targeted.

• Ecologically, this result represents a very small shift in the hydrology across scenarios for small freshes to connect habitats, including waterholes.

Page 38

The sequencing of the small fresh flow indicator has been considered to assess the number of periods between the flow event lasting longer than 2 and 3 years (Table 8). These intervals between the flow events were selected to reflect the intention to have a small fresh occur regularly and close to annually. These intervals are not considered a critical interval based on evidence of ecological thresholds, but do provide another line of evidence to consider differences between the water recovery scenarios.

The spells analysis shows that under most water recovery scenarios, there is no reduction from 2009 baseline conditions in the number of two year spells (11 spells). The 320 GL pro rata, 390 GL and 415 GL scenarios are better at reducing the number of extended three year dry periods from two to one.

The modelled water recovery scenarios result in limited improvement in frequency and sequencing of the flow indicator from the 2009 baseline conditions. All scenarios are well short of the target range, with the spells analysis indicating that the 320 GL pro rata, 415 GL and 390 GL scenarios are marginally better at reducing extended dry periods. Fish and other aquatic species would likely be under similar levels of stress due to prolonged periods between small freshes through the river system, with implications for environmental outcomes such as less resilience in the populations of aquatic plants and animals. The maximum period between events is 3.5 years for the 2009 baseline conditions and all water recovery scenarios.

Similar to the waterhole refuge flow indicator, the low flow (and therefore low volume) nature of this flow indicator means there is scope to consider management actions to deliver on ground outcomes for small freshes, such as temporary trade or protection of flow events.

Table 8: Spell statistics for the 1,000 ML/d for seven day flow indicator at Brenda on the Culgoa River

Water Recovery Scenario Maximum period between events (years)

No. of spells longer than 2 years


Baseline - 2009 pre-water recovery 3.5 11 2

278 GL - water recovery as at December 2015

3.5 11 2

320 GL A - recovery below Basin Plan

3.5 11 2

320 GL B pro rata - recovery below Basin Plan

3.5 11 1

350 GL- targeted recovery 3.5 11 2

390 GL - fully implemented Basin Plan

3.5 11 1

415 GL - recovery above Basin Plan 3.5 11 1

Page 39




Without development 1.7 0 0

Large in-channel freshesLarge in-channel freshes provide opportunities for aquatic animals to move larger distances up and down the river channel. Flow dependent native fish for example make use of these flows to find new habitats and suitable feeding areas, and to breed and mix with other populations (which provides genetic diversity and resilience). In particular, large freshes and bank-full flows provide access to a diverse range of in-stream habitats, including those associated with snags, in-channel benches and nearside vegetation.

Two flow indicators have been identified to represent large in-channel flow events that are almost bank-full in the Culgoa and Narran rivers. These large in-channel freshes are likely to result in greater than 2 metre rises in the river (above commence to flow) and achieve a 0.3 metre per second flow velocity along much of the Culgoa and Narran rivers. These factors are considered important in triggering the movement of fish. This is based on applying the findings of research from other catchments, including in the Moonie River (Marshall et al. 2016) and in the southern basin (Reynolds 1983; Koehn et al. 2009; Mallen-Cooper and Zampatti 2015). The Queensland government also has a fish movement study, using acoustic tagging, in the lower Balonne floodplain. Initial findings have shown one species (Hyrtl's tandan) moved in response to a small rise in the river of 2,500 ML/d at the Brenda gauge (P. Woods, pers. Comm). The study is ongoing and is expected to provide a technical report in 2017. It is anticipated that this type of flow would provide fish with opportunities to overcome in-channel barriers to move between waterholes and through parts of the system.

The duration of the flow indicators is 14 days based primarily on the hatch time for Murray cod eggs which encompasses the needs of other native fish in the region. This duration has been checked against the hydrology of the lower Balonne and is within the median duration for these events under without development conditions. The frequency is for the large freshes to occur roughly every 2 years based on advice from experts about the requirements of fish for dryland rivers (NSW DPI 2015).

CB4. Large in-channel flow - 1,700 ML/day on the Narran River, measured at the Wilby Wilby gauge, for a minimum of fourteen days, between August and May. The frequency target is to have the flow event occur between 40-60% of years.

Page 40

• Under baseline conditions (2009 levels of development before water recovery), the Narran River large in-channel flow indicator occurs much less often than without development conditions (from 25% of years with an event compared to 61% of years with an event).

• The modelled water recovery scenarios do improve the frequency of the Narran large in-channel flow indicator.

• The 278 GL and 320 GL scenarios do not increase the frequency of this type of event as much as the other scenarios. This is likely because the 278 GL and 320 GL scenarios have a smaller volume of water recovery in the Narran system (7 GL and 17 GL as compared to 21 GL to 30 GL for the others).

• The 320 GL pro rata, 350 GL and 415 GL scenarios reach the target range, with both 390 GL scenarios very close (within 1% of the lower end of the target range).

• Ecologically, the 320 GL pro rata, 350 GL and 415 GL scenarios provide more opportunities for aquatic biota to move, feed and breed, which is especially important for flow dependent native fish species such as golden perch, silver perch, spangled perch and Hyrtl’s tandan. Improving this part of the flow regime is expected to improve the population condition and resilience of a range of aquatic species including fish, frogs and turtles.

• Under the recovery scenarios that do not meet the target range, there will be less opportunities for fish to move and breed to improve population condition.

• Increased frequencies of this type of flow event also help to improve the condition of near channel vegetation, which can help increase bank stability and provide additional food resources.

Critical thresholds (spells) have been used to consider if there are ecologically significant sequencing differences between the water recovery scenarios. A moderate risk threshold of longer than four years between the event, and a high risk threshold of longer than 10 years was adopted. These thresholds are based on thresholds of concern used by the Queensland government for water resource planning to represent risks to the spawning and recruitment and therefore the population of golden perch in dryland river systems (DSITI 2015).

The spells analysis (see Table 9) shows that under the 320 GL pro rata, 350 GL, 390 GL and 415 GL scenarios the high risk critical threshold of 10 years is eliminated. This high risk threshold is breached under 2009 baseline conditions, and in the 278 GL and 320 GL water recovery scenarios. The moderate risk threshold of four years is breached under all water

Page 41

recovery scenarios but there is a reduced for the 320 GL pro rata, 350 GL, 390 GL and 415 GL scenarios (from seven to six moderate risk spells). This compares to only two moderate risk and no high risk spells under without development.

The maximum period between events shows a change between scenarios. All water recovery scenarios show a reduction from baseline conditions, the 278 GL and 320 GL scenarios are 10.1 years (i.e. just breaching the high risk critical threshold) and the 350 GL, 390 GL and the 415 GL scenarios all have a maximum period between events of 8 years.

This maximum spell is significant for low-level areas of the floodplain where water can break out as a result of this large fresh. Nine years between inundation events has been identified as a critical threshold for low-level wetlands, as it represents the viability of wetland plant seeds based on research on Ribbon weed (Vallisneria australis). Ribbon weed is a common aquatic plant, having seeds that will remain viable in dry wetland sediments for up to nine years (Roberts and Marston 2011). The maximum spell for the 278 GL and the 320 GL scenarios breach this threshold, representing an increased risk to this parts of the floodplain. All other scenarios eliminate this critical threshold.

Table 9: Spell statistics for the 1,700 ML/d for fourteen days flow indicator at Wilby Wilby on the Narran River.


Number of spells longer than 4 years (moderate risk)

Number of spells longer than 10 years (high risk)

Baseline - 2009 pre-water recovery

12.7 7 1


10.1 7 1


10.1 7 1


8 6 0

350 GL- targeted recovery 8 6 0


8 6 0

415 GL - recovery above Basin Plan

8 6 0


Page 42

CB5. Large in-channel flow - 3,500 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of fourteen days, between August and May. The frequency target is to have the flow event occur between 40-60% of years.

• Under baseline conditions (2009 levels of development before water recovery), the Culgoa River large in-channel flow happens much less often than without development conditions (from 30% of years with an event compared to 68% of years with an event).

• The modelled water recovery scenarios all significantly improve the frequency of the large in-channel indicator, with all achieving the target range.

• The 350 GL scenario results in the smallest increase in frequency (41%); likely because there was less recovery of low-mid channel entitlement types in the Culgoa River (as compared to the other scenarios) with recovery volume focused on the mid-high entitlement types to improve floodplain connectivity. The highest frequency increase is achieved under the 415 GL; both 390 GL scenarios; and 320 GL pro rata scenarios (46%).

• Ecologically, all scenarios are likely to provide more opportunities for aquatic biota to move, feed and breed, which is especially important for flow dependent native fish species such as golden perch, silver perch, spangled perch and Hyrtl’s tandan. Improving this part of the flow regime is expected to improve the population condition and resilience of a range of aquatic species, including fish, frogs and turtles. The condition of near channel vegetation is also likely be improved. This can also increase bank stability and provide additional food resources.

• Improvements are expected to be greatest under the 415 GL, 390 GL and 320 GL pro rata scenarios.

The same critical threshold analysis as done for the large Narran River large fresh flow indicator has been applied to this flow indicator.

