Reports

Previous stratigraphic and dating studies conducted primarily on the Sir Richard Peninsula but also on the adjacent Younghusband Peninsula are reviewed. A tentative evolutionary history and stratigraphic analysis is produced for the Sir Richard Peninsula based on new drilling and coring across the Beacon 19 track. Previous research indicates initial retrogradation of a coastal barrier over a palaeo-lagoon system starting ~7,000 to 8,000 years BP in the NW portion of the barrier. In the Beacon 19 Track area, the barrier is tentatively considered to have formed by spit extension and migrated alongshore over riverine/lagoonal sediments. This was followed by a narrow band of progradation. Several dunefield phases were possibly formed over time. Vegetation and geomorphological changes are also examined, and significant changes have taken place since the 1700’s and in the post-1930’s period. The Peninsula was characterised by significant areas of active mobile blowouts and parabolic dunes in the northwest portion, and transgressive dunefields in the southeast portion in the 1930’s to 1940’s. By 1975 these were considerably less active and by 2025 the entire Peninsula, apart from the very distal spit terminus, is largely vegetated. The vegetation cover has increased by at least 30% between 1945 and 2025. Significant invasion of Thinopyrum junceiforme (sea wheat grass) has taken place since the early 1980’s.

The Coorong, Lower Lakes and Murray Mouth (CLLMM) region covers approximately 142,500 hectares and includes areas of local, national, and international significance. The CLLMM region is highly vulnerable due to the combined pressures of upstream water regulation, inherent ecological sensitivity and climate change. Climate change impacts will be increasingly realised as declining freshwater inflows, increasing ambient temperatures, and rising sea levels pose significant challenges to the region. The region is ecologically, culturally, and socially important which makes the CLLMM a highly valued and dynamic landscape involving diverse stakeholders, including First Nations community and organisations, community and community organisations, researchers, industry, and management agencies. Increasingly, incorporating local knowledge through community consultation is recognised as essential to effective environmental research and management. To support shared contributions that guide the future direction of knowledge generation in the region, we applied an inclusive two-phase horizon scan methodology to engage a diverse range of stakeholders. Phase 1 involved a survey (N=639) capturing perspectives on key challenges, opportunities, and novel unasked questions. Phase 2 consisted of a stakeholder workshop to identify priority issues through deeper analysis of the survey data. Results from the survey highlighted a wide range of themes, reflecting the region’s complexity. The outcomes of the horizon scan readily synthesise into five main pillars: Water Security and System Management, Climate Change and Future Extremes, Ecosystems, Biodiversity, and Fauna, Environmental Protection, Rehabilitation, and Restoration, and Human Systems, Governance, and Community Engagement. The pillars reflect the top five challenges, opportunities and novel unasked questions of the horizon scan. The identified pillars direct research that will enhance predictive understanding through forecasting and scenario planning. Additionally, they improve the evidence-base for biodiversity conservation and restoration actions. Further, the pillars support better governance and collaboration, strengthened education, science communication, and community-based involvement. A focus on this research direction will generate knowledge to support integrative and adaptive management under uncertainty that can translate into robust decision-making and action to shape the future of the CLLMM region.

