Murky waters running clearer? Monitoring, reporting and evaluation of the state of the Murray–Darling Basin after more than three decades of policy reform

The volume of water owned or controlled by Aboriginal organisations is increasing.

The Act states that the Basin Plan must have regard to ‘social, cultural, Indigenous and other public benefit issues’ (S21(4)(c)(v)) and include a description of the Basin water resources and the uses to which the Basin water resources are put (including by Indigenous peoples) (S22.1(b)). Water-resource plans must include requirements in relation to social, spiritual and cultural matters relevant to Indigenous people (S22(3)(ca)) and the MDBA must engage with the Indigenous community on the use and management of Basin water resources (S172(1)(ia)). The Water Amendment (Restoring Our Rivers) Act 2023 (Cth) states that the Basin Plan must ensure that the use and management of Basin water resources take into account spiritual, cultural, environmental, social and economic matters relevant to Indigenous peoples, including in relation to their knowledge, values, uses, traditions and customs (Commonwealth of Australia 2023, S3fa) and the MDBA must review and report on:

the extent to which the Basin Plan, including requirements relating to water resource plans, recognises and protects the interests of Indigenous people…opportunities for Indigenous people to participate in determining and developing priorities and strategies for the development or use of Basin water resources [S14B].

For over two decades, Indigenous peoples have sought to advance their objectives in water governance through Basin programs and policies (Godden et al. 2020; Hartwig et al. 2020). A central tenet of Indigenous ‘cultural flows’ is to increase access to and control of water through rights and entitlements (Moggridge and Thompson 2021). We consider the volume of water that Aboriginal organisations hold represents a suitable provisional indicator, noting refinement of such an indicator is a matter for Indigenous peoples to agree (Hartwig et al. 2021).

The volume of Indigenous water holdings in the New South Wales part of the Basin is very small and declined from 14.7 GL year⁻¹ in 2008–09 (66 entitlements held by 27 Aboriginal organisations; 0.17% of total non-environmental entitlements) to 12.1 GL year⁻¹ in 2018–19 (55 entitlements, 25 organisations and 12 GL year⁻¹ in 2020–21; 0.1% of total non-environmental entitlements for both 2018–19 and 2020–21) (Fig. 2a; Hartwig et al. 2020, 2021). We consider that the target for this indicator has not been met.

Fig. 2.

Indicators, Indigenous theme. (a) Water holdings by Indigenous organisations in the New South Wales Basin. (b) Volume of Commonwealth environmental water released to wetlands as flooding flows in areas represented by Indigenous organisations. Data sources are provided in the sections for Indicators 1 and 2 the Supplementary material.

The volume of water released by the CEWH and State agency partners as flooding flows to wetlands in areas under the jurisdiction of Indigenous organisations is steady or increasing. These areas include those under Native Title claims, Aboriginal Land Councils in New South Wales, Registered Aboriginal Parties in Victoria and Cultural Heritage Bodies in Queensland (Costanza-van den Belt et al. 2022, fig. 1 therein).

As for Indicator 1, implementation of the Basin Plan must have regard to the uses to which the Basin water resources are put, including by Indigenous peoples (S22.1(b)). The MDBA must engage with the Indigenous community on the use and management of Basin water resources (S172(1)(ia)).

We used flooding flows as an indicator because they represent a closer expression of Indigenous cultural relations with water than do in-channel flows (Moggridge and Thompson 2021). Ensuring that managed flow releases provide benefits to Indigenous peoples is required under the Basin Plan but has not yet been realised (Costanza-van den Belt et al. 2022). The MDBA stated a ‘clear and committed pathway for First Nations social and economic outcomes’ is necessary (Murray–Darling Basin Authority 2020a, p. 125). CEWO stated that they ‘value the ongoing contribution that First Nations people make to the planning and delivery of water for the environment’ (Commonwealth Environmental Water Holder 2024), although they do not report on the nature and extent of this engagement (Department of Climate Change, Energy, Environment and Water 2023b). Meaningful engagement would involve Indigenous peoples in cultural water planning (Murray–Darling Basin Authority 2020d, p. 3; Woods et al. 2022), including recognition of Indigenous watering objectives (Department of Climate Change, Energy, Environment and Water 2024a). The Monitoring, Evaluation and Research program (MER 2.0), run by CEWH, has committed to providing a minimum of 10% of catchment budgets for engagement with Aboriginal partners.

The volume of environmental water released by CEWO, in partnership with State agencies, as flooding flows to wetlands that are under the jurisdiction of Indigenous organisations varied between 102 and 721 GL year⁻¹ between 2012–13 and 2022–23, representing an annual average of 18% of the total volume of environmental water being released. Most of this water was accounted for at major wetlands such as Macquarie Marshes, Gwydir Wetlands and those of the central and lower Murray and lower Murrumbidgee. The time-series indicates a variable and fluctuating pattern, with no discernible trend (Fig. 2b).

The targets are (3) mean personal community income is increasing in real terms and (4) disparity between lowest and highest incomes is stable or declining.

The objective is to optimise economic outcomes within the Basin (S20d).

Commonwealth government water-market reforms and water-efficiency projects have been implemented with the intent of generating economic benefits for irrigators under conditions whereby less water is available. Basin governments and their agencies have recognised that changes in water policy and management could have negative effects on the livelihoods of irrigation communities. Such impacts have been widely reported (Sefton et al. 2020; Schirmer et al. 2021). Accordingly, it is important to measure the economic outcomes for Basin communities. Nominal personal income of communities (not adjusted for inflation) is reported at Local Government Area (LGA) scale (Australian Bureau of Statistics 2021). We included all LGAs that had most of their area located within the Basin and calculated the mean of annual personal incomes, adjusted to real (current) values, with 2011–2012 as the reference year, and compared LGAs containing irrigation districts with non-irrigation LGAs, as well as the difference between LGAs with highest and lowest incomes within each group as a measure of income disparity. Increase in income inequality in Australia since the 1980s has prompted a range of public policies to address economic disadvantage and its consequences (Senate Community Affairs References Committee 2014). Inequality in wages and wages growth is only one of several measures of income inequality and is less comprehensive than the more commonly used Gini coefficient (Fletcher and Guttmann 2013). We use change in wages disparity herein because data are publicly available at the LGA scale and can be scaled up to whole-of-Basin scale and partitioned into LGAs within and outside irrigation districts.

Between 2011–12 and 2020–21 there was little difference in income between LGAs in irrigation districts and non-irrigation LGAs in the northern and southern Basin. The grand mean for northern Basin irrigation LGAs was A$48,613 person⁻¹ year⁻¹, for northern Basin non-irrigation LGAs, it was A$46,839, for southern Basin irrigation LGAs, it was A$47,871, and for southern Basin non-irrigation LGAs, the grand mean was A$49,850. Income increased by between A$515 person⁻¹ year⁻¹ (southern Basin non-irrigation LGAs) to A$1082 (northern Basin non-irrigation LGAs) over the period of record (Fig. 3a).

Fig. 3.

Indicators, economic theme. (a) Mean personal community income in irrigation LGAs and non-irrigation LGAs in the northern and southern Basin. (b) Mean income disparity (range) in irrigation and non-irrigation LGAs in the northern and southern Basin. (c) Gross value of irrigated agricultural production (GVIAP; bars) and as a percentage of national GVIAP (line). (d) Production value of irrigation water (solid black line), with line of best fit (red dashed line) and irrigation water use. (e) Mean irrigation farm cash income (bars) and Commonwealth annual environmental water allocations. (f) Annual rate of return for irrigation farms. (g) Mean Basin farmland price compared with national mean. (h) Surface-water diversions by Basin State and total area irrigated, showing periods of drought (yellow) and flood (blue), and key events (vertical dashed lines).

Income disparity between the top and bottom five irrigation and non-irrigation LGAs in the northern and southern Basin showed the lowest difference was in southern Basin irrigation LGAs (grand mean A$9250 person⁻¹ year⁻¹), then northern Basin irrigation LGAs (A$15,074 and, northern Basin non-irrigation LGAs (A$15,579). Income disparity for these three groups showed inter-annual variation but was stable. Highest income disparity was for southern Basin non-irrigation LGAs (A$23,671) which increased over the period of record (from A$19,457 in 2011–12 to A$28,444 in 2020–21) (Fig. 3b). This increased disparity in southern Basin non-irrigation LGAs may reflect a strong wages growth in large regional centres such as Queanbeyan, Orange and Bathurst.

The data on personal income and income disparity do not support the narrative that irrigation communities have experienced significant economic disadvantage since the implementation of the Basin Plan, as asserted by Sefton et al. (2020), who specified 12 towns experiencing ‘acute social and economic conditions’ (p. 46), yet provided no data in their report to support this claim. Of the 12 towns mentioned, 6 were in Basin LGAs in the upper 50% for mean personal income and 3 were in the upper 20%.

(a) Gross value of irrigated agricultural production (GVIAP) is stable or increasing and (b) the trend is equal to or greater than the national average.

The objective is to have regard to the consumptive and other economic uses of Basin water resources (S21(4)(c)(ii)).

Almost half the value of Australian irrigated agricultural production is from the Basin (cf. findings below), underpinning national food security and regional economic output. The indicator can be used to assess whether the irrigation sector has been adversely affected economically by the implementation of the Basin Plan.

Real GVIAP in the Basin (Australian Bureau of Statistics 2019) has remained stable between 2010–11 and 2019–20 (grand mean A$6.83 billion year⁻¹; range: A$5.9 billion–8.42 billion year⁻¹) (Fig. 3c) but has declined as a proportion of national GVIAP from 46% in 2010–11 to 39% in 2019–20 (long-term average: 47%). Values were lowest in 2010–11 and 2012–13, which coincided with the end of the Millennium Drought (September 2010) and highest in 2017–18 following a wetter than average year in 2016–17. During the Millennium Drought, irrigators in the Basin received only approximately one-third of their water allocations. A 67% decline in water use from 2000–01 to 2008–09 resulted in a 14% reduction in nominal GVIAP and a 20% reduction in price-adjusted gross value (Kirby et al. 2014). Many irrigators were able to adapt their production systems by substituting different crops and inputs and by improving irrigation efficiency. Accordingly, GVIAP on its own appears not to be particularly sensitive in the short term to reductions in irrigation water availability because of the economic adaptability of irrigators (cf. also findings for Indicator 6, below).

GVIAP per unit of irrigation water used is steady or improving.

The objective is to achieve efficient and cost-effective water management (S3g).

The Commonwealth Government has funded irrigation-efficiency schemes to improve productivity and decrease irrigation water use, providing a rationale for including an indicator of the economic value of irrigated agriculture and efficiency of water use. GVIAP per unit of irrigation water use is an indicator of the production efficiency of water use.

GVIAP per gigalitre of irrigation water use doubled over the period from 2010–11 to 2020–21, from A$0.75 million GL⁻¹ to A$1.5 million GL⁻¹ (calculated from line of best fit, Fig. 3d). Production value tended to be lower in wet years (2010–11 to 2012–13), during which time irrigation water use was relatively high and greater areas of crops and pastures were grown as a consequence, indicating relatively low water use efficiency (Marsden Jacob Associates 2020, fig. 2 therein). In dry years, and with lower irrigation water diversions (2014–16 and 2018–21), production value was high, indicating increased water-use efficiency. This pattern reflects changes in irrigation water availability and diversions in relation to rainfall, but also the high adaptability of irrigators during dry periods in switching crops, decreasing the area under production and substituting irrigation water sources (cf. findings for Indicator 5 above). Also, in the northern Basin, which accounts for an average GVIAP of 26% of the Basin total and where floodplain water harvesting is widespread, total on-farm storage capacity is ~3375 GL (Brown et al. 2022). These dams fill in very wet years, providing water for irrigation in subsequent drier years. But floodplain harvesting is not accounted for specifically in the ABS water use on Australian farm reports (Australian Bureau of Statistics 2022). Volumes for the category of water source listed as ‘water taken from on-farm dams or tanks’ are grossly underestimated.

It would be incorrect to infer that the increase in production value of irrigation water has been due mainly to improved irrigation efficiency. Since 2012–13, there were major increases in plantings of high-value almonds and cotton in the southern Basin (SunRISE Mapping and Research 2022; Kennedy and Mackintosh 2024), as well as improvements in production. For example, cotton yield in New South Wales increased from 10.5 to 13.8 tonnes (Mg) ha⁻¹ between 2000–01 and 2018–19 (Marsden Jacob Associates 2020, p. 15). Such changes accord with the objective of the NWI, whereby adaptation could be achieved through water trading to shift water use to higher values and profitability. But the setting of this objective in the NWI did not account for the extent to which water availability would be limited by the Basin Plan and climate change. To limit competition for increasingly scarce irrigation water, the Almond Board of Australia urged the New South Wales and South Australian governments to follow the Victorian government and impose a moratorium on water licences for new plantings in the Lower Murray (Kennedy and Mackintosh 2024). The upward trend for this indicator cannot continue indefinitely. As demand for water outstrips supply and limits production, the area of permanent plantings of high-value crops will contract and the production value will decline accordingly.

Cash income and rate of return of irrigation farms is increasing.

The objective is to optimise economic outcomes within the Basin (S20d).

Cash income and rate of return can be used to assess whether irrigators have been disadvantaged economically by the recovery of environmental water under the Basin Plan, as claimed in submissions to inquiries into the Basin Plan (Productivity Commission 2018, pp. 157–159; Senate Select Committee 2021, pp. 201–202).

Cash income is a more direct indicator than is GVIAP, which is influenced by changes in commodity prices. Data on cash income (Ashton 2020) cover only 4 years since commencement of the Basin Plan (2006–07 to 2015–16), but the time series corresponds with a period during which the majority of environmental water holdings were acquired, including by the Sustainable Rural Water Use and Infrastructure program (Murray–Darling Basin Authority 2023b, p. 55).

Trends in cash income do not necessarily reflect changes in profitability. The rate of return represents the net profit of irrigation farm investment. Data on rates of return for dairy and rice farms at Basin scale cover the period from 2006–07 to 2018–19 (to 2017–18 for cotton) (Goesch et al. 2020). Data for horticulture farms at Basin scale are available for the period from 2006–07 to 2015–16 (Ashton 2020).

Mean annual cash income showed inter-annual variation, reflecting commodity prices and water availability. Over the 10-year period of assessment, cash income rose from A$101,625 to A$301,775, a three-fold increase (calculated from lines of best fit; Fig. 3e). Increase in cash income for rice farms was highest (from A$51,400 to A$231,800; a 4.5-fold increase), then dairy (from A$53,200 to A$202,900; a 3.8-fold increase), cotton (from A$248,400 to A$655,200; a 2.6-fold increase) and then horticulture (from A$53,500 to A$117,200; a 2.2-fold increase). During the same period, Commonwealth cumulative environmental water entitlements increased from less than 20 GL to 2432 GL and annual allocations from about 10 GL to 1682 GL (Fig. 3e). These figures indicate that when most water was recovered for the environment, irrigators were not adversely affected economically, despite fluctuations in commodity prices and water availability (Ashton 2020).

Between 2006–07 and 2015–16, the mean annual rate of return was highest for cotton farms at 4.5%, increasing from 1.9% in 2006–07 to 7% in 2011–12 after the breaking of the Millennium Drought (Fig. 3f), before declining to 1.4% in 2017–18 after the onset of drought in 2018. Horticulture farms (including almond orchards) were the next-most profitable, with a rate of return of 2.1%, recovering from 0.9% in 2006–07 during the Millennium Drought to 3.3% in 2015–16. Dairy farms had a mean of 2%, recovering from −0.3% in 2006–07 to 6.1% 2013–14, before falling back to 1.4% in 2018–19. Rice farms, hit hardest by the Millennium Drought, had a mean of 1.6%, recovering from −0.3% in 2006–07 to 4.9% in 2014–15, then declining to −0.2% in 2018–19 after the onset of drought that year. The period of greatest acquisition of environmental water coincided with a period of marked increases in profitability of irrigation farms.

Farmland price is improving.

The objective is to optimise economic outcomes within the Basin (S20d).

Farmland price (A$ ha⁻¹) is the average price paid for broadacre farmland (land used for large-scale crop production) (2006–07 to 2022–23), on the basis of data from Australian Bureau of Agricultural and Resource Economics and Sciences (2024). It is an indicator of the value of farmland as a capital asset.

The mean price per hectare (real dollar value) of broadacre farmland in the Basin increased from A$2259 in 2006–07 to A$9277 in 2022–23, a 4.1-fold rise, compared with a national increase from A$4055 to A$8769, a 2.2-fold rise. Prices rose exponentially from 2014–15 to 2022–23 (Fig. 3g). These figures represent a mean annual rate of increase of 25.7% for the Basin and 13.5% nationally. Farmland price growth over the 17 years was highest for the New South Wales Riverina (36.6% per year) and lowest for Queensland Darling Downs and Central Highlands (17.9% per year).

It is important to note that this indicator includes land used for dryland and irrigation cropping (Australian Bureau of Agricultural and Resource Economics and Sciences 2024), so a note of caution is required in interpreting the implications for irrigation enterprises. However, in regions dominated in extent by irrigated agriculture (including the New South Wales Riverina and the central north of Victoria), prices were well above the average for the Basin. The high rate of increase in the price of broadacre farmland indicates strong growth in the capital value of irrigation farms, with clear economic benefits for farmers. According to ABARES, there has been a decline in the rate of consolidation of farms, indicating fewer small farmers are exiting the industry. Fewer farms are being sold and at much higher prices than in 2010 (Chan 2024).

The volume of surface water diversions for consumptive purposes is declining.

The objectives are to ensure the return to environmentally sustainable levels of extraction for water resources that are overallocated or overused (S3di), and to ensure the establishment and enforcement of environmentally sustainable limits on the quantities of surface water and ground water that may be taken from the Basin water resources (S20b).

The total surface-water diversions for consumptive purposes by each Basin State and Territory represent an expression of the magnitude of the economic benefit provided by Basin water resources. The ESLT is the level at which water can be taken from a water resource which, if exceeded, would compromise key environmental assets, ecosystem functions, the productive base or key environmental outcomes. The vast majority of surface-water diversions from the Basin are used for irrigation; hence, this indicator is included under the economic theme. So, to restore water to the environment, the volume of take for consumptive purposes must be reduced. The SDL caps the volume of Basin water that can be taken for consumptive use and must reflect the ESLT. We used data from annual water-take reports (e.g. Murray–Darling Basin Authority 2023b) for the period from 1983–84 to 2021–22. We note that the period covers three different water-accounting frameworks (Basin Cap from 1996–97 to 2018–19, transitional SDLs with no compliance for 2019–20 onwards and SDLs with compliance for WRPs accredited before 2020–21). The accounting methods differ for each. Because such a large proportion of surface-water diversions is used for irrigation, we examined the relationship between take and area irrigated, using area data from the Water Use on Australian Farms survey (e.g. Australian Bureau of Statistics 2022) between 2005–06 (when data at Basin scale were first published) and 2020–21, the most recent published survey.

Irrigation diversions accounted for 92.2% of surface-water take between 1999–2000 and 2011–12. Diversions follow rainfall and river inflows during wet and dry years (Fig. 3h; Murray–Darling Basin Authority 2023b). Between 1983–84 and 1999–2000, surface-water take was ~11,000 GL year⁻¹, before falling sharply during the Millennium Drought (2001–10) to 4000 GL year⁻¹ in 2008–09, with declines in all States except South Australia (with an annual entitlement of up to 1850 GL year⁻¹ under the Murray–Darling Basin Agreement, but less during dry years). Take recovered after the Millennium Drought broke, reaching a maximum of nearly 14,000 GL in 2012–13 at the commencement of the Basin Plan. Since then, in New South Wales, the largest consumptive water user (mean 55% of total volume, 1983–84 to 2021–22), Victoria (mean 32% of total volume) and Queensland (7% of total volume), take declined by between one-third and one-quarter from 2012–13 to 2021–22. Basin-wide, take declined from almost 14,000 GL year⁻¹ in 2012–13 to ~11,000 GL year⁻¹ in 2020–21. Yet, this reduction has not been reflected by increases in river flows (cf. findings for Indicator 11, below).

Increased irrigation water-use efficiency does not necessarily mean a reduction in water consumption, because increased efficiency tends to result in increased water use and total area irrigated (Adamson and Loch 2018; Wheeler et al. 2020). Increase in irrigated area is a major driver of increased water consumption (Puy et al. 2021). The area equipped for irrigation in Australia (i.e. the area that could be irrigated; AEI) increased between the calendar years 2000 and 2015, mostly in the Basin according to Mehta et al. (2024, table 2 and fig. 4 therein). These authors consider that AEI, rather than the area actually irrigated, allows for better assessment of broad temporal trends and minimises errors from attempting to integrate different agricultural census data. But, in reality, the actual area of the Basin irrigated each year shows considerable inter-annual variation and has not increased overall between the water years 2005–06 and 2020–21 (mean 1.3 × 10⁶ ha; Fig. 3h).

Ramsar wetlands are flooded with managed environmental flows at frequencies and extents that meet their environmental water requirements.

To give effect to relevant international agreements (to the extent to which those agreements are relevant to the use and management of the Basin water resources) (S3(b)) and the Basin Plan must promote the conservation of Ramsar wetlands (S21(3)b).

The Ramsar Convention obliges the Commonwealth Government to promote the wise use of all wetlands and maintain the ecological character of Ramsar sites. Maintenance of ecological character of those Ramsar wetlands located on floodplains depends on floods of a frequency, extent, depth and duration to meet the environmental water requirements of their constituent species and vegetation communities, for example, river red gum (Eucalyptus camaldulensis) forests and woodlands, black box (E. largiflorens) woodlands and lignum (Duma florulenta) shrublands (Chen et al. 2021; Kirsch et al. 2022). Floods may be generated by natural flows and releases of held environmental water by Commonwealth and State agencies.

Changes in the characteristics of flow and flood regimes of the major wetlands of the Basin are not regularly assessed by government agencies. Two-monthly data from 1988–2022 on the extent of flooding at the Basin scale are publicly available (Penton et al. 2023), but require specialist spatial-analysis tools to consolidate, visualise and assess the annual extent of flooding of natural wetlands. As a proxy measure, we used the maximum annual extent of open water and wet ground of Ramsar wetlands (1999–2000 to 2020–21), by using the Digital Earth Australia Wetlands Insight Tool (Geosciences Australia, see https://maps.dea.ga.gov.au, accessed 28 July 2024; Dunn et al. 2023). A limitation of this indicator is that Ramsar wetlands represent only a small proportion of the total area of natural wetlands that can be flooded with managed environmental flows.

Maximum extent of flooding and wet areas varied with wet and dry years, with highest values in 2010–11 when the Millennium Drought broke, and in 2016–17, a year with above-average rainfall in most of the southern Basin and the upper reaches of the Border Rivers, Gwydir and Namoi catchments (Stewardson and Guarino 2018, fig. 4 therein) (Fig. 4a). During the period after the Basin Plan commenced (2012–13 to 2020–21), maximum flood extent declined slightly, on the basis of lines of best fit to time series in Fig. 4a, at Riverland (14.8–12.1% of 30,615 ha), Barmah–Millewa Forest (24.2–23.5% of 66,535 ha) and Gunbower–Koondrook–Perricoota Forest (18.1–16.3% of 56,043 ha), and increased at Macquarie Marshes (10.2–21.9% of 19,850 ha), Gwydir Wetlands (8.4–13.9% of 823 ha) and Narran Lakes (9.3–20.2% of 8454 ha). Less than 17% of the total area (193,035 ha) of these six Ramsar sites was flooded in 6 of 9 years, representing 67% of the time for which data are available since commencement of the Basin Plan. At Macquarie Marshes, environmental watering of the Southern Nature Reserve has effectively ceased, with only one flood between 2016–17 and 2020–21, whereas the Northern Nature Reserve has received environmental water annually during this period (Schweizer et al. 2022).

Fig. 4.

Indicators, environmental theme. (a) Percentage of Ramsar wetland area flooded, showing periods of drought (yellow) and flood (blue); (b) Ramsar wetlands: frequency of years of large and small floods; (c) Ramsar wetlands: vegetation condition, based on mean annual leaf area index, showing periods of drought (yellow) and flood (blue); (d) Ramsar wetlands: frequency of years of good or poor vegetation condition; (e) river flows: observed v. expected; (f) abundance of waterbirds and number of breeding species; (g) frequency of occurrence of threatened species with lines of best fit; and (h) number of fish kills in the New South Wales Basin. G-K-P, Gunbower–Koondrook–Perricoota; d/s, downstream.

Ramsar wetlands remain dependent on intermittent, unpredictable, high unregulated flows for the maintenance of their ecological character. Environmental watering had some buffering effect during dry years. The frequency of years during which the extent of flooding was particularly high or low (greater than the mean ± 0.5 standard deviations, s.d.) supports this finding (Fig. 4b). The proportion of years of low flood extent for the six wetlands prior to commencement of the Basin Plan was 0.47 (6.2 years in 13) compared with 0.26 (2.3 years in 9) after the commencement of the Basin Plan. The proportion of years of high flood was 0.26 (3.3 years in 13) prior to the Basin Plan and 0.15 (1.3 years in 9) after. The Basin was in drought for 8/13 years prior to the Basin Plan but only 2/9 years after (Fig. 4a).

Between 1999–2000 and 2020–21, an average of 17.9% of the total area of 193,085 ha of the six wetlands was flooded, or at least wetted, which is somewhat higher than the estimate of 10% (2000–01 to 2020–21) based on the area flooded to a depth of ≥0.5 m (Chen et al. 2021). The small extent flooded indicates that environmental water requirements of Ramsar wetlands are not being met, despite delivery of environmental flows to these wetlands in almost all years, except Narran Lakes, which was watered only in 2016–17. Only 21% of the volume of managed environmental flows released annually (2012–13 to 2020–21) (mean 1886 GL) was delivered as flood events (Chen et al. 2021).

Vegetation condition in Ramsar wetlands is maintained or improving.

The objective is to promote sustainable use of Basin water resources to protect and restore the ecosystems, natural habitats and species that are reliant on the Basin water resources and to conserve biodiversity (S21(2)(b)), as well as to give effect to relevant international agreements (to the extent to which those agreements are relevant to the use and management of the Basin water resources) (S3(b)).

River red gum forest and woodland, coolibah (Eucalyptus coolabah) woodland, black box woodland and lignum shrubland are the wetland vegetation communities prioritised for managed environmental flows in the Basin-wide Environmental Watering Strategy (BWS), a statutory instrument of the Basin Plan (Murray–Darling Basin Authority 2019b). The BWS contains targets for maintenance of these communities in each catchment. Condition of river red gum forests and woodlands has been monitored (Cunningham et al. 2009) and reported in the Basin Plan Evaluation (Murray–Darling Basin Authority 2020a). Leaf area index (LAI) is a measure of crown extent (Asner et al. 2003). Floodplain trees shed their leaves to reduce transpiration and water uptake as a response to low water availability during drought (Gibson et al. 1994). Flooding recharges soil water and subterranean palaeochannels filled with coarse sediment (Colloff 2014, p. 9), leading to increased water uptake, leaf area (by epicormic growth) and transpiration, resulting in improved tree condition (Doody et al. 2015). We used mean annual LAI of vegetation, derived from remote sensing (Centre for Water and Landscape Dynamics 2020), in six Ramsar wetlands as an indicator of wetland woody-vegetation condition.

Mean annual LAI tends to be higher in wetlands with extensive forests of river red gum (Barmah–Millewa Forest, Gunbower–Koondrook–Perricoota Forest and Macquarie Marshes) or stands of water couch (Paspalum distichum) (Gwydir Wetlands) than in wetlands with more scattered lignum shrublands and open riparian woodland (Riverland and Narran Lakes). LAI at the six Ramsar sites showed no trend between calendar years 2000 and 2023 (Fig. 4c), other than improvements during wet years and declines during droughts. This pattern becomes more apparent when comparing frequency of years during which LAI was particularly high or low (greater than the mean ± 1 s.d.) (Fig. 4d). The mean proportion of low-LAI years for the six wetlands prior to implementation of the Basin Plan (2000–11) was 0.29, compared with 0.1 after (2012–23) and the proportion of high-LAI years was 0.15 before and 0.21 after. These figures might suggest that the Basin Plan is associated with improved vegetation condition at these wetlands, but the Basin was in drought for 9 of the 12 years prior to the Basin Plan and only 3 of 12 years after; so, no such inference can be drawn.

River flows at hydrological indicator sites increase as predicted under the Basin Plan.

The objective is to promote sustainable use of Basin water resources so as to protect and restore the ecosystems, natural habitats and species that are reliant on the Basin water resources and to conserve biodiversity (S21(2)(b)).

To restore and maintain healthy flow-dependent ecosystems, the volume, duration and frequency of flow events must accord with standards at the 124 ‘hydrological indicator sites’ (HISs). Infrequent high flows are needed to inundate wetland ecosystems at higher elevations on the floodplains, for example, black box woodlands.

The Wentworth Group (2020) reported on observed v. expected flows under the Basin Plan at river gauges at 27 HISs across the Basin, with data adjusted for climatic variability, providing a basic indicator of whether sufficient water is available to maintain in-channel and wetland habitats and biota. An estimated 5591 GL was expected to flow across the South Australian border between 2012–13 and 2019–20, but the actual flows were 22% lower. Flows at 24 of the 27 sites were lower than expected, 13 received less than 75% of the expected flows and three received less than half (Fig. 4e). Year 2020 was the first year when the Wentworth Group analysis was undertaken, and it is likely to be ongoing. Modelled flow is based on past events and flows now are less than in the past, indicating a declining trend.

Abundance of populations of waterbirds is increasing.

The objective is to give effect to relevant international agreements (to the extent to which those agreements are relevant to the use and management of the Basin water resources) (S3(b)), and to protect, restore and provide for ecological values (S3(d)(ii)).

Waterbird populations are closely linked to the ecological condition of rivers and wetlands (Reid et al. 2013). Since calendar year 1983, the annual Eastern Australian Waterbird Survey has monitored abundance of waterbirds, numbers of species breeding and wetland area (Kingsford et al. 2020; Porter et al. 2020, 2024). Waterbird abundance and the number of breeding species are indicators of the availability of suitable habitat and resources within wetlands, which, in turn, are driven by rainfall, river inflows and flooding (Kingsford et al. 2017).

Annual waterbird abundance and number of breeding species in the Basin (1983–2023) have declined since the high values between 1983 and 1985 (Fig. 4f). Low values coincide with the Millennium Drought (2000–10) and the 2017–20 drought, with increases in 2010–11 after the Millennium Drought broke and after wet years in 2016–17 and 2022–23.

Under current and predicted future climates, waterbird targets under the Basin Plan (increase in abundance by 20–25%, in frequency of breeding events by 50%, in breeding abundances by 30–40% and maintenance of species diversity; Murray–Darling Basin Authority 2019a, pp. 44, 45) are unlikely to be met (Bino et al. 2021). Abundance is predicted to continue to decline and remain below targets without improved policies and management for environmental watering to support waterbird populations.

Frequency of occurrence of selected flow-dependent threatened species is maintained or improving.

The objective is to give effect to relevant international agreements (to the extent to which those agreements are relevant to the use and management of the Basin water resources) (S3(b)), and to protect and restore ecosystems, natural habitats and species that are reliant on the Basin water resources and to conserve biodiversity (S21(2)(b)).

Monitoring of changes in distribution, abundance and occurrence of flow-dependent threatened species is required for their conservation, as detailed in mandated Species Recovery Plans (SRPs). But monitoring is fragmented and only a minority of flow-dependent threatened species have SRPs. Increases in occurrence and abundance provide evidence that environmental flows are meeting the water requirements of threatened species (Ryan et al. 2021); so, improved monitoring should be a priority for the Basin Plan.

We used data on annual frequency of occurrence (the number of times per year a species was recorded) from the Atlas of Living Australia (see https://www.ala.org.au/, accessed 8 February 2024) for five of the species selected by Ryan et al. (2021), updated for calendar years from 2019 to 2023, to assess the effectiveness of environmental flows for threatened species conservation (southern bell frog, Litoria raniformis; Australasian bittern, Botaurus poiciloptilus; Australian painted snipe, Rostratula australis; trout cod, Maccullochella macquariensis; and silver perch, Bidyanus bidyanus). These species are listed as Threatened under the Environmental Protection and Biodiversity Conservation Act 1999 (Cth) and one or more State and Territory conservation acts. They are distributed predominantly in the regulated catchments of the southern Basin, where almost 90% of environmental water was released between 2012–13 and 2018–19 (Chen et al. 2021). There are SRPs for southern bell frog and trout cod and draft plans for Australasian bittern and Australian painted snipe. Silver perch is Critically Endangered under the Environmental Protection and Biodiversity Conservation Act but has no SRP.

Frequency of occurrence for all species between 2012 and 2023 was highly variable among years; however, on the basis of lines of best fit to time-series in Fig. 4g, increased markedly for southern bell frog (mean annual occurrence 11.6–164.9), increased slightly for Australasian bittern (55.2–75.3), stayed more-or-less the same for silver perch (12.9–16.8) and trout cod (13.7–9.1) and declined markedly for Australian painted snipe (33.8–7.2). The overall pattern is variable, with no clear benefits of environmental watering for flow-dependent threatened species. The distribution of species except Australasian bittern appears to have contracted since 1990 (Ryan et al. 2021, fig. 1 therein). Environmental watering has had some transient positive outcomes for some threatened species in some locations on some occasions, but monitoring and reporting is fragmented and inconsistent (Ryan et al. 2021). If the Basin Plan is to deliver on its environmental objectives, including Australia’s international environmental treaty obligations, a comprehensive monitoring framework is required to support a more effective, targeted approach to managing environmental water requirements for threatened species.

The number of water quality-related fish kills is declining.

The objective is to protect, restore and provide for the ecological values and ecosystem services of the Murray–Darling Basin (taking into account, in particular, the impact that the taking of water has on watercourses, lakes, wetlands, ground water and water-dependent ecosystems that are part of the Basin water resources and on associated biodiversity) (S3(d)(ii)).

Freshwater fishes are sensitive to declines in water quality, particularly low dissolved oxygen concentration, which is often exacerbated by low flows (Australian Academy of Science 2019). Changes in frequency and magnitude of fish kills over time reflect changes in water quality, which may include de-oxygenation caused by blackwater events, cyanobacterial (blue–green algal) blooms or runoff from bushfire-affected areas (cf. Indicator 23 below). We used data from the New South Wales Department of Primary Industries fish kill database to estimate frequency and magnitude of fish kills (>1000 individuals killed).

Between calendar years 1980 and 2023, the mean annual number of fish kills in the New South Wales Basin increased from 1.5 to 3.6 (calculated from the line of best fit, Fig. 4h). During the period after the Basin Plan was implemented (2012–23), the annual number of events rose from 1.3 to 3.9, compared with an increase from 3 to 4.6 for the 12 years prior (2000–21). There have been six very large fish kills (>100,000 individuals) during the 2012–23 period, including the 2019–20 and 2023 kills at Menindee Lakes in New South Wales (Australian Academy of Science 2019; Sheldon et al. 2022; NSW Chief Scientist 2023), compared with only 2 in the 12 years prior. Accordingly, there is no evidence that the Basin Plan settings are associated with a reduction in the number of fish kills. The number of very large events appears to be increasing.

(16) Electrical conductivity of Murray River waters is below target levels and (17) mean annual discharge of 2 × 10⁶ Mg of salt from the Murray Mouth is met.

Environmental outcomes can be enhanced by discharging 2 × 10⁶ Mg of salt per year from the Murray–Darling Basin as a long-term average (S86AA(2)(d)).

Highly saline water inhibits animal health, the production of irrigated crops and pastures and flow-dependent ecosystems (Nielsen et al. 2003). Saline domestic water sources are a health risk for communities.

Ensuring that salinity is managed below target levels is important for agriculture, communities and ecosystems. Basin States monitor electrical conductivity at multiple sites (e.g. at selected river gauges; cf. Murray–Darling Basin Authority 2020e, table 16 therein). The Basin Plan salinity target is to have <800 EC at Morgan for at least 95% of the time, with 95% targets for four other sites as follows: on the Murray River at Lock 6 (580 EC), Murray Bridge (830 EC) and Milang (1000 EC) on the lower Darling at Burtundy (830 EC) (Murray–Darling Basin Authority 2015, 2020e, table 8 therein).

The annual target for salt discharged from the Murray Mouth is 2 × 10⁶ Mg (on the basis of 3-year rolling averages), set by MDBA, on the basis of modelling of 1.95 × 10⁶ Mg year⁻¹ under the Basin Plan 2800 GL year⁻¹ scenario for return of water to the environment (Murray–Darling Basin Authority 2012b, p. 31). Salt interception schemes diverted over 4 × 10⁶ Mg (2009–10 to 2019–20; Murray–Darling Basin Authority 2020e). State and Territory governments report on salinity control, recorded by MDBA in registers of salinity credits and debits.

The salinity (EC) targets for Morgan, Murray Bridge and Lock 6 have been met and maintained since 2010 (Murray–Darling Basin Authority 2020e, p. 10). Salinity at Burtundy was above target for an average of 29% of the time and at Milang for 7% of the time between 2009 and 2020 (Fig. 5a).

Fig. 5.

Indicators, environmental theme. (a) Proportion of days salinity exceeds target values; (b) salt discharge through Murray Mouth; (c) proportion of days water-quality standards for total Kjeldahl nitrogen are exceeded; (d) proportion of days water-quality standards for total phosphorus are exceeded; (e) monthly difference in temperature upstream and downstream of dams with and without thermal pollution control operating; (f) difference in mean summer temperature upstream and downstream of dams with and without thermal pollution control; (g) fish populations, New South Wales and Victorian Basin, the red dashed line indicates no proportional change for Victoria; (h) proportion of time Murray Mouth is open and dredged. Tg, teragram (i.e. 1 × 10⁶ tonnes).

The target for annual discharge of salt from the Murray Mouth shows that the target has not been achieved since 2011–12. The mean annual amount discharged since Basin Plan implementation in 2012–13 is only 0.8 × 10⁶ Mg year⁻¹ (Fig. 5a) and appears to be declining: from 1 × 10⁶ Mg  year⁻¹ in 2012–13 to 0.6 year⁻¹ between 2019 and 2022 (calculated from the line of best fit, Fig. 5b). The reason given for the failure to meet the target was because of ‘relatively low inflows since the Basin Plan was introduced’ and that ‘the overall salt export objective should also be revisited in the context of the Basin’s variable climate’ (Murray–Darling Basin Authority 2017, p. 84). The MDBA has examined the appropriateness of the target and concluded that it was ‘not effective’ and should be ‘improved ahead of the 2026 review of the Basin Plan’ (Murray–Darling Basin Authority 2022b, p. 16). A factor not considered by MDBA regarding the salt export target is the failure to meet flow targets; actual flow at Lock 1, the nearest HIS to the Murray Mouth for which observed v. expected flows were assessed, was only 70% of the expected flow between 2012–13 and 2019–20 (cf. Indicator 12 above). Minimum flow required to keep the Murray Mouth open was estimated to be 730–1090 GL year⁻¹ (Department for Environment and Water 2019, p. 29), yet average annual discharge over the Lower Lakes barrages since 2012–13 was 3925 GL year⁻¹ (cf. Indicator 21 below), mainly owing to very high flows between 2021 and 2023, although flows were below or close to the upper bound of the minimum discharge volume (1090 GL year⁻¹; blue dashed line in Fig. 5h) in 2015–16 and 2017–20.

Reduce nitrogen and phosphorus concentrations in river water towards Australian and New Zealand Environment and Conservation Council standards over time.

The Basin Plan must include a water-quality management plan (S21(1)), including objectives and targets (S25(1)(b)).

High nitrogen and phosphorus concentrations can trigger water-quality threats, including cyanobacterial blooms. State agencies monitor water quality at a range of sites, but information is publicly available only for those sites included in the River Murray Water Quality Monitoring program (Henderson et al. 2013), although data for Morgan were obtained from MDBA on request. Basin Plan targets for nutrient concentrations are listed in Schedule 11 of the Basin Plan, being based on standards for nutrients in waterbodies in south-eastern Australia (Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand 2000).

Weekly or monthly mean concentrations of Kjeldahl nitrogen (N) and total phosphorus (P), measured at locations representing the upper, central and lower Murray River (Kiewa River at Bandiana, Murray River at Swan Hill and Murray River at Morgan respectively) between 2000–01 and 2022–23, were above limits for the following proportions of time: Bandiana: N, 0.19; P, 0.34; Swan Hill: N, 0.2; P, 0.21; Morgan: N, 0.16; P, 0.08 (Fig. 5c and d). Since 2012–13, the proportion of time that N targets were exceeded has improved at Bandiana (0.23–0.13; calculated from line of best fit), Swan Hill (0.3–0.16) and Morgan (0.29–0.8). Exceedance of P targets has barely changed at Bandiana (0.41–0.39), increased slightly at Swan Hill (0.19–0.24) and declined slightly at Morgan (0.08–0.05). Peak exceedances were associated with high unregulated flows during 2010–13, 2016–17 and 2021–23, but, overall, some reductions are evident in N and P concentrations in the Murray since implementation of the Basin Plan.

Cold-water pollution is declining.

The Basin Plan must include a water-quality management plan (S21(1)), including objectives and targets (S25(1)(b)), so as to protect and restore the species that are reliant on the Basin water resources (S21(2)(b)).

Water released from the bottom of large dams is colder than water at the surface or the upstream rivers (Pittock and Hartmann 2011; Lugg and Copeland 2014). The release of cold water lowers the temperature of the river downstream, causing harmful ecological effects, including limiting spawning, survival and growth of fishes if water temperature is consistently below thermal preferences (Koehn et al. 2020). The water-quality target for temperature in the Murray–Darling Basin Plan (Schedule 11) is for the monthly median to be within the range of the 20th and 80th percentile of the natural monthly water temperature.

Installation of thermal-pollution control devices (TPCDs) at dams (Department of Primary Industries 2017) or use of other mitigation methods (Gray et al. 2019; Michie et al. 2020) has been undertaken in New South Wales in a program to restore natural water temperatures and mitigate ecological damage (Department of Primary Industries 2017, 2021). Despite the publication of an action statement on cold-water pollution in Victoria (Department of Sustainability and Environment 2003) and an assessment of the prevalence of the problem over two decades ago (Ryan et al. 2001), we found no evidence for the installation of TPCDs on dams in the Victorian Murray–Darling Basin. Accordingly, the data for this indicator are from the New South Wales Basin only.

Dams with operating strategies for cold-water pollution (Burrendong and Windamere) had a maximum mean monthly downstream water temperature 4.9°C lower than in upstream (2012–13 to 2022–23), compared with 8°C for dams without such strategies (Burrinjuck, Hume, Keepit and Wyangala) (Fig. 5e). Burrinjuck has a TPCD installed but an operating protocol is yet to be fully implemented. Temperatures downstream of the Hume Dam increased to a maximum of 4°C cooler between spring and summer, potentially disrupting spawning of fishes (Sherman et al. 2007). Keepit Dam, considered of low potential for cold-water pollution (Department of Primary Industries 2021), had a maximum difference of only 2.9°C cooler downstream (December), whereas maximum difference at Wyangala Dam was 11.2°C (January).

Mean temperature difference during summer (December–February) between 2012–13 and 2022–23 showed consistently smaller differences at dams with cold-water pollution strategies in operation (grand mean 4°C, range 1.1–7.5°C) than in those without such strategies (grand mean 6°C, range 3.5–8°C; Fig. 5f), although with considerable variation. High values were recorded in 2012–13 and 2016–17. The TPCD installed at Burrendong had been operational only since May 2014 (Gray et al. 2019). There was some indication of a slight decreasing trend over time for both types of dams, but the high inter-annual variation urges caution in making such an inference.

The indicator value is the temperature difference during the summer months, when fishes are spawning, for dams with and without operating TPCDs. If more dams are retrofitted, the gap would be expected to narrow. Our assessment was hindered by incomplete water temperature records at appropriate river gauges (Supplementary Table S1). An improved temperature-monitoring program has been implemented to address this deficit (Department of Primary Industries 2017, p. 12, appendix 1 therein).

Populations of fishes are maintained or improved.

The objective is to protect, restore and provide for the ecological values and ecosystem services of the Murray–Darling Basin (taking into account, in particular, the impact that the taking of water has on watercourses, lakes, wetlands, ground water and water-dependent ecosystems that are part of the Basin water resources, and on associated biodiversity) (S3(d)(ii)).

Populations of native fishes have been under threat from flow alteration associated with irrigation diversions, drought and climate change, poor water quality, barriers to movements, invasive species, habitat alteration, cold-water pollution and commercial and recreational fishing (Koehn et al. 2020). Populations have declined substantially since European settlement (Murray–Darling Basin Authority 2020f). The Native Fish Recovery Strategy includes the objective that threats to native fishes are identified and mitigated (Murray–Darling Basin Authority 2020f). The BWS contains several recommendations relating to river flows and connectivity for improving fish populations (Murray–Darling Basin Authority 2019b).

We used data on relative abundance of two large-bodied species, Murray cod and golden perch, from New South Wales Basin (calendar years 1995–2022) from the New South Wales Department of Primary Industries Freshwater Ecosystem database (Crook et al. 2023) and the annual proportional change in abundance of the same species from the Victorian tributaries of the Murray River between 1999 and 2019 (Yen et al. 2021). The units used to express abundance are incompatible for these two datasets, but the overall trends can be compared.

New South Wales populations showed marked increases over the 28-year period of record, but with considerable variation among years (Fig. 5g). The commercial inland fishery for native finfish in New South Wales was progressively restricted during the 1990s (Reid et al. 1997) and ceased altogether in September 2001 (Department of Primary Industries 2024). Increases in populations are likely to reflect reduced harvesting pressure as well as extensive stocking with hatchery fishes (Crook et al. 2023). Regulation of recreational fishing, use of managed environmental flows, provision of fishways and habitat restoration may also have contributed to increased abundance. Peak abundance of Murray cod in 2008 during the Millennium Drought may be due to the selective advantage of this channel-dwelling species compared with the floodplain-spawning golden perch, which showed a peak in abundance 2 years after the drought broke. Subsequent fluctuations of Murray cod, and declines for golden perch, may be due in part to the six very large fish kills (>100,000 individuals) in the New South Wales Basin during the period 2012–23 (cf. Indicator 15 above).

The increases in populations of Murray cod and golden perch in the New South Wales Basin were generally similar to population trends in the Victorian Murray tributaries, although the increase in golden perch populations commenced earlier than for Murray cod (2007–08 v. 2010–11) (Fig. 5g). High spring discharge and a reduction in the number of low-discharge days relative to the long-term average probably influenced the increase in abundance of the two species (Yen et al. 2021).

The Murray Mouth is open >95% of the time without dredging.

Environmental outcomes can be enhanced by ensuring that the mouth of the Murray River is open without the need for dredging in at least 95% of years, with flows every year through the Murray Mouth barrages (S86AA(2)(c)).

The target for keeping the Murray Mouth open without dredging is mandated in the Water Act and the Basin Plan (Schedule 5, paragraph 2(c)).

The target was not met (Fig. 5h). The minimum annual flow required to keep the Murray Mouth open has been estimated by the MDBA to be 730–1090 GL year⁻¹. The 2020 Basin Plan evaluation concluded the following: ‘it appears that under the drying climate the target for the Murray Mouth opening is unachievable’ (Murray–Darling Basin Authority 2020a, p. 53).

Improved water security for regional towns with historically insecure water supply, whereby (22) there are fewer days of water restrictions and (23) the number of drinking-water quality incidents (e.g. boil water alerts) is declining.

The objective is to improve water security for all uses of Basin water resources (S3e).

Many communities have experienced reduced water security and water quality because of decreased flows and low volumes in public dams during recent droughts (Barbour et al. 2020). There are little in the way of publicly available data on town water security and quality Basin-wide, but number of days per year in which restrictions on domestic water use were in place has been available for LGAs in the New South Wales Basin, as were data on the annual number of water-quality incidents (boil water notices) by LGA (Department of Planning, Industry and Environment 2024).

The mean number of days with water restrictions for 10 LGAs that had water restrictions in place for some of the time between 2013–14 to 2019–20 increased sharply in 2018–19 (194 days) and 2020–21 (232 days) from a relatively stable period between 2013–14 and 2017–18 (mean 62 days) (Fig. 6a). The annual total number of water-quality incidents rose sharply from two in 2012–13 to 13 in 2022–23, with peaks in 2016–17 (9 incidents) and 2019–20 (15 incidents).

Fig. 6.

Indicators, social theme. (a) Town water security, New South Wales Basin: number of days of water restrictions (bars) and number of drinking-water quality incidents (line); (b) number of water-quality threat events (blue–green algal blooms, low dissolved oxygen, hypoxic blackwater and high turbidity).

Water-quality threats are declining in frequency.

Include water-quality objectives and targets for Basin water resources (25(1)(b)). Critical human water needs are the highest-priority water use for communities who are dependent on Basin water resources (S86A(1)(a)).

This indicator overlaps with indicators from other themes. Cyanobacterial (blue–green algal) blooms, hypoxic blackwater events, high turbidity owing to high flows or debris from bushfires have detrimental cultural, environmental and social impacts. Low dissolved oxygen has been responsible for major fish deaths at Menindee Lakes, associated with cyanobacterial blooms (Australian Academy of Science 2019) and high degradable biomass (NSW Chief Scientist 2023). The ecological condition of rivers and wetlands is inherently connected with the health, livelihoods and wellbeing of local communities and Indigenous peoples (Bates et al. 2023). Poor water quality restricts use of water for irrigation, stock and domestic purposes, recreational use of waterways and tourism, and negatively affects personal, community and economic wellbeing. Water-quality threats are likely to intensify in frequency and severity under climate change (Baldwin 2021).

The data were derived from water-quality threat maps published several times per year by MDBA, commencing in October 2020 (e.g. Murray–Darling Basin Authority 2023c).

Some 464 water-quality threat events were identified between October 2020 and December 2023, of which 25% were due to cyanobacterial blooms, 15% to low dissolved oxygen, 9% to hypoxic blackwater and 2% to high turbidity. The mean number of threats per season showed highest numbers during summer and autumn, but no apparent trend over the relatively short period of record (Fig. 6b). The high number of threats in the summer of 2022–23 was due to increased risks of low dissolved oxygen and blackwater event following high flows and flooding during that year.

Target for Indicator 25 is to ensure that SDLs for surface-water resource units are met and Indicator 26 to ensure that the adjusted cumulative balance for each resource unit is stable or increasing, i.e. trending away from the SDL.

The objective is to ensure the return to environmentally sustainable levels of extraction for water resources that are overallocated or overused (S3(d)(i).

The MDBA determines the SDL for each SDL resource unit (or catchment). SDLs are the long-term average volume of water that can be diverted from each SDL resource unit for consumptive use and represent the level of take that is considered economically, socially and environmentally acceptable (Murray–Darling Basin Authority 2011, pp. 2–3). This definition does not equate with the volume required to sustain healthy, flow-dependent ecosystems. This so-called ‘triple-bottom-line’ definition is not legislated in the Water Act, and is considered unlawful (Walker 2019, p. 53). SDLs are part of the legislative mechanism for reducing diversions in the Basin, so it is important to determine whether they are being complied with. We used the adjusted cumulative balance, expressed as a proportion of the long-term SDL, for each surface-water SDL resource unit for 2019–20, 2020–21 and 2021–22, to indicate whether take was increasing or decreasing and moving towards or away from the SDL. Accounting for New South Wales SDL resource units is subject to change on the basis of accreditation of methods used to estimate annual permitted and annual actual take. New South Wales water-resource plans for surface SDL resource units were accredited between March and June 2024, but those for Gwydir and Namoi remained outstanding at the time of writing (August 2024).

SDL compliance reports and registers of take for 2019–20, 2020–21 and 2021–22 indicate that water take has been below the SDL compliance trigger in all SDL resource units, except for Barwon–Darling, in all 3 years and Gwydir in 2021–22 (Fig. 7a). For the Barwon–Darling, exceedance is apparently due to ‘modelling limitations and incomplete environmental water recovery’ and in the Gwydir due to ‘the result of the growth in take by floodplain harvesting’ (Murray–Darling Basin Authority 2023d, pp. 5–7).

Fig. 7.

Indicators, compliance theme. (a) SDL compliance; (b) adjusted cumulative SDL balance; (c) breaches of water laws: number of prosecutions, penalty notices, directions and enforceable undertakings issued.

In the 2021–22 SDL compliance statement, the Inspector-General of Water Compliance stated the following: ‘The situation in New South Wales is deeply concerning, particularly as there is an increasing number of areas on the interim SDL accounts pointing to an SDL excess beyond the SDL compliance threshold, specifically, the Barwon–Darling watercourse by 40%, Gwydir surface water by 21% and the Murrumbidgee is trending toward the SDL compliance threshold at 18% SDL exceedance’ (Inspector General of Water Compliance p. 12). On the basis of this concern, we consider that Indicator 25 has not been met.

Exceedance of the SDL compliance trigger does not provide an indication of trajectories of change in take relative to the SDL. We found that the adjusted cumulative balance of annual permitted take minus annual actual take was negative (i.e. in debit) in 2019–20 in one surface-water SDL resource unit (Barwon–Darling), in five units in 2020–21 (Barwon–Darling, Gwydir, New South Wales Border Rivers, Murrumbidgee and South Australian Murray) and in three units in 2021–22 (Barwon–Darling, Gwydir and Murrumbidgee) (Fig. 7b). The mean trajectory of change in adjusted cumulative balance over the 3-year period declined in eight surface-water SDL resource units (Barwon–Darling, Gwydir, New South Wales Border Rivers, Murrumbidgee, Loddon, Victorian Murray, Eastern Mount Lofty Ranges and South Australian Murray), was stable in three (change was within 10% of the mean; Intersecting Streams, Namoi and Broken) and increased in the remaining 18 units. The number of SDL resource units in which annual actual take was greater than 90% of annual permitted take doubled from 6 in 2019–20 to 12 in 2020–21 and 2021–22. We consider that Indicator 26 has not been met.

A shortcoming of the SDL accounting framework is that surface-water SDLs can be increased over time when estimates of baseline diversion limits (BDLs; the volume of water being used prior the Basin Plan) are updated and increased (Murray–Darling Basin Authority 2023d, p. 7). Evidence for justifying changes in BDLs has generally been poor, rarely adequately articulated and always involve increases. The BDL has been adjusted upward for the Murrumbidgee, with consequent increases in the SDL. Accordingly, the trend towards exceedance that we found is likely to be greater and more extensive had it not been for repeated increases in BDLs and SDLs.

Estimated take from floodplain harvesting for SDL resource units in the northern Basin is reported as identical, 448.2 GL year⁻¹, for 2020–21 and 2021–22 (Murray–Darling Basin Authority 2022c, p. 34; Murray–Darling Basin Authority 2023d, p. 37). This volume is likely to be a considerable underestimate of the actual volume harvested (Brown et al. 2022, p. 51). Accordingly, annual actual take relative to annual permitted take in these SDL resource units is likely to be significantly higher than in the SDL compliance reports and registers of take.

Compliance with the Water Act and State law; number of enforcement actions is declining.

This indicator relates to Part 2, management of Basin water resources; Division 3A, offences and civil penalty provisions and, specifically, S73B, taking water when not permitted under State law. The Commonwealth and the Basin States and Territory agreed to the Murray–Darling Basin Compliance Compact in June 2018 to address compliance and integrity of Basin water management (Murray–Darling Basin Authority 2018). The Compact is intended to ensure that water users comply with State, Territory and Commonwealth laws.

In July 2017 the Australian Broadcasting Corporation aired a Four Corners report, ‘Pumped’, into theft by irrigators of taxpayer-funded water for the environment from the Barwon–Darling River. The report prompted an independent investigation into compliance in New South Wales (Matthews 2017a, 2017b), which led to the establishment in 2018 of the Natural Resources Access Regulator (NRAR), responsible for enforcement of water laws in that State. We used data from the NRAR Public Register on enforcement actions (successful prosecutions, penalty notices, directions and enforceable undertakings; New South Wales Natural Resources Access Regulator 2024) as well as data on successful prosecutions in Queensland, Victoria and South Australia (cf. Indicators 1 and 2 in the Supplementary material).

Legal structures, processes and administration of penalties operate quite differently in each jurisdiction and data cannot readily be sourced, aggregated and compared at whole-of-Basin scale (Seidl and Wheeler 2024). The number of enforcement actions in any jurisdiction will vary according to the degree to which enforcement is implemented. In Victoria, alleged compliance breaches increased since 2014–15 because of improved telemetry and metering to detect unlawful water extractions, whereas in New South Wales, alleged breaches tripled following establishment of the NRAR (Seidl and Wheeler 2024).

Despite the limitations of this indicator, some broad comparisons can be made regarding prosecutions. In the Queensland Basin, there was only one prosecution of an offence committed between 2018–19 and 2021–22 (Queensland Government 2024), and only two in South Australia between 2014–15 and 2022–23 (Department for Environment and Water 2024). In Victoria, there were 48 successful prosecutions between 2018–19 and 2022–23 (Department of Energy, Environment and Climate Action 2024) (Fig. 7c). The small number of successful prosecutions in the New South Wales Basin by NRAR between 2018–19 and 2022–23 indicates the low likelihood of detection and prosecution for serious breaches of the Water Management Act 2000 (NSW) (Fig. 7c). Some 17 convictions are listed on the Public Register (New South Wales Natural Resources Access Regulator 2024), plus 3 in the 2022–23 annual progress report (New South Wales Natural Resources Access Regulator 2023), from 40 prosecutions initiated between 2018–19 and 2022–23, with a conviction rate of 50%. In total, 303 penalty notices, directions and enforceable undertakings were issued over the period, with 56% within LGAs in the northern Basin and 44% in the southern Basin. Although these figures appear to indicate a reasonably even geographical distribution between north and south, the number of enforcement actions per irrigation farm is four times higher in the northern Basin; of an estimated 8390 irrigation farms in the Murray–Darling Basin, only 2001 (24%) are in the northern Basin (Australian Bureau of Statistics 2022). No apparent trend is detectable from what is only 5 years-worth of data. It would be incorrect to assume that the low number of enforcement actions indicates a downward trend over the previous 3–4 years. Rather, because 2022–23 was a particularly wet year, it is likely that there was less incentive for irrigators to break the law.

Successful prosecutions in the New South Wales Land and Environment Court tend to carry much higher fines and costs for common offences of unlawful taking of water and unlawful construction of water supply works (mean fine A$173,281, n = 7; mean costs A$162,127, n = 5) than in the New South Wales Local Court system (mean fine A$15,486, n = 7; mean costs A$8125, n = 6). Fines and costs in the Land and Environment Court typically remain lower than the value of the water taken unlawfully, lessening the incentive for compliance. Under the Water Management Act 2000 (NSW), the Minister for Water has the power to cancel the water licence of a person convicted of an offence. We could find no record of any such cancellation. By comparison, fines in Victorian Magistrates courts for unauthorised take of water under the Water Act 1989 (Vic.) are extremely lenient (mean A$1900, n = 27; costs A$ 2985, n = 41). In 58% of proven cases, no conviction was recorded. As for New South Wales, there is no apparent trend in 5 years-worth of data (Fig. 7c). Seidl and Wheeler (2024) found that the probability of detection and prosecution for water theft in New South Wales is very low, resulting in an average penalty value of stealing water well below existing water market prices, thus providing an incentive for non-compliance.

Underpinning the stark differences among jurisdictions in the enforcement of water laws is the mean annual number of successful prosecutions per capita (2018–19 and 2022–23), namely, 0.3 per thousand irrigation businesses in Queensland, 0.4 in South Australia, 2.8 in Victoria and 1.3 in New South Wales. In the report of the Murray–Darling Basin Royal Commission, Walker (2019, p. 67) observed that there is a ‘high degree of inconsistency between Basin States in relation to matters including the range of offence and penalty provisions, and the use of administrative orders. Basin States may wish to give consideration to whether their respective offence and penalty provisions properly reflect community expectations.’