The spells analysis in Table 10 shows that under the 320 GL, 320 GL pro rata, 390 GL, and 415 GL scenarios the high risk critical threshold of 10 years is eliminated. This high risk threshold is breached under 2009 baseline conditions twice, and in the 278 GL and 350 GL water recovery scenarios once. The high risk threshold being breached in the 350 GL scenario is a result of the targeted nature of the water recovery for that scenario, with

Page 43

entitlements associated with floodplain connectivity (overland flow licences) being more heavily targeted in the Culgoa River.

The moderate risk threshold of four years is breached under all water recovery scenarios but there is a reduced risk from baseline conditions for all scenarios (from seven spells). All scenarios except the 350 GL scenario reduce this to five spells and the 350 GL scenario reduces this to six spells. This compares to only one moderate risk and no high risk spells under without development.

The maximum period between events also shows a change between scenarios. The water recovery scenarios show varying levels of reduction from baseline conditions (maximum period between events of 10.8 years). The 278 GL and 350 GL scenarios shows the least reduction at 10.1 years (i.e. just breaching the high risk critical threshold). The 320 GL scenario has a maximum period of 8 years, and the 320 GL pro rata, 390 GL and the 415 GL scenarios have a maximum period between events of 7.6 years.

This maximum spell is significant for low-level areas of the floodplain where water can break out as a result of this large fresh. Nine years between inundation events has been identified as a critical threshold for low-level wetlands, as it represents the viability of wetland plant seeds based on research on Ribbon weed (Vallisneria australis). Ribbon weed is a common aquatic plant, having seeds that will remain viable in dry wetland sediments for up to nine years (Roberts and Marston 2011). The maximum spell for the 278 GL and the 350 GL scenarios breach this threshold, representing an increased risk to this parts of the floodplain. All other scenarios eliminate this critical threshold.

Table 10: Spell statistics for the 3,500 ML/d for fourteen day flow indicator at Brenda on the Culgoa River

Water Recovery Scenario Maximum time between events (years)




10.8 7 2


10.1 5 1


8.0 5 0


7.6 5 0

350 GL- targeted recovery 10.1 6 1


7.6 5 0

415 GL - recovery above 7.6 5 0

Page 44

Water Recovery Scenario Maximum time between events (years)



Basin Plan


Floodplain and wetland connectivity results (lower Balonne)

Riparian, inner, mid and outer floodplain connectivityRiparian, inner, mid and outer floodplain connectivity is achieved through flooding flows that results in water breaking out from the channels and spread outwards (laterally). The size of the flow (and the shape of the land) determine how far out away from the river channel the water spreads. These floodplain connectivity flows are an essential component of many dryland river systems. When rivers overflow and floodwaters extend across the floodplain there is an exchange of nutrients which replenishes the floodplain and river. There is also a dispersal of seeds and organisms, which are important processes for the lifecycle needs of many animals including fish, waterbirds, and other aquatic life.

There are four floodplain connectivity flow indicator, of increasing size (in terms of how far they spread out).

1. Riparian connectivity — this flow event is expected to flood (inundate) riparian and near channel areas of the Culgoa River and larger distributary channels. At this flow, there would be inundation of river red gum, ephemeral wetlands and lignum communities with about 3% of the floodplain becoming connected to the river (including both areas of Queensland and New South Wales). Other benefits include the maintenance of habitat areas for animal species such as amphibians and reptiles; and maintaining features within the river channel, such as benches and waterholes, which also provide habitat and refuge for a range of species. By providing water to maintain healthy riparian vegetation, this flow would also help to reduce river bank collapse and erosion.

2. Inner floodplain connectivity — this flow event is expected to flood (inundate) riparian and wetland communities, as well as inundating around 15% of the floodplain. In doing so, at least a third of coolibah and lignum along the channels of the Culgoa and Narran rivers would be inundated. This flow is also high enough to allow anabranches to carry water out onto the floodplain, particularly in the northern floodplain sections. Inundation associated with this flow would maintain or improve the health of wetlands on the inner floodplain, which would provide foraging habitat for waterbird, including migrating birds.

3. Mid floodplain connectivity — this flow event is a significant floodplain event whereby floodwaters would emerge from the Culgoa River in the vicinity of bifurcation 1, travel across the floodplain and re-enter the Culgoa River downstream near the Woolerbilla

Page 45

gauge. This event would also be expected to enter woodlands, and provide direct pathways for the exchange of material between the river and floodplain.

4. Outer floodplain connectivity — this flow event, though infrequent, has been shown to be vital for maintaining the productivity of grasslands on the outer floodplain, which make up nearly 40% of the total flood dependent vegetation. The maintenance of these grasslands is important for providing terrestrial animal habitat and energy sources for aquatic animals. Infrequent large-scale events such as these also facilitate mass exchanges of nutrients, sediment and animals, and are important for sustaining a wide variety of communities across the floodplain, and providing longer term moisture to these areas. Most of the ephemeral wetlands across the floodplain would also be inundated at these flows.

The durations for each of the four flow indicators represents the median event duration for all events that exceeded the threshold under the without development scenario. The selection of this median duration was to select an actual event duration that is typical for the threshold and is not skewed by the very large events (which would occur if the mean duration was selected).

For each of the flow indicators, timing is not constrained as floodplain flows historically occur at any time of year and water on floodplains and in wetlands is generally retained beyond the flow event. The frequency target range is expressed as the average number of years between an events. The target considers the types of floodplain vegetation that are inundated by the different flows and their average watering requirements (Casanova 2015; Roberts and Marston 2011). These frequencies also consider observed and modelled hydrology data to ensure consistency with the typical hydrology of the lower Balonne.

CB6. Riparian connectivity - 9,200 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of twelve days, any time of the year. The frequency target is to have, on average, no longer than 2-3 years between events.

• Under baseline conditions (2009 levels of development before water recovery), the time between the Culgoa River riparian connectivity flow occurring is much longer compared to without development conditions (5.6 years between an event compared to 1.3 years between an event). The target range is for this flow event to occur, on average, between two and three years. Many wetlands, creeks and river channels within the riparian zone would have been inundated with at least this frequency under

Page 46

without development conditions. This frequency is also consistent with the flooding requirements of river red gum forest and lignum.

• The modelled water recovery scenarios do result in a noticeable improvement in reducing the time between riparian connectivity events (between 4 and 3.4 years depending on scenario). The difference is significant as it represents a six month reduction in the period between events on average, which would result in improved inundation frequency for the riparian zone.

• The 415 GL scenario has the greatest improvement compared to the other scenarios. The 278 GL and 320 GL scenarios result in the lowest increases in frequency.

• Not meeting the target, particularly in prolonged dry periods, may result in riparian red gum vegetation being in poorer condition. There could be reduced quality and extent of lignum, which is relied upon by many animals (including waterbirds) for habitat.

• There will be increased opportunities for floodplain specialist fish species to access anabranches and secondary channels for all scenarios, with the opportunities increasing with the larger recovery volume scenarios. This is significant as these fish generally have a life-span less than five years and benefit from having access to off-river habitats. These habitats form an important nursery for juveniles with significant habitat diversity and aquatic productivity.

• This type of flow is important for improving riparian vegetation condition and for geomorphic processes such as waterhole scouring and maintenance of benches and other floodplain features. Improvements in the frequency of this flow could reduce risks such as river bank collapse and sediment infilling of waterholes.

Spell analysis is discussed as a group (of floodplain indicators) further below.

CB7. Inner floodplain connectivity - 15,000 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of ten days, any time of the year. The frequency target is to have, on average, no longer than 3-4 years between events.

• Under baseline conditions (2009 levels of development before water recovery), the time between the Culgoa River inner floodplain connectivity flow occurring is much longer compared to without development conditions (7.1 years between events

Page 47

compared to 1.9 years between events). The target range is for this flow event to occur, on average, every three and four years. This frequency is consistent with the watering requirements for vigorous growth in river red gum woodlands and black box woodlands and to sustain lignum shrublands.

• The modelled water recovery scenarios do result in an improvement in reducing the time between inner floodplain connectivity events occurring (between 6.3 and 5.4 years depending on scenario). The difference is significant as it represents close to a year reduction in the period between events on average, with improved condition of this part of the floodplain.

• The 390 GL scenarios and the 415 GL scenario have the most improved frequency result compared to the other scenarios; however, these scenarios are well short of achieving the target range.

• Ecologically, over prolonged dry periods, these frequencies of inundation could result in vegetation being in poorer condition, and could reduce habitat diversity, including foraging habitat for migrating birds. As the frequency reduces, access to temporary, opportunistic and productive habitat would also be reduced for floodplain fish specialists and other water-dependent biota. The average period between events for some of the scenarios is longer than five years which is longer than the life-span for many of the floodplain specialist fish.

Spell analysis is discussed as a group (of floodplain indicators) further below

CB8. Mid floodplain connectivity - 24,500 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of seven days, any time of the year. The frequency target is to have, on average, no longer than 6-8 years between events.

• Under baseline conditions (2009 levels of development before water recovery), the time between the Culgoa River mid floodplain connectivity flow occurring is much longer compared to without development conditions (8.7 years between an event compared to 3.5 years between an event). The target range is for this flow event to occur, on average, between eight and six years.

• The modelled water recovery scenarios do result in an improvement in reducing the time between mid-floodplain connectivity events occurring (between 8.1 and 7.6 years depending on scenario). This represents a six-month reduction in the period between events across the scenarios.

Page 48

• The 278 GL, 320 GL, 390 GL and 415 GL scenarios meet the target range. The other scenarios (320 GL pro rata and the 350 GL) are very close to also achieving the target range.

Page 49

• Ecologically, there is a reasonable likelihood of scenarios that meet the target supporting woodland communities including blackbox and coolibah. As a secondary consideration, these scenarios would be expected to support river red gum and lignum in the middle part of the floodplain, although their condition is likely to be poorer, and more frequent subsequent watering may be required to improve their health and vigour. This flow event is important as it represents conditions where there is more significant flooding with floodwater returning to the rivers. Exchange of nutrient rich flows between the river and floodplain links terrestrial and aquatic ecosystems, supporting food webs, is also likely to be enhanced across the scenarios and especially under the scenarios that with the largest improvement in frequency.

• Under extended dry periods, there may be increased risks of declining condition of woodland communities for the 320 GL and 350 GL scenarios compared to the other scenarios.

Spell analysis is discussed as a group (of floodplain indicators) further below.

CB9. Outer floodplain connectivity - 38,000 ML/day on the Culgoa River, measured at the Brenda gauge, for a minimum of six days, any time of the year. The frequency target is to have, on average, no longer than 10-20 years between events.

• Under baseline conditions (2009 levels of development before water recovery), the time between the Culgoa River outer floodplain connectivity flow occurring is much longer compared to without development conditions (28.5 years between an event compared to 9.5 years between an event). The target range is for this flow event to occur, on average, between 10 to 20 years. It is based on the requirements of grassland communities similar to those of the lower Balonne.

• The modelled water recovery scenarios both result in an increase and decrease in the time between outer floodplain connectivity events occurring (between 38 and 16 years, depending on scenario).

• The 390 GL and 415 GL scenarios achieves the target range; but the 415 GL has a greater level of frequency improvement (16.3 years, on average, between events).

Page 50

• The 278 GL and 320 GL scenarios result in a poorer outcome from the baseline condition. For these scenarios, inundation of this part of the floodplain is expected to be similar to baseline conditions. The large shift in frequency is caused by the reduction in the peak flow of one outer floodplain connectivity event. For this flow event, the required duration is not met, and the loss of this event results in the large change in the period between events as there are so few events of this magnitude.

• Ecologically, both 390 GL scenarios and the 415 GL scenario are likely to provide the greatest improvement to the productivity of grasslands on the outer floodplain, which is important for providing terrestrial animal habitat and energy sources for aquatic animals. Improvements to the mass exchanges of nutrients, sediment and animals, resulting from the increased number of large inundation events, will help sustain a wide variety of communities across the floodplain, and provide longer term moisture to these areas. The likelihood of a healthy outer floodplain is considerably reduced under the 320 GL and 278 GL scenarios.

Spell analysis is discussed below and covers the set of Lower Balonne floodplain and wetland connectivity indicators.

There are critical intervals between inundation events that are described in the literature that can be considered for lower Balonne floodplain and wetland connectivity. These are based on a range of information sources including: a review of the scientific literature regarding the water requirements of five floodplain tree species from across the Murray–Darling Basin (Casanova 2015); an assessment of the viability of wetland plant seeds based on research on Ribbon weed (Roberts and Marston 2011); and anecdotal information from landholders from the lower Balonne region. These thresholds, while not prescriptive and based on average requirements, do provide a means to compare modelled water recovery scenarios. The thresholds identified include:

• 5 years between inundation events for the inner floodplain (area inundated by a flow rate of 9,200 ML/d at Brenda). This is based on anecdotal information indicating that coolibah trees begin to die after this time period.

• 9 years between inundation events for low-level wetlands, representing the viability of wetland plant seeds based on Ribbon weed (Vallisneria australis), a common aquatic plant, having seeds that will remain viable in dry wetland sediments for up to nine years (Roberts and Marston 2011)

• 10 and 13 years between inundation events for river red gum (Eucalyptus camaldulensis) forests and woodlands, representing an indicative period to cause decline from good to critical condition for modelling state and transitions in floodplain vegetation (Casanova 2015).

• 14 years between inundation events for black box (Eucalyptus largiflorens), representing an indicative period to cause decline from good to critical condition for the purposes of modelling state and transitions in floodplain vegetation (Casanova 2015).

The maximum period between events for the floodplain and wetland connectivity flow indicators is shown in Table 11. These periods are much longer than the identified thresholds. However, the maximum period between events shows that there are major reductions between scenarios for the 9,200 ML/d flow indicator and the 15,000 ML/d flow

Page 51

indicator. This shows that some water recovery scenarios can shift the high risk prolonged dry spells. The major differences include:

The 415 GL, 390 GL, 350 and 320 GL B scenarios show the largest change with reductions in the 9,200 ML/d flow indicator.

Only the 415 GL and 390 GL scenarios show change (reduction) for the 15,000 ML/d flow indicator.

The 320 GL A and 278 GL scenarios show no change for any of the flow indicators The 24,500 ML/d and 38,000 ML/d flow indicators show no change in the maximum

period between events for any of the water recovery scenarios.

The sequencing of the flow indicator events results in very long spells that are longer than the indicative thresholds identified in the literature. To further consider this, analysis was done to relax the duration requirement. This analysis yielded results that better align with the scientific literature regarding thresholds and shows differences between scenarios.

Table 12 and 13 presents the results of this analysis for the 9,500 ML/d and 15,000 ML/d indicators.

Table 11: Maximum period between events for the four floodplain and wetland connectivity flow indicators in the Lower Balonne for the modelled scenarios.

Water Recovery Scenario

9,200 SFI Maximum time between events (years)





28.5 55.1 55.1 55.1


28.5 55.1 55.1 55.1


28.5 55.1 55.1 55.1


19.2 55.1 55.1 55.1

350 GL- targeted recovery

19.2 55.1 55.1 55.1


19.2 28.5 55.1 55.1

415 GL - recovery 19.2 28.5 55.1 55.1

Page 52






above Basin Plan

Without development 5.3 8.9 10.9 55.1

Table 12: Spell statistics for the 9,200 ML/d threshold


4 years between events





11 0 0 0


7 0 0 0

320 GL - recovery below Basin Plan

7 0 0 0


7 0 0 0


7 0 0 0


7 0 0 0

Without development 1 0 0 0

Table 13: Spell statistics for the 15,000 ML/d threshold.






Baseline - 2009 pre- 8 4 4 1

Page 53






water recovery


10 3 1 0

320 GL - recovery below Basin Plan

10 3 1 0


9 3 1 0


10 1 1 0


10 1 1 0

Without development 4 0 0 0

Page 54

Narran Lakes flow indicators resultsThe Narran Lakes flow indicators include floodplain and wetland connectivity indicators (and their sub-components) is shown in Figure 1.

Figure 22: Breakdown of Narran Lakes flow indicators into flow component groups and sub-groups

Floodplain and wetland connectivity results (Narran Lakes)

Provision of key habitatThe Narran Lakes includes landforms of channels, wetlands, and floodplains; it doesn’t just refer to the actual lakes. Narran Lakes is known an internationally significant site with respect to migratory birds and waterbirds, and the entire system provides a range of important ecological functions. Many of these functions are driven by large-scale wetland filling, where upstream waters from the Narran River pour into the system filling lakes and flooding floodplains.

When the wetlands dry up, the dead aquatic vegetation, invertebrates and fish form a rich organic substrate, which in turn supports a different vegetation community (e.g. herb fields). When the next watering event occurs, the organic substrate and decaying vegetation provides for a boom of food resources for quickly developing populations of macroinvertebrates and wetland plants, and the process starts again.

Narran Lakes can also provide a drought refuge as the lakes can retain water for up to two years following inundation. These refuges may protect populations which can re-colonise the Narran River system, and potentially other distributary rivers, during subsequent watering events.

The conditions of the Narran Lakes are highly variable and include times when the wetlands and lakes are filled with water as a result of large -scale wetland filling. Large-scale wetland filling in the Narran Lakes tends to fill from north to south (given big enough volumes), with

Page 55

Narran Lakes UEA

4 flow indicators


4 flow indicators

Provision of key habitat

4 flow indicators

the exception that Narran Lake in the south progressively receives some water at even relatively low inflows. In the northern lakes, flows generally fill Clear Lake first, and when the lake is at sufficient water levels, flow then fill Back Lake and Long Arm. However, the complex landform nature of the Narran Lakes system means that the pattern of inundation is also very complex, and may change over time.

Wetland filling provides ideal growing conditions for a range of vegetation, and provides critical breeding and nursery habitat for native fish species (such as the critically endangered silver perch and the endangered olive perchlet) and nesting waterbirds (such as the straw-necked ibis). When conditions are suitable, straw-necked ibis waterbirds congregate in large numbers to breed in the Narran Lakes (with nest numbers that can easily exceed 100,000). Large-scale wetland filling in the Narran Lakes also provides habitat for amphibians, water-dependent invertebrates, and reptiles.

Four site-specific flow indicators have been selected for the Narran Lakes to reflect a flow regime that would support the vital habitat and foraging requirements of migratory birds and waterbirds.

1. Key rookery habitat - This indicator targets the key rookery habitat of the northern lakes (Clear Lake, Back Lake, Long Arm and the adjacent lignum swamp).

2. Habitat of the northern lakes and floodplains - this indicator targets the habitat of the northern lakes and northern floodplains.

3. Large-scale waterbird breeding - this indicator targets providing opportunities for large-scale waterbird breeding events; and based on historic data, results in a guaranteed successful breeding outcome (though not always at a large scale) for straw-necked ibis. A flow event of this magnitude would be beneficial for vegetation on a third of the northern floodplain and about 20% of the central floodplain.

4. Habitat of the Narran Lakes, and northern central floodplains - this indicator targets habitat across all of the Narran Lakes and much of the northern and central floodplain.

For each of the flow indicators, timing is not constrained as floodplain flows historically occur at any time of year and water on floodplains and in wetlands is generally retained beyond the flow event.

NL1. Key rookery habitat - 25,000 ML total volume over a 60 day period on the Narran River, measured at the Wilby Wilby gauge, any time of the year. The frequency target is to have, on average, no longer than 1-1.3 years between events.

Page 56

• Under baseline conditions (2009 levels of development before water recovery), the average time between the Narran River key rookery habitat flow occurring is much longer compared to without development conditions (1.3 years between an event compared to 0.6 years between an event). The target range is for this flow event to occur, on average, between 1.3 and 1 year. This is to provide inundation regularly enough to allow large clumps of lignum (approximately 60% cover) to thrive in the key bird rookery habitats.

• The modelled water recovery scenarios do result in an improvement in reducing the time between the key rookery habitat flow events occurring (between 1.2 and 1 year depending on scenario); and because of the distribution of the target range, all scenarios achieve the lower end of the targeted ecological outcomes.

• The 278 GL scenario is the only scenario that performs differently, and also results in the smallest increase in frequency away from the baseline condition.

• Ecologically, the long-term health of vegetation in the parts of the northern lakes inundated by this event is likely to be maintained (Clear Lake, Back Lake, Long Arm and the adjacent lignum swamp). This is vital habitat for waterbird rookery sites. These flows may also support waterbird breeding (though unlikely to be in large numbers) and can provide breeding and nursery habitat for native fish species (the endangered olive perchlet). The most risk is represented under the 278 GL scenario as this scenario shows the least improvement towards the upper end of the target range.

NL2. Habitat of the northern lakes and floodplains - 50,000 ML total volume over a 90 day period on the Narran River, measured at the Wilby Wilby gauge, any time of the year. The frequency target is to have, on average, no longer than 1.3-1.7 years between events.

Page 57

• Under baseline conditions (2009 levels of development before water recovery), the average time between the Narran River northern lakes and northern floodplain habitat flow occurring is much longer compared to without development conditions (2 years between an event compared to 0.8 years between an event). The target range is for this flow event to occur, on average, between 1.7 and 1.3 years. This is to provide inundation regularly enough to allow large clumps of lignum (approximately 20% cover) to thrive in the key bird rookery habitats.

• The modelled water recovery scenarios result in a considerable improvement in reducing the time between events occurring (between 1.7 and 1.4 years). The scenarios with an average period of 1.4 years include 11 additional events compared to the 278 GL scenario and 5 more compared to the 320 GL scenario. This increase in the number of events provides a range of benefits including improvements to the long-term condition of vegetation and more opportunities for aquatic biota and waterbirds to make use of the Narran Lakes.

• All scenarios meet the target range, with the 320 GL pro rata, 350 GL, both 390 GL scenarios and the 415 GL scenario having the greatest improvement to the top of the target range.

• Ecologically, the long-term health of vegetation in the northern lakes is likely to be maintained (Clear Lake, Back Lake, Long Arm and the adjacent lignum swamp). The increased inundation, as compared to the 25 GL flow indicator, is likely to improve the filling of the northern lakes, inundate larger areas of lignum that are important foraging sites and improve the condition of riparian vegetation. These flows are also more likely to support waterbird breeding (though unlikely to be in the thousands of nests). The 278 GL scenario has relatively more risk as this scenario shows the least improvement towards the upper end of the target range.

Page 58

NL3. Large-scale waterbird breeding - 154,000 ML total volume over a 90 day period on the Narran River, measured at the Wilby Wilby gauge, any time of the year. The frequency target is to have this event occur, on average, at least once every 4-5 years.

• Under baseline conditions (2009 levels of development before water recovery), the average time between the Narran River large-scale waterbird breeding habitat flow occurring is much longer compared to without development conditions (8.3 years between an event compared to 2.6 years between an event). The target range is for this flow event to occur once, on average, every four and five years. This is to provide conditions suitable for breeding to take place twice in a straw-necked ibis' breeding lifespan.

• The modelled water recovery scenarios do result in an improvement in reducing the average period of time between an event occurring (between 7.2 and 6.3 years depending on scenario). However, none of the modelled scenarios reach the target range. This is a significant difference as it represents

• The 415 GL, 350 GL and 320 GL pro rata scenarios perform equally well reducing the average period of years between events to 6.3. The 278 GL results in the smallest improvement from baseline.

• Based on the (average) frequency results above, it is likely that some waterbirds (e.g. straw-necked ibis) will only have one opportunity to breed in their lifetimes. At this breeding recurrence, some waterbird populations are likely to continue to decline, particularly if on-going research shows that other suitable breeding areas are usually synchronised with Narran Lakes.

Spells analysis looked to see how often two events in any eight (and ten) year period occurred (not on average) (see Table 14). Even under without development condition, there were periods where straw-necked ibis would not have had suitable conditions in the Narran Lakes to breed (for example, there was a fifty-year dry period near the turn of the century (1800 to 1900) where two opportunities in every eight/ten did not occur). Notwithstanding, the 415 GL scenario results in much stronger frequency of the large-scale waterbird breeding habitat flow than the other scenarios with 32% of all eight year sequences including two breeding opportunities (compared to a 27% for all other scenarios). Analysis of the two opportunities in any ten year period resulted in the same trend, with the 415 GL resulting 42% of all ten years period having two breeding opportunities (with all other scenarios resulting a range from 37%-36%).

Page 59

Table 14: Spells analysis for the waterbird breeding flow indicator (2 in 8/10 years)

Water Recovery Scenario % of time 2 events occur in any 8 year period

% of time 2 events occur in any 10 year period

Baseline - 2009 pre-water recovery 20 27


27 36

320 GL - recovery below Basin Plan 27 36

350 GL- targeted recovery 27 37

390 GL - fully implemented Basin Plan 27 37

415 GL - recovery above Basin Plan 32 42

Without development 79 90

NL4. Habitat of the Narran Lakes, and central and northern floodplains - 250,000 ML total volume over a 180 day period on the Narran River, measured at the Wilby Wilby gauge, any time of the year. The frequency target is to have, on average, no longer than 8-10 years between events.

• Under baseline conditions (2009 levels of development before water recovery), the average time between the Narran Lakes and central and northern floodplains flows occurring is much longer compared to without development conditions (13.8 years without an event compared to 5.3 years without an event). The target range is for this flow event to occur, on average, between eight and ten years. This range represents the critical period for survival of lignum in Narran Lakes (in small form and sparse) and other key vegetation species on northern and central floodplains.

Page 60

• The modelled water recovery scenarios do result in a noticeable improvement in reducing the time between riparian connectivity events occurring (between 12.2 and 9.1 years depending on scenario).

• The 415 GL, 350 GL and 320 GL pro rata scenarios achieve the target range, with the 320 GL pro rata performing strongest (9.1 years). The 390 GL scenario does not perform well (11 years), but is still better than the 287 GL scenario (12.2 years).

• The ecological consequence of the 320 GL, 350 GL and 415 Gl scenarios is likely to be healthy vegetation habitat across all of the Narran Lakes and much of the northern and central floodplains. This are is vital habitat for waterbird rookery sites and can provide breeding and nursery habitat for native fish species (such as the critically endangered silver perch and the endangered olive perchlet).

• Under the 278 GL, 320 GL and both 390 GL scenarios, the lack of suitably frequent inundation is likely to be a reduction in health of the floodplain vegetation (compared to those scenarios that achieved the target range), with flow on consequences for the extent and availability of floodplain habitats.

Page 61

Barwon–Darling flow indicators The 11 Barwon–Darling flow indicators include in-channel connectivity indicators (and their sub-components) and floodplain and wetland connectivity indicators (and their subcomponents) as shown in Figure 2

River channel connectivity results (Barwon–Darling)

Small in-channel freshesSmall in-channel freshes are important to provide regular opportunities for aquatic animals to access habitats (including those associated with snags and in-channel benches). It also allows some recruitment of fish and other aquatic species; and the absorption of additional carbon and nutrients into the aquatic food web to increase in-stream primary production.

There are three (Bourke, Louth and Wilcannia) flow indicators specified for protecting small in-channel freshes in the Barwon–Darling.

1. The Bourke flow indicator small fresh event should provide regular access to in-channel habitat and movement opportunities along much of the Barwon–Darling River for many aquatic animals. Based on the mapping done, this event would inundate 60% of snags and 90% of the mapped benches between Walgett to Brewarrina, and around 50% of the snags and 80% of mapped benches between Brewarrina and Bourke. It should also provide fish movement opportunities between Brewarrina and Bourke (hundreds of km), and into tributaries such as the Culgoa River, and between Bourke and the next weir downstream towards Louth (Weir 19A), noting that the weir at Bourke remains a barrier to fish passage at this flow.

2. The Louth flow indicator small fresh event should provide improved opportunities for in-channel specialists such as Murray cod through better access to key snag habitat between Brewarrina and Tilpa. This small in-channel event should also inundate 40% of mapped snags on the Darling River between Bourke and Tilpa, and 80% of

Page 62

Barwon Darling UEA

11 flow indicators


7 flow indicators

Small in-channel freshes

3 flow indicator

Large in-channel bankfull flows

4 flow indicators


4 flow indicators

Riparian, inner, mid & outer floodplain

connectivity4 flow indicators

mapped benches between Bourke and Tilpa. Such a flow would allow better access for Murray cod to critical snag habitat which is a key requirement to initiate breeding.

3. The Wilcannia small fresh flow indicator event is specified to occur twice yearly. The specification of two flow events in a year at Wilcannia reflects that this type of flow in this reach of the Darling River typically occurs both in summer and winter. These two in-channel events would inundate around 90% of these in-channel benches. As a result, carbon and nutrients will be absorbed into the river, which, depending on a number of related factors, can trigger in-stream primary production which fuels aquatic ecosystems, resulting in improved condition of fish and other aquatic animals.

The duration of the in-channel flow indicators at each flow indicator gauge is consistent with the system hydrology and has been informed by ecological considerations (outlined in the EWR Reports). Generally the durations target spawning and hatch requirements, and carbon and nutrient absorption rates.

The small in-channel flow indicators associated with habitat access and in-stream primary productivity (i.e. the Bourke and Wilcannia flow indicators) are considered to provide benefits to fish throughout the year and can therefore be met at any time. The indicator at Louth aims to improve opportunities for habitat access and fish spawning, and therefore excludes the two coldest months when fish responses are expected to be less vigorous.

BD1. Small in-channel fresh - 6,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of fourteen days, any time of the year. The frequency target is to have the flow event occur between 80-90% of years.

• Under baseline conditions (2009 levels of development before water recovery), the percentage of years with a Darling River (Bourke) small in-channel flow is much less often than without development conditions (66% of years with an event compared to 96% of years with an event). The target range is for this flow event to occur 80-90% of years.

• The modelled water recovery scenarios markedly improve the frequency of the Bourke small in-channel indicator.

• All scenarios except the 278 GL scenario and the 390 GL current river operations scenario achieve the target range. Improvement is equal amongst those that achieve

Page 63

the range, except for the 320 GL pro rata which results in a slightly lower frequency result.

• This frequency range aims to provide regular maintenance and condition opportunities for the fish community, particularly important for short-lived fish that need regular freshes to complete important life-cycle stages.

• The ecological outcome from the scenarios that meet the target range (and also very likely for the 278 GL scenario given how close it is to the target range) is the inundation of 60% of snags and 90% of the mapped benches between Walgett to Brewarrina, and around 50% of the snags and 80% of mapped benches between Brewarrina and Bourke. Further, native fish will have the opportunity to move between Brewarrina and Bourke (hundreds of km), and into tributaries such as the Culgoa River, and between Bourke and the next weir downstream towards Louth. The duration of the flow events and its frequency are particularly important for short-lived fish populations. Access to varied habitat and the ability to move freely gives native fish the opportunity to find varied food resources; access different habitats, and ultimately helps establish and/or support healthy and resilient populations.

BD2. Small in-channel fresh - 6,000 ML/day on the Darling River, measured at the Louth gauge, for a minimum of twenty days, between August and May. The frequency target is to have the flow event occur 70% of years.

• Under baseline conditions (2009 levels of development before water recovery), the percentage of years with a Darling River (Louth) small in-channel flow is much less often than without development conditions (58% of years with an event compared to 91% of years with an event). The target point for this flow event to occur 70% of years.

• The modelled water recovery scenarios slightly improve the frequency of the Louth small in-channel indicator with the 350 GL scenario increasing the frequency the most (64% of years with an event).

• No scenario achieves the target point, and they do not perform distinctly differently (between 61-64%).

• The ecological outcome of this result, under all scenarios, is that access to snag habitat (between Brewarrina and Tilpa) is likely to be often compromised for native fish populations. This is particularly relevant for in-channel specialists such as Murray cod where access to snag habitat is a key requirement to initiate breeding. Under all

Page 64

scenarios regular opportunities for fish movement and in-channel specialist spawning (for species such as Murray cod and freshwater catfish) is also likely to be compromised This can lead to reduced population size, health and/or resilience.

BD3. Small in-channel fresh - 6,000 ML/day on the Darling River, measured at the Wilcannia gauge, for a minimum of fourteen days, twice each year. The event can occur any time of the year. The frequency target is to have the flow event occur between 45-60% of years.

• Under baseline conditions (2009 levels of development before water recovery), the percentage of years with a Darling River (Wilcannia) small in-channel flow is much less often than without development conditions (42% of years with an event compared to 77% of years with an event). The target range is for this flow event to occur 45-60% of years.

• The modelled water recovery scenarios do improve the frequency of the Wilcannia small in-channel indicator from baseline, and all scenarios reach the target range.

• While only small differences between the scenarios exist, the 390 GL current river operations results in the largest increase in frequency (50%).

• The ecological outcome of this result is the inundation of around 90% of in-channel benches between Tilpa and Wilcannia, which would lead to increased productivity benefits to the river system's food web. Subject to light penetration, this will trigger in-stream primary production which fuels aquatic ecosystems, resulting in improved condition of fish (especially important for pre-spawning fitness) and other aquatic biota

Large in-channel freshesThe main environmental outcomes associated with the large in-channel freshes in the Barwon–Darling system are regular opportunities for movement (and migrations for spawning), extensive access to habitats (including those associated with snags and in-channel benches) as well as hydrodynamic diversity, opportunities for recruitment, and absorption of additional carbon and nutrients into the aquatic food web to increase in-stream primary production.

There are four (Bourke x 2, Louth and Wilcannia) flow indicators specified for protecting large in-channel freshes in the Barwon–Darling.

Page 65

1 & 2. The two Bourke indicators aim to provide fish passage between Walgett and Menindee, a river distance of over 1,100 km. Such flows provide opportunities for fish migration, especially for fish that undertake large-scale migrations such as golden perch and silver perch. The 'two event' flow indicators would enhance recruitment of larvae and juveniles as the second flow event provides for dispersal, access to habitat, and suitable prey.

3 & 4. The Louth and Wilcannia indicators are designed to provide additional large-scale river channel connectivity. The flows would be expected to inundate over 95% of snags and in-channel benches between Brewarrina and Wilcannia, and provide upstream and downstream access to in-channel habitat to promote pre-spawning, larval, and juvenile fitness for many native fish. The flows would also result in greater access to off-channel wetlands, important for many fish such as olive perchlet and Darling River hardyhead (as well as other aquatic species).

Research in the Barwon–Darling catchment (Brennan et al. 2002) showed that large in-channel flows (around 20,000 ML/d at Bourke, Louth and Wilcannia) inundated between 15% and 60% of the wetlands along the river, depending on the location along the river system between Mungindi and Menindee. These flows would also improve the condition of riparian vegetation, particularly in the middle parts of river banks

The duration of these flow indicators differ for each flow indicator and they do serve different purposes. Generally, they relate to the hatch time of native fish eggs (first Bourke indicator); breeding outcomes (second Bourke indicator and Louth); and absorption of carbon and nutrients from inundated benches and the dispersal and movement of aquatic species. The Wilcannia duration also links to spawning and recruitment outcomes for fish species that are opportunistic and quicker to respond such as bony bream, golden perch and spangled perch.

The timing for the two 10,000 ML/d flow indicators at Bourke and the 21,000 ML/d flow indicator at Louth is August to May. The two coldest months of the year are excluded because fish responses are expected to be subdued. This timing is also consistent with existing knowledge of spawning and recruitment requirements for Murray cod (August to December) and the other native fish species in the northern basin. The timing for the Wilcannia indicator is any time of year, as the benefits of increased in-stream primary productivity provided by the release of carbon and other nutrients into the water column would provide benefits to aquatic food webs throughout the year, and also provide expanded access to habitat.

BD4. Large in-channel freshes - 10,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of fourteen days, between August and May. The frequency target is to have the flow event occur 60-80% of years.

Page 66

• Under baseline conditions (2009 levels of development before water recovery), the percentage of time with a Bourke large in-channel flow event is much longer compared to without development conditions (54% of the time compared to 89% of the time). The target range is for this flow event to occur 60-80% of the time.

• The modelled water recovery scenarios do result in an improvement in reducing the time between Bourke large in-channel flow events (between 57-60% depending on scenario). While all scenarios perform similarly, only the 415 GL scenario actually reaches the target range.

• The ecological outcome of this result is likely to be sufficient opportunities for regular fish passage between Walgett and Menindee, which is of significant important for those species that undertake large-scale migrations such as golden perch and silver perch. Benches and snags will also be inundated more regularly (than baseline), which provides increased access to aquatic habitat. Combined, the opportunity to move, spawn and access habitat is likely to improve mixing of populations (genetic diversity) and enhance conditions for breeding, especially for flow dependent specialist fish.

BD5. Large in-channel freshes - 10,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of fourteen days, twice each year. The event should occur between August and May. The frequency target is to have the flow event occur between 25-35% of years.

Page 67

• Under baseline conditions (2009 levels of development before water recovery), the percentage of time with a Bourke large in-channel flow event (two per year) is much longer compared to without development conditions (20% of the time compared to 42% of the time). The target range is for this double flow event to occur 25-35% of the time.

• The modelled water recovery scenarios do result in an improvement in reducing the time between Bourke large in-channel double flow events.

• The water recovery scenarios perform similarly (between 22-23% of years); however, no scenario reaches the target range.

• The ecological outcomes of this result is that under prolonged dry conditions, there would be considerable stress on the recruitment of native fish larvae and juveniles, particularly as the second flow event provides for dispersal, access to habitat, and suitable prey. Golden perch and freshwater catfish may also not receive sufficiently regular cues for migration, which may result in a reduction in the population size or health over time.

BD6. Large in-channel freshes - 21,000 ML/day on the Darling River, measured at the Louth gauge, for a minimum of twenty days, between August and May. The frequency target is to have the flow event occur 40% of years.

• Under baseline conditions (2009 levels of development before water recovery), the percentage of time with a Louth large in-channel flow event is much longer compared to without development conditions (32% of the time compared to 54% of the time). The target range is for this flow event to occur 40% of the time.

• The modelled water recovery scenarios do not result in any improvement in how often Louth large in-channel flow events occur, and remain at baseline condition (32% of years), despite the recovery scenario applied.

• The likely ecological outcome of this result is that in-channel specialist native fish, such as Murray cod, will have compromised opportunities for re-colonisation, breeding and habitat access. During prolonged dry periods, this may result in reduced population size, health and/or resilience.

Page 68

BD7. Large in-channel freshes - 20,000 ML/day on the Darling River, measured at the Wilcannia gauge, for a minimum of seven days, any time of the year. The frequency target is to have the flow event occur 45-60% of years.

• Under baseline conditions (2009 levels of development before water recovery), the percentage of time with a Wilcannia large in-channel flow event is much longer compared to without development conditions (39% of years compared to 70% of years). The target range is for this flow event to occur 45-60% of the time.

• The modelled water recovery scenarios do result in an improvement in reducing the time between Wilcannia large in-channel flows occurring (between 44%-45% of years depending on scenario).

• While all scenarios are close to achieving the target range, the 278 GL and 415 GL fall just short (44%).

• Particularly for those scenarios that meet the target range, the ecological outcome is a sufficiently regular access to off-channel wetlands, important for many fish (such as olive perchlet and Darling River hardyhead). Access to the wetlands provides a boom in food supply and access to ideal spawning and nursery habitats. Further, regular nutrient cycling from the inundation of benches will also be suitably frequent which in turn promotes increased in-stream primary productivity.

Floodplain and wetland connectivity results (Barwon–Darling)

Riparian, inner, mid and outer floodplain connectivityFloodplains (and connected wetlands) are important for the ecology of rivers, and have important ecological values in their own right. Many, if not all, of the species that live in rivers depend in some way on connecting with floodplains (Mussared 1997). Flooding flows that punctuate dry spells and inundate floodplains are an essential component of dryland river systems (Leigh et al. 2010; Sheldon and Thoms 2006). When rivers overflow and floodwaters extend across the floodplain there is an exchange of nutrients which replenishes the floodplain and river, and a dispersal of seeds and organisms (Thorp, Thoms and Delong 2008).

Floodwaters inundate a diverse suite of habitats, including wetlands and floodplain vegetation communities such as forests, woodlands and grasslands (Figure 3). This provides

Page 69

habitat for many different animals, both terrestrial and aquatic, and triggers off-stream primary production. Animals such as amphibians and reptiles depend on areas of the river banks and floodplain (DSITIA 2013, Ayers, Mazzer and Ellis 2004). Wetlands, creeks and anabranches provide foraging habitat for bird species (Wen et al. 2011). Fish move into the anabranches and wetlands that are inundated with floodwaters to feed in these nutrient-rich areas, and many small bodied fish use these habitats to breed (Nichols et al., 2012, NSW DPI 2015). Periodic inundation is important for the health of the river system and the floodplain (Thorp, Thoms and Delong 2008).

There are four (Bourke x 3 and Wilcannia) flow indicators specified for protecting floodplain and wetland connectivity in the Barwon Darling.

1. The riparian connectivity indicator at Bourke of 30,000 ML/d in the Darling River at Bourke for a minimum of 24 consecutive days is expected to provide water to river red gum, ephemeral wetlands, and lignum communities. These flows (around bankfull) are important for maintaining geomorphic features within the bed of the river channel, such as sand bars, benches and waterholes (DSITIA 2013; Wilkinson, Keller and Rutherfurd 2004). Maintaining flows around bankfull provides for riparian vegetation that assists in reducing river bank erosion; and may start to inundate near-channel wetlands.

2. The inner floodplain connectivity indicator of 45,000 ML/d in the Darling River at Bourke for a minimum of 22 consecutive days will fill near-channel habitats and floodplain wetlands. This size of flow (around 45,000 ML/d at Bourke) is important in connecting the majority of wetlands with the main-channel (i.e. inundation of 90% of wetlands). The majority of river red gum communities between Bourke and Louth would be inundated. Further, this inner floodplain flow would exceed the commence-to-flow of channels connecting lakes known to be important for waterbirds, including Poopelloe Lake, Victoria Lake and Eucalyptus Lake in the Talyawalka–Teryaweynya system (Brennan et al. 2002). The flow would be sufficient to water river red gum woodlands communities fringing the Talyawalka Anabranch. The recent inundation analysis (MDBA 2016d) suggests that at this flow at least 10% of coolibah woodlands and floodplain grasslands in the Bourke to Louth region would also be inundated.

3. The mid floodplain connectivity indicator of 65,000 ML/d in the Darling River at Bourke for a minimum of 24 consecutive days will inundate a large proportion of the floodplain (MDBA 2016d). This facilitates the dispersal of organic matter (e.g. seeds, animals and carbon), water, nutrients and sediment across the floodplain, creeks and wetlands (Thoms 2003). More than 30% of the floodplain grasslands and coolibah woodlands found at higher elevations become inundated at this flow rate (MDBA 2016d). This flow also inundates black box woodland communities around the Talyawalka -Teryaweyna system (MDBA 2012a).

4. The outer floodplain connectivity indicator of 2,350 GL into Talyawalka Creek within a single flow event is associated with filling of lakes in the Talyawalka -Teryaweynya system, as well as inundating wetlands and vegetation communities on the outer Darling floodplain more broadly.

The Talyawalka–Teryaweynya system supports extensive areas of floodplain vegetation (Jenkins and Briggs 1997) including black box, lignum, and grasslands. When inundated, the system’s lakes provide foraging habitat for large numbers of waterbirds (DEWHA 2010). In particular, Kingsford et.al (1997) identified Pelican Lake within the system, and the Darling

Page 70

River floodplain near Louth, as places known or predicted to support 20,000 or more waterbirds.

For each of the flow indicators, timing is not constrained as floodplain flows historically occur at any time of year and water on floodplains and in wetlands is generally retained beyond the flow event.

BD8. Riparian connectivity - 30,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of twenty four days, any time of the year. The frequency target is to have the flow event occur, on average, once every 2-3 years.

• Under baseline conditions (2009 levels of development before water recovery), the average time between a Bourke riparian connectivity flow event is much longer compared to without development conditions (4.1 years without an event compared to 1.8 years). The target range for this flow event is for there to be between 2-3 years on average between events.

• The modelled water recovery scenarios do result in a small improvement in reducing the time between Bourke riparian connectivity flows (between 3.9 to 3.8 3.6 years depending on scenario).

• The 320 GL B pro rata, 350 GL, 390 GL and 415 GL scenarios are most effective at decreasing the time between events (3.8 years), with the other two scenarios performing equally at 3.9 years. The target range is not met under any scenario.

• The ecological consequence of this result is that under prolonged dry periods, the flooding requirements to sustain vigorous river red gum forests and lignum, and to re-fill wetlands near the channel may not be achieved. Further, without the ideal flow frequency (such as in prolonged low flows), increases in river bank collapse and erosion problems may also occur as a result of decreased riparian vegetation cover. Maintenance of benches and waterholes for hydraulic diversity is also more likely to be compromised.

BD9. Inner floodplain connectivity - 45,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of twenty two days, any time of the year. The frequency target is to have the flow event occur, on average, once every 3.5-4 years.

Page 71

• Under baseline conditions (2009 levels of development before water recovery), the average time between a Bourke inner floodplain connectivity flow event is much longer compared to without development conditions (5.1 years without an event compared to 3.3 years). The target range for this flow event is for there to be between 4-3.5 years on average between events. This range is within the range of the watering requirements of river red gum woodlands, and would sustain wetland communities including lignum.

• The modelled water recovery scenarios do not result in any reduction in the average time between Bourke inner floodplain connectivity flows.

• All scenarios perform equally (and equal to baseline condition) and do not meet the target range.

• The ecological consequence of this result is that under prolonged dry periods, the flooding requirements to sustain river red gum woodlands (such as areas around the Talyawalka Anabranch) and wetland communities including lignum may not be achieved. Lakes such as the Poopelloe Lake, Victoria Lake and Eucalyptus Lake in the Talyawalka–Teryaweynya system would not have sufficiently regular inflows, which would have likely flow on consequences for waterbirds, including migrating birds.

• The result would also likely reduce access to temporary, opportunistic and productive floodplain and wetland habitat for many floodplain fish specialists and other water-dependent biota (aquatic insects as a key part of the aquatic food web).

BD10. Outer floodplain connectivity - 65,000 ML/day on the Darling River, measured at the Bourke gauge, for a minimum of twenty four days, any time of the year. The frequency target is to have the flow event occur, on average, once every 6-8 years.

Page 72

• Under baseline conditions (2009 levels of development before water recovery), the average time between a Bourke mid floodplain connectivity flow event is much longer compared to without development conditions (8.7 years without an event compared to 5.6 years). The target range represents the type of frequency these flows would have occurred historically in the Darling River, and is also reasonably consistent with and is close to the critical interval of river red gum woodlands and lignum.

• The modelled water recovery scenarios do not result in any reduction in the average time between Bourke mid floodplain connectivity flows.

• All scenarios perform equally (and equal to baseline condition) and do not meet the target range.

• The ecological consequence of this result is that under prolonged dry periods, the flooding requirements to sustain vigorous growth for black box would not be achieved. Further, given the target range was based on close to the critical interval of river red gum woodlands and lignum, these species/communities would also be at considerable risk under all scenarios during prolonged dry periods. However, it is acknowledged that these two species may remain alive if inundation occurs at a lower frequency, their condition is likely to be poor, and more frequent subsequent wetting may be required to improve their health and vigour.

BD11. Outer floodplain connectivity - Greater than 30,000 ML/d to achieve an annual inflow volume of 2,350 GL. This is measured on the Darling River at the Wilcannia gauge, and can occur at any time of the year. The frequency target is to have at least 7-10% of years include this event.

Page 73

• Under baseline conditions (2009 levels of development before water recovery), the average percentage time with a Wilcannia outer floodplain connectivity flow event is much longer compared to without development conditions (7% of years with an event compared to 11% of years). The target range represents the type of frequency these flows would have occurred historically in the Darling River.

• The modelled water recovery scenarios do not result in any increase in the percentage years with a Wilcannia outer floodplain connectivity flow.

• All scenarios perform equally (and equal to baseline condition); however due to the position of the target range, all scenarios meet the target; although no progress is made toward the 'low uncertainty' end of the target.

• The ecological consequence of this result is that under all scenarios the filling of lakes in the Talyawalka–Teryaweynya system, as well as inundating wetlands and vegetation communities on the outer Darling floodplain, is likely to be frequent enough to sustain broadly under to sustain many of the vegetation communities in those areas. The communities/species targeted by the frequency of this indicator include black box, lignum, and grasslands. Further, when inundated, the system’s lakes provide foraging habitat for large numbers of waterbirds, and also mass exchange of nutrients, sediment and recruitment and dispersal opportunities for a range of native aquatic animal and plant species.

Page 74

Border Rivers flow indicators The three Border Rivers flow indicators identify in-channel connectivity indicators (and their sub-components) only (as shown below in Figure 2).

Page 75

Border Rivers UEA

3 flow indicators


3 flow indicators

Large in-channel flows

3 flow indicators

Page 76

Gwydir River flow indicators The nine Gwydir River flow indicators identify in-channel and floodplain and wetland connectivity flow indicators (and their sub-components) as shown below in Figure X.

Page 77

Gwydir River UEA

9 flow indicators


2 flow indicators

Baseflows and large in-channel

flow2 flow indicators

Wetland and floodplain


Low lying wetlands and near channel

4 flow indicators

Low, mid outer floodplain

3 flow indicators

Page 78

Page 79

Page 80

Namoi River flow indicators The three Namoi River flow indicators identify in-channel and floodplain and wetland connectivity flow indicators (and their sub-components) as shown below in Figure X.

Page 81

Namoi River UEA

3 flow indicators


2 flow indicators

Baseflows and small in-channel

fresh2 flow indicators


connectivity1 flow indicator

Wetland and floddplain inundation

1 flow indicator

Page 82

Macquarie River flow indicators The four Macquarie-Castlereagh flow indicators identify floodplain and wetland connectivity flow indicators only (and their sub-components) as shown below in Figure X.

Page 83

Macquarie River UEA

4 flow indicators



Near channel

1 flow indicator

Low level floodplain

1 flow indicator

Mid level floodplain

1 flow indicator

Oute floodplain

1 flow indicator

Page 84

Appendix 2: Lower Balonne flow indicator assumptions and relative confidence in the evidence High – where specific studies or multiple lines of evidence have been used (including actual gauge data)

Pragmatic – where there aren’t specific studies or pieces of evidence, but an analysis of the characteristics of modelled baseline or without development flows has been undertaken

Flow indicator Main logic and assumptions Flow threshold/volume confidence

Duration confidence Timing confidence Frequency confidence (target range)

2 ML/d for 1 day, any time of the year at Weilmoringle (refuge waterholes on the Culgoa River). Frequency: 350-430 days average of the longest 10% periods. between events

The flow indicator identifies the longest 10% of periods between flows and compare this to the persistence time of waterholes. The logic is that any flow downstream of the refuge waterhole will have filled the river channel and the upstream waterholes.

Assumption 1: A network of waterholes in the lower part of the system with at least 0.5m water depth is needed to maintain resilient aquatic communities during dry spells.

Assumption 2: The 0.5m water depth is an appropriate estimate for the threshold to maintain suitable habitat.

Assumption 3: Jack Taylor Weir, Beardmore dam and the Barwon-Darling also provide drought refuges but there are issues with fish movement past structures.

Assumption 4: the top 10% of periods between flows are all longer no-flow spells that would stress refuge waterholes.

High

Critical lower threshold for flow to be observed downstream of refuge waterholes

High

One day is the minimum time step to observe flow in the model

High

The timing is not relevant as the aim is to reduce the longer and more risky no-flow spells whenever possible.

High-pragmatic

Based on waterhole persistence research (DSITI 2015). The 350 day target is to maintain at least seven waterholes along the Culgoa River, with at least four of these to a depth of more than 0.5 m. The 430 days target provides at least two refuge waterholes to at least 0.5 m.

The research gives high confidence regarding the availability of water in waterholes along the Culgoa at the end of dry periods. However there is uncertainty about the differences in quality of waterhole habitat.

A depth of 0.5 m is assumed to be an adequate water depth to allow the waterhole to persist in a suitable condition for aquatic species however, there is uncertainty about whether this is appropriate for all waterholes.

2 ML/d for 1 day, any time of the year at Narran Park (refuge waterholes on the Narran River). Frequency: 350-430 days average of the longest 10% periods. between events

The flow indicator identifies the longest 10% of no flow periods identified in the model and compares this to the persistence time of waterholes (350 days). The logic is that any flow recorded downstream of the refuge waterhole will have filled the river channel and topped up the refuge waterholes on the way through.

Assumption 1: A network of waterholes in the lower part of the system with at least 0.5m water depth is needed to maintain resilient aquatic communities during dry spells.

Assumption 2: The 0.5m water depth is an appropriate estimate for the threshold to maintain suitable habitat.

Assumption 3: Jack Taylor Weir, Beardmore dam and the Barwon-Darling provide drought refuges but there are issues with fish movement past structures.

High

Critical lower threshold for flow to be observed downstream of refuge waterholes

High

One day is the minimum time step to observe flow in the model

High

The timing is not relevant as the aim is to reduce the longer and more risky no-flow spells whenever possible.

High-pragmatic

Based on waterhole persistence research (DSITI 2015). The 350 day target is to maintain at least six waterholes along the Narran River, with at least four of these to a depth of more than 0.5 metres. The 470 days target provides at least two refuge waterholes to at least 0.5 m.

The research gives high confidence regarding the availability of water at the end of dry periods. However there is uncertainty about the quality of the different refugia and the 0.5 m depth threshold.

Page 85



Assumption 4: The top 10% of no-flow periods represent the longest set of dryspells experienced in the flow record that would have resulted in stress (reduction of loss) of refuge waterholes in the lower parts of the system.

1,000 ML/d for 7 days, any time of the year at Brenda (small fresh for channel connectivity). Frequency: 80% to 90% of years

The flow indicator describes a small in-channel fresh that would flow through the Culgoa and Narran rivers and at least part of the Bokhara River. This is based on analysis of observed flows for 1965-1975 (before development) and 2005-2015 (after development), and on modelled flows from the baseline scenario.

Assumption 1: This is the minimum flow to be confident about system scale connectivity (in reality smaller flows often provide the same level of connectivity).

Assumption 2: Small freshes need to occur close to annually to support a healthy river.

Assumption 3: A small fresh at any time of the year will provide benefits to the environment - irrespective of water temperature.

High

Observed and modelled flow data used to determine a flow size that has a high likelihood of riverine connectivity across the Lower Balonne.

High

Observed and modelled flow data used to determine a flow size that has a high likelihood of riverine connectivity across the Lower Balonne.

High-pragmatic

The high flow season is generally summer to autumn, and research from the nearby Moonie River shows fish move any time of the year (linked to water level rise, temp above 15 degrees and the first post winter flow). Average temp at Brenda is above this threshold between August and May, but it is assumed that a small fresh provide benefit to the ecosystem any time of the year.

Pragmatic

Based on expert advice for this flow type in the dryland rivers of the Northern Basin. It is not based on site-specific information (other than this small fresh occurring in 98% of years under WOD).

3,500 ML/d for 14 days, between Aug and May at Brenda (large fresh in the Culgoa). Frequency: 40% to 60% of years

This large in-channel pulse (close to bankfull) is expected to provide a range of outcomes with the main one being improved movement and breeding opportunities for flow dependent fish. Research from other catchments (including the nearby Moonie) shows rises in the river, velocity, the first post winter flow and temperature to be important to trigger fish to move larger distances.

Assumption 1: Fish in the Lower Balonne will respond to this type of flow (no site-specific data).

Assumption 2: The flow will provide opportunities for fish to move large distances and past in-channel barriers.

Assumption 3: Large freshes are important to maintain populations of flow dependent fish and to provide a range of other benefits (for example for riparian vegetation).

High

Based on providing a river rise of at least 2 m and an average velocity of 0.3 m/s. These flow attributes have been linked to fish movement responses in other catchments. The Qld government have other projects in the Lower Balonne to look at fish movement and the type of flows to overcome barriers that will further improve our understanding over time.

Pragmatic

Based primarily on the hatch time of Murray cod eggs which represents the longest requirement for native fish in the area. Duration is consistent with the typical duration for this type of event (median duration under without development is 17 days).

High

Linked to research about the water temperature for golden perch to move large distances (15-16 degrees). Golden perch is the main flow dependent fish in the region. Average temp at Brenda is above this threshold between Aug and May.

Pragmatic

Based on MDBA assessment of general advice from experts (50% of years would generally have a large in-channel fresh in the dryland rivers of the Northern Basin).

1,700 ML/d for 14 days, between Aug and May at Wilby Wilby (large fresh in the Narran). Frequency: 40% to 60% of years

This large in-channel pulse (close to bankfull) is expected to provide a range of outcomes with the main one being improved movement and breeding opportunities for flow dependent fish. Research from other catchments (including the nearby Moonie) shows rises in the river, velocity, the first post winter

High-emerging

Based on providing a river rise of at least 2 m and an average velocity of 0.3 m/s. These flow attributes have been linked to fish

Pragmatic

Based primarily on the hatch time of Murray cod eggs which represents the longest requirement for native fish in the area.

High

Linked to research about the water temperature for golden perch to move large distances (15-16 degrees). Golden perch is the main flow dependent fish

Pragmatic

Based on MDBA assessment of general advice from experts (50% of years would generally have a large in-channel fresh in the dryland rivers of the Northern Basin.

Page 86



flow and temperature to be important to trigger fish to move larger distances.

Assumption 1: Fish in the Lower Balonne will respond to this type of flow (no site-specific data).

Assumption 2: The flow will provide opportunities for fish to move large distances and past in-channel barriers.

Assumption 3: Large freshes are important to maintain populations of flow dependent fish and to provide a range of other benefits (for example for riparian vegetation).

movement responses in other catchments. The Qld government have other projects in the Lower Balonne to look at fish movement and the type of flows to overcome barriers that will further improve our understanding over time.

Duration is consistent with the typical duration for this type of event (median duration under without development is 15 days).

in the region. Average temp at Brenda is above this threshold between Aug and May.

9,200 ML/d for 12 days, any time of the year at Brenda (riparian connectivity). Frequency: average period between events of 2 to 3 years.

This event is expected to inundate riparian and near channel areas across the Lower Balonne, particularly river red gum, ephemeral wetlands and lignum communities.

Assumption 1: Flows at Brenda represent wider floodplain connectivity for the Lower Balonne.

Assumption 2: Changes in the spatial distribution of water recovery (between channels) can be represented by flows at Brenda.

Assumption 3: Frequency aims to represent the average flooding requirements of vegetation in this zone.

Assumption 4: Some of the past floodplain connectivity research used flows at St George. The MDBA related these flows to the Brenda gauge using average attenuation relationships. There is a large variation in the attenuation relationship. However, this information is not relied upon as the only source and forms part of the multiple lines of evidence.

Assumption 5: Between flooding, vegetation will use a variety of water sources including groundwater.

High

Based on multiple lines of evidence regarding floodplain connectivity, including past research (based on flows at St George), new research by the MDBA (based on historic inundation for river reaches across the Lower Balonne), assessments of bankfull flow rates and mapping of floodplain vegetation.

Pragmatic

An analysis of without development events was undertaken to determine the median duration. Flow events in the Lower Balonne are shorter than the period of inundation identified in the literature for vegetation in this part of the floodplain (up to several months). The median duration has been used to represent events that would typically inundate riparian and near channel areas in the Lower Balonne.

Pragmatic

Floodplain inundation is preferable in summer/autumn to correspond with the high flow season for the Lower Balonne but timing has not been constrained to reflect that the flows are unregulated.

High-Pragmatic

Consistent with the scientific literature describing the watering requirements of river red gum forests and lignum. Tested against the baseline and without development frequency to ensure consistency with the hydrology of the Lower Balonne.

15,000 ML/d for 10 days, any time of the year at Brenda (inner floodplain connectivity). Frequency: average period between events of 3 to 4 years.

This event is expected to inundate riparian and wetland communities as well as around 15% of the floodplain.



High

Based on multiple lines of evidence regarding floodplain connectivity, including past research (based on flows at St George), new research by the MDBA (based on historic inundation for river reaches across the Lower

Pragmatic

An analysis of without development events was undertaken to determine the median duration. Flow events in the Lower Balonne are shorter than the period of inundation identified in the literature for vegetation in this part

Pragmatic


High-Pragmatic

Consistent with the scientific literature describing the watering requirements of river red gum and black box woodlands and to sustain lignum. Tested against the baseline and without development frequency to ensure consistency with the hydrology of the Lower Balonne.

Page 87



Assumption 3: Frequency aims to represent the average flooding requirements of vegetation.

Assumption 4: Some of the past floodplain connectivity research used flows at St George and MDBA related these flows to Brenda using average attenuation relationships. There is large variation in attenuation. However, this information is not relied upon as the only source and forms part of the multiple lines of evidence.


Balonne), assessments of bankfull flow rates and mapping of floodplain vegetation.

of the floodplain (up to several months). The median duration has been used to represent events that would typically woodlands in the Lower Balonne.


This event represents large scale floodplain inundation whereby floodwaters break out from the Culgoa River around bifurcation 1 and travel across the floodplain. This event inundates about 40%, especially for woodlands, and provides direct pathways for exchange of material between the river and floodplain.






High

Based on multiple lines of evidence regarding floodplain connectivity, including past research (based on flows at St George), new research by the MDBA (based on historic inundation for river reaches across the Lower Balonne), assessments of bankfull flow rates and mapping of floodplain vegetation.

Pragmatic

An analysis of without development events was undertaken to determine the median duration. Flow events in the Lower Balonne are shorter than the period of inundation identified in the literature for the vegetation in this part of the floodplain (up to several months). The median duration has been used to represent events that would typically inundate the woodlands in the Lower Balonne.

Pragmatic


High-Pragmatic

Consistent with the scientific literature describing the watering requirements black box woodlands and to support coolibah in this zone. Tested against the baseline and without development frequency to ensure consistency with the hydrology of the Lower Balonne.


This infrequent event plays an important role in maintaining the productivity of grasslands on the outer floodplain. The maintenance of these grasslands is important for providing terrestrial animal habitat and to facilitate mass exchanges of energy, nutrients, sediment and animals between the river and floodplain.

Assumption 1: Flows at Brenda represent wider floodplain connectivity for the lower Balonne.

High

Based on multiple lines of evidence regarding floodplain connectivity, including past research (based on flows at St George), new research by the MDBA (based on historic inundation for river

Pragmatic

An analysis of without development events was undertaken to determine the median duration. The median duration has been used to represent events that would typically achieve broad scale

Pragmatic


Pragmatic

Consistent with the scientific literature describing the watering requirements of similar grasslands in parts of the Gwydir Wetlands and to support coolibah woodlands. Tested against the baseline and without development frequency to ensure consistency with the hydrology of the Lower

Page 88







reaches across the lower Balonne), assessments of bankfull flow rates and mapping of floodplain vegetation.

floodplain inundation. Balonne.

The outer areas of the floodplain tend to be flooded infrequently and have a lower degree of flood dependence. The vegetation communities found in these zones are adapted to large variations in flooding frequency, being able to withstand long periods of dry. There is also less site specific scientific information to provide additional information about the water requirements of the coolibah and grasslands in the Lower Balonne, and how this may vary with flood frequency. As such, the requirements for flooding are less certain and broad (10 to 20 years).

Page 89

Appendix 3: Narran Lakes flow indicator assumptions and relative confidence in the evidence (Still in preparation)

Appendix 4: Barwon–Darling flow indicator assumptions and relative confidence in the evidence (Still in preparation)

Page 90

Documents

Environmental Outcomes Report V5€¦ · Web viewThis is to provide inundation regularly enough to allow large clumps of lignum (approximately 60% cover) to thrive in the key bird