The blue and teal carbon project is a flagship research initiative within the Coorong Lower Lakes and Murray Mouth Research Centre’s climate mitigation theme. This report documents the findings of Task 1, Objective 1.1 of the project, which involved a stocktake of existing knowledge, available data, and maps concerning the blue and teal carbon ecosystems in the CLLMM region. The stocktake also investigated the environmental drivers and threats impacting blue and teal carbon ecosystems in the region, and identified key knowledge gaps. The materials discovered through the stocktake, along with the resulting metadatabase and geospatial database, will support the remaining tasks being undertaken as part of the blue and teal carbon project. Blue carbon ecosystems (e.g., saltmarsh and supratidal forest) and teal carbon ecosystems (e.g., freshwater wetlands) are crucial for mitigating climate change by sequestering and storing carbon. These ecosystems also provide extensive co-benefits, including supporting biodiversity and fisheries, protecting shorelines, and maintaining cultural significance for First Nations communities. The broader aims of the blue and teal carbon project are to establish the carbon abatement potential of these ecosystems in the CLLMM region, which include saltmarsh, supratidal forests, coastal sedges, and reed beds. The stocktake compiled and assessed existing knowledge and spatial data relevant to the location of blue and teal carbon ecosystems, above or below-ground carbon stocks, sequestration rates, or ecosystem restoration within the CLLMM region. Data was sourced through expert elicitation, Google searches for grey literature, and a Web of Science search. The stocktake discovered a total of 83 documents and 43 spatial data files, which were examined and summarised. The information was entered into a metadatabase (Excel file), and the spatial data was compiled into a geodatabase in ArcGIS Pro. The majority of information sources originated from state government or council reports, with relatively few sources derived from published scientific literature. The most consistent and long-term data available is from The Living Murray initiative's Lower Lakes Vegetation condition monitoring program (2008-2023 included here, but monitoring is ongoing as part of TLM program), which is focused on aquatic vegetation (i.e. teal carbon ecosystem) monitoring. The stocktake identified 55 different vegetation habitats, grouped into 18 categories, including Reedbeds, Saltmarsh/Samphire shrublands, and various types of Wetlands. Though there is little existing data on soil organic carbon, what exists suggests that the highest concentrations occur within the surface sediments of the Coorong, and that poorly flushed parts of the system, such as the Coorong South Lagoon and Lake Albert, tend to have high carbon accumulation. There is no existing information on teal carbon stocks. The project identified 42 threats, summarised into 29 classes, with the most commonly mentioned threats to blue and teal carbon ecosystems being altered hydrological regimes, salinity changes, and invasive species. The geodatabase contains 43 geospatial datasets, with the majority (23 layers) focused on soil properties that may influence carbon storage. Despite extensive research conducted in the CLLMM region, particularly after the Millennium Drought, no targeted blue or teal carbon studies have yet been undertaken. Key data deficiencies that prevent reliable carbon stock estimation include significant spatial gaps in soil organic carbon sampling coverage across the region and a lack of above-ground biomass data for many species. The absence of targeted carbon studies and specific above-ground biomass data means that ecosystem-based carbon stock estimates cannot currently be made. To generate robust, fine-scale, region-wide estimates of future carbon opportunities, updated vegetation mapping and future projections of vegetation extent under varying conditions (e.g., sea-level rise and changing hydrology) are necessary. The new targeted data collection planned by this project is intended to address some of these gaps and lay the foundations for realising blue and teal carbon wetland restoration in the CLLMM region.

This report provides the first regional assessment of aquatic methane (CH4) emissions across the Coorong, Lower Lakes, and Murray Mouth (CLLMM) region. It combines direct emissions measurement of CH4 (a greenhouse gas with ~80× the warming effect of CO2) with microbial community analysis, and habitat-informed climate-scenario estimates, to examine how CH4 varies across freshwater, brackish, saline, and hypersaline environments in one of Australia’s most important and climate-exposed wetland systems. The report found that CH4 emissions across the CLLMM could not be explained by the typical assumption that increasing salinity suppresses CH4. Ebullition (bubble-mediated emissions) was confined to low-salinity waters, but diffusive emissions persisted across the entire salinity gradient, including in hypersaline habitat. Hydrological and organic-matter-related variables (including tide height, dissolved organic matter, sediment organic matter, dissolved oxygen, and barrage flows) were each more informative controls on emissions than salinity alone. These findings indicate that CH4 emissions reflect interacting effects of organic-matter availability, oxygen conditions, hydrological connectivity, water residence time, and physical transport. Methane-cycling microbial communities also changed systematically across the CLLMM’s salinity gradient. Aerobic methane oxidisers (CH4-consumers) were relatively enriched in fresher and moderately saline habitats, whereas methylotrophic methanogens (CH4-producers) were more strongly associated with saline and hypersaline habitat. This suggests that salinity-related habitat change may alter the balance between methane production and consumption, potentially allowing diffusive emissions to persist even where CH4 bubbles are suppressed. Habitat-informed climate-scenario estimates suggest that Coorong lagoon CH4 emissions may increase under hotter, drier future conditions, while greater marine influence and sea-level rise may produce only small reductions in emissions. Salinisation should therefore not be considered a reliable CH4-mitigation pathway, particularly given its substantial ecological costs. The results also show that hydrological interventions may produce both benefits and trade-offs for methane-cycling. Freshening, flushing, and improved oxygenation may reduce CH4 accumulation and support oxidation, whereas inflows rich in nutrients or labile organic carbon may stimulate CH4 production where residence times are long. These findings do not support reducing or withholding environmental flows. Rather, CH4 outcomes should be considered alongside ecological, cultural, biodiversity, water-quality, and habitat objectives, with particular attention to flow timing, source-water quality, connectivity, oxygenation, and net flushing. Overall, the study establishes a regional baseline for aquatic CH4 emissions and provides a process-based foundation for incorporating CH4 into climate adaptation, environmental-water management, and future carbon accounting in the CLLMM and comparable flow-managed coastal systems.

Historical foredune and beach profile data for the Victor Harbor to Kingston region collected by the Department of Environment and Water, together with LiDAR datasets from 2008 and 2018, were analysed to assess the magnitude of topographic change and to investigate potential shoreline responses to sea level rise. The analysis shows that beach–dune systems in the region exhibit high spatial and temporal variability, with profiles indicating accretion, erosion, or relative stability depending on location. While a small number of sites, such as Victor Harbor and Cape Jaffa, show persistent erosional behaviour and very few (e.g. Goolwa) display accretion, no consistent long-term erosion trend attributable to sea level rise is evident over more than 40 years of observations. Instead, short- to medium-term fluctuations driven by storm-induced erosion and subsequent accretion during calmer wave conditions appear to dominate the record, masking any potential sea level rise signal.

The Coorong, Lower Lakes, and Murray Mouth (CLLMM) region, a nationally and internationally significant wetland system, faces an unprecedented climate challenge that threatens ecosystem services and regional prosperity. Reflecting the magnitude of values at risk, previous valuation studies indicate that the community’s willingness to pay for environmental improvements in the region is substantial, with estimates of approximately $13 billion, including $5.8 billion specifically for the Coorong. Prolonged droughts reduce freshwater inflows, elevate salinity, and drive habitat losses that affect water supply, fisheries, agriculture, and tourism. In the CLLMM, these impacts can be mediated or exacerbated not only by climatic conditions but also by Murray-Darling Basin water governance, including upstream extraction, allocation settings, and environmental water policy. The Millennium Drought highlighted this vulnerability, resulting in widespread and prolonged ecological damage and significant impacts on local communities and businesses. These impacts triggered more than $810 million in emergency spending, including $208.6 million for water infrastructure, $174.8 million to compensate for tourism losses, and $163.1 million for habitat restoration. With ongoing climate change, rising sea levels threaten to transform the region from a predominantly freshwater system into a more marine‑dominated environment, fundamentally altering the ecosystem services it can provide and its economic baseline. This report describes two parallel and complementary engagements with the community in the CLLMM, exploring perspectives on climate adaptation. One engagement used an economic approach, costing impacts, actions, and outcomes. The other used a creative arts approach, co‑developing a community‑led short animation that articulated community concerns and aspirations for the CLLMM in the face of the climate crisis. These engagements are reviewed through a climate‑readiness lens, with themes reflecting specific climate adaptation imperatives. The economic analysis drew on ecosystem valuation from previous studies, agricultural opportunity costs derived from entitlement markets and hydro‑economic literature, and recent experiences of community adaptation gathered through a new targeted survey. A synthesis of Murray-Darling Basin hydro‑economic, econometric, and computable general equilibrium studies indicates that reallocating water to the environment may, in some scenarios and under specific assumptions, deliver aggregate social benefits, grounded in stated community values for ecological improvement, that exceed estimated agricultural opportunity costs. High‑security water entitlements provide one indicator of water value and have recently traded in the range of $7,000–$9,000 per megalitre, with annualised equivalents of approximately $550–$750 per megalitre. These entitlement values are not a universal metric across all sectors or jurisdictions, given differences in entitlement classes, reliance on annual trade, and regional production systems. Basin‑scale evidence further suggests that adaptation across the Murray-Darling Basin can buffer impacts on agricultural profit, with modelling indicating that a 35% reduction in water availability may be associated with an estimated 18% decline in regional production value. However, these aggregate results can mask severe localised transitional costs, including lost production, lost income, and adjustment pressures for individual enterprises and communities. Despite the economic imperative to adapt, local water‑dependent enterprises encounter barriers to adjustment and adaptation. Qualitative evidence from open‑ended survey responses by active CLLMM businesses indicates that many organisations have adopted flexible booking practices, new water sources, and water‑efficiency technologies. However, escalating capital and insurance costs, information gaps, and complex regulatory environments can keep operators in reactive crisis management rather than enabling proactive, strategic adaptation. The community‑led short animation was a key activity and output of the project. It provided a creative process through which participants identified and expressed local priorities, lived experiences of environmental change, and aspirations for climate adaptation. The animation complemented the economic engagement by providing a different form of evidence about what climate change means for the community and how adaptation is understood at a local level. Together, the economic and creative arts engagements effectively involved community members in exploring current and emerging priorities for building resilience to climate change in the CLLMM. Across both approaches, participants consistently emphasised a shift from reactive emergency mitigation toward more proactive, coordinated, and system‑level responses. Key themes raised through the two activities included securing baseline ecological flows to reduce reliance on emergency responses; improving access to clear, plain‑English information to support climate‑smart decision‑making; providing financial assistance, such as small grants and low‑interest loans, to accelerate adaptation; and adopting coordinated, cross‑sectoral planning approaches that prioritise collective action. The economic evidence and the engagement activities concluded that many local community members view downstream environmental health as closely linked to regional prosperity, rather than as a competing objective. While there is substantial local‑scale activity to respond to climate events, there was a strong and consistent sentiment that the scale of change required to address climate impacts exceeds the capacity of local communities alone, and that large policy and institutional responses are required.

This report summarises the field investigation and borehole drilling activities conducted on the Sir Richard Peninsula, South Australia, during 7–14 October 2025. This work is a part of the fourth milestone (“Groundwater monitoring”) of the project “Coastal beach-dune dynamics and management under future climate change”. A team from Flinders University, led by Professor Patrick Hesp, visited the proposed drilling sites and performed an extensive field investigation. The activities include drilling and coring of boreholes, hand-augered holes, installing of wells, deploying water level loggers, undertaking primary water quality measurements and high resolution RTK measurements of borehole positions. These investigations will be followed by subsequent site visits, each with a duration of 1 to 3 days, to monitor groundwater levels and assess water-quality data into the foreseeable future.

Revegetation, human assisted regeneration and care of remnant vegetation have been occurring for many decades, and are part of the social fabric of the Coorong, Lower Lakes and Murray Mouth (CLLMM) region. Recent ecological surveys of remnant ecosystems and areas that were revegetated fifteen years ago have demonstrated multiple benefits of revegetation at the landscape scale for plants, birds and overall ecological health. Surveys of landholders and practitioners involved in revegetation have yielded additional information on revegetation success factors and key considerations for adapting revegetation techniques to a changing climate. The results of these surveys were presented at a workshop in March 2026 that brought together First Nations, scientists, local practitioners and community members. They distilled the key findings of the surveys, reviewed draft revegetation principles and considered what an adaptive management framework for future revegetation and human-assisted regeneration might need in terms of guiding principles, a community of practitioners, funding continuity and monitoring for evaluation. Key reflections of the workshop included the importance of: · Understanding your site in terms of regional context and changing climatic conditions, · Working with neighbours and other communities to connect vegetation across the landscape, · Incorporating cultural knowledge, including cultural burning, into best-practice to achieve the best biodiversity and ecological function outcomes, · Adaptive planning that uses different species and/or techniques over time, · Long-term funding to enable planting when and where conditions are suitable, and · Monitoring and sharing results of success and failures to continuously improve our practices. This report is a snapshot of practitioner and researcher responses to the results of the surveys and the resultant discussions around principles for future revegetation in the CLLMM region. The next steps needed to develop a regional revegetation framework are outlined and focus on development of documented principles and practices, formalisation of demonstration sites and deeper examination of the plant species-specific climate projections to ensure the most appropriate species are being planted. This workshop was a significant step in the pathway towards a co-designed future revegetation framework that is tailored for the CLLMM region and based on decades of experience and the latest research.

On the Younghusband Peninsula, soaks are open waterholes with salinity low enough to support human or animal consumption. This project focused on understanding where we should expect to find low salinity soaks along the Younghusband Peninsula, developing a method to identify where they are from satellite imagery, and carrying out targeted fieldwork at selected soaks to begin to better understand how they function. The specific outcomes from undertaking this research in alignment with land management requirements are: Remote sensing mapping and ground-truthing of the freshwater soaks across Younghusband Peninsula, Improving cultural connection with research and management of freshwater soaks through on-country activities with FN groups, DEW, and the Landscape Board, Understanding of hydrological dynamics of key freshwater soaks, Summary of freshwater soak vulnerability to climate change, and A graphical representation (infographic) of soak hydrology for broader dissemination and synthesis report suitable for publication in peer-reviewed journal. Using maps from the early 20th century, we digitised documented soaks along the Younghusband Peninsula. Visual inspection of satellite imagery from the past 15 years indicates that few of these soaks likely exist in the modern landscape. Using multi-spectral imagery, we sought to precisely locate currently existing soaks, which are typically small and often easy to confuse with shadows on satellite imagery. This process was refined following ground-truthing of initially identified, potential soaks. The salinity of the identified soaks ranged from brackish (1,600–16,000 mS/cm) to salty (>16,000 mS/cm; seawater ~55,000 mS/cm). Most current soaks are located on the Coorong side of the peninsula, at low elevations relative to Coorong water level elevation. The presence of Phragmites australis (common reed), a fresh-brackish plant species, was found to be a good indicator of low salinity soak occurrence. While the reed analysis was developed as a proof of concept during the current project, further development of this methodology provides an opportunity to locate and monitor soak distribution throughout the longitudinally expansive and logistically challenging extent of the Younghusband Peninsula. The soaks exist in a relatively simple but delicate hydrological cycle. Rainfall on Younghusband Peninsula is approximately 500 mm per year. This rainfall percolates through, and gets stored in, the dunes; how much water flows into a given dune each day is dependent on how big the dune is that replenishes the soak, and how much evaporation and plant water use is occurring (e.g., what season it is). This evaporation and transpiration accounts for most of the annual rainfall; evapotranspiration rates were calculated to match or exceed average annual rainfall for the study soaks. Any “leftover” rainfall flows through the soak and into the Coorong. The soaks exist in an environment that has been heavily impacted by post-colonial development due to vegetation impacts from feral animal species and introduced weeds; however, it has little infrastructure or human water withdrawal. Nevertheless, both Ngarrindjeri and Burrandies communities expressed the wish that vehicle traffic along the Younghusband Peninsula be restricted to protect the delicate dune environment, and this may be a good way to limit future disturbances to the soaks. Climate change is already affecting South Australia through alterations in the timing and intensity of storm events. While analyses of long-term rainfall records in the Coorong do not show statistically significant changes at an annual scale, previous studies indicate that dune vegetation is sensitive to interannual shifts in rainfall and to sudden storm events. Now that the locations of current soaks in the northern section of the Coorong have been identified, next steps are to: Improve the mapping method moving from a mixture of manual and machine learning to more automated deep learning AI methods, Extend the mapping to all sections of the Coorong, Understand how the hydrology and ecology of the soaks interact, Continue longer-term water level and water quality monitoring to understand how they change through time both seasonally and annually, and Monitor the impact of sea level rise and periodic flooding events.

In the report we review the various methods to estimate the response of sandy beaches to sea level rise. We then conduct an analysis of future sea level rise on representative modal beaches in the Coorong, Lower Lakes and Murray Mouth (CLLMM) region and beyond. These beaches are Waitpinga Beach, a modal intermediate beach dominated by large rip systems, Goolwa Beach, a modal longshore-bar-trough beach dominated by a wide multi-bar, largely dissipative system, and Horseshoe Bay, a modal reflective beach dominated by a narrow surfzone and steep, narrow beach. In the worst-case scenarios, the results show that Goolwa Beach could retreat by up to 96 m, Waitpinga Beach by up to 85 m, and Horseshoe Bay by up to 30 m by 2100. The results indicate that there is an order of magnitude difference in degree of shoreline recession between Robin’s Rule and the Bruun model, with the XBeach modelling sitting approximately in the middle of the two extremes. This indicates the difficulty that exists when attempting to determine a line or zone of setback for future development, and also a line or zone of coastal land that could be impacted by a future sea level rise of 1 m by ~2100.

The Coorong, a Ramsar‑listed coastal lagoon of ecological, cultural and economic importance, is undergoing change as a result of climate‑driven shifts in freshwater inflows, temperature, sea level and extreme weather events. Phytoplankton are important bioindicators of these changes due to their rapid turnover rates and are foundational to the functioning of the Coorong’s food web. This report synthesises 16 years of phytoplankton and environmental data (2007–2023) to examine how climate‑mediated flow variation shapes the Coorong’s microalgal communities, supporting a predictive framework to understand future environmental trajectories and phytoplankton responses under climate change. Through an integrated analysis combining long‑term monitoring, statistical modelling, machine learning and habitat‑based ecological characterisation, this study delivers a comprehensive assessment of how climate variability affects phytoplankton community dynamics and harmful algal bloom (HAB) risk in the Coorong. A full range of hydrological conditions were captured in this study, from the Millennium Drought (extreme dry) to a large River Murray flood (extreme wet). These contrasting climate periods produced dramatic shifts in barrage flows, water levels, salinity, temperature and nutrient concentrations, which in turn restructured phytoplankton abundance, diversity and community composition. The Coorong’s environmental conditions significantly differed across flow periods. During the dry and extreme dry periods, freshwater inputs declined lowering water levels and increasing salinity to levels exceeding 190 PSU in the South Lagoon. Wet and extreme wet periods reversed these conditions, producing rapid freshening, increased hydrological flushing that removed nutrients and salt from the system and increased water levels, that further diluted salt and nutrients. Phytoplankton communities demonstrated strong, statistically significant responses to these changes, with significantly lower abundances under both extreme wet and dry conditions, but higher richness under wetter conditions and higher evenness during drought. Across an ensemble of analytical approaches, including multiple correlation analysis, ordination, distance‑based linear modelling and generalised additive models, salinity and water level emerged as the dominant drivers of phytoplankton abundance, diversity and community structure. These variables are tightly coupled and both controlled by freshwater inflows. Higher water levels and lower salinities promoted Cyanobacteria and Chlorophyte abundance. Conversely, increasing salinity altered species composition across all taxonomic groups, with increasing dominance of Diatoms and Dinoflagellates. Other important secondary drivers of phytoplankton community dynamics included temperature, distance from the Murray Mouth (representing hydrodynamic and salinity gradients) and nutrient concentrations (particularly NOX and PO₄), although their influence was weaker and more variable across statistical models. To characterise responses to hydrological change, phytoplankton communities were classified into six salinity‑derived habitats spanning freshwater (H1) to extreme hypersalinity (H6). Community composition, abundance and diversity significantly differed between habitats. Key patterns included: · Freshwater and low‑brackish habitats (H1–H2) had the highest richness but lower evenness, driven by Cyanobacteria and Chlorophyte dominance, particularly under wetter periods. However, H2 demonstrated the most equal representation of all taxonomic groups, · High‑brackish to marine conditions (H3) displayed moderate diversity and high evenness and were dominated by Diatoms and Dinoflagellates, · Hypersaline habitats (H4–H5) supported very high picophytoplankton abundance and Diatom dominance, and · Extreme hypersaline conditions (H6) had the lowest richness but highest evenness and were dominated by Diatoms and Dinoflagellates with limited representation from other groups. These habitat‑based community descriptions provide a strong foundation for forecasting future ecological states using hydrological model outputs that predict salinity. A total of 106 potentially harmful algal bloom forming species (HABs) were identified across the study period, including 82 toxin-producing species. Most HABs occurred in low‑salinity habitats (H1–H2) in the North Lagoon, especially during the post‑drought wet period in 2012. Although HAB events were generally infrequent and short‑lived, two notable bloom events occurred: 1. A Cyanobacteria bloom from January to June 2012 in the North Lagoon consisting of multiple species, though dominated by Aphanocapsa sp. and exceeding medium to high risk thresholds, and 2. A Dinoflagellate bloom during the extreme dry period, where Alexandrium spp. reached action‑level abundances across all sampling sites. Despite overall low HAB prevalence, the action-level abundances of Alexandrium species capable of producing toxins warrants ongoing vigilance, particularly under future climate scenarios that may increase the frequency and intensity of low‑flow, high‑salinity conditions or sharp nutrient pulses. An extreme gradient boost (XGBoost) machine learning model was used to predict HAB species abundance from environmental variables. The model performed strongly across much of the dataset, explaining 77% of the variation in HAB abundance, with a normalised root mean square error of 5.7%. The model’s predictive performance declined under very high observed HAB abundances (> 1.0 x 108 cells L-1), suggesting that extreme bloom events are harder to predict due to their rarity and non‑linear ecological triggers. SHapley Additive exPlanations (SHAP) values were used to quantify each environmental predictor variable’s contribution to the variation explained by the model. Salinity was revealed as the most influential predictor, aligning with results from other analyses, with higher salinity strongly suppressing HAB formation. pH was the second most important variable, exerting mostly negative effects on HAB abundance. NOX and distance from the Murray Mouth were moderate predictors linked to nutrient supply and spatial-chemical gradients, with NOX exerting a positive influence on HAB formation, while HAB abundance declined with distance from Murray Mouth. The XGBoost model was developed as a tool to better understand the broad context of harmful algal proliferation in the Coorong, but not to be used as a primary method for risk assessment. However, this model has the potential to be further developed to predict the abundances of individual species that have defined cell count risk thresholds and pose a significant threat to the Coorong. Climate change projections for the Coorong predict continued declines in freshwater inflows, increased air and water temperatures, higher variability in droughts and floods and increasing marine influence with sea-level rise. Under future conditions, the Coorong is expected to experience more frequent and prolonged hypersalinity, particularly in the South Lagoon, resulting in a decline in low‑salinity habitats (H1–H2) that currently support higher species richness and Cyanobacteria and Chlorophyte diversity and abundance. Picophytoplankton, Diatoms and Dinoflagellate abundances are predicted to increase, with decreasing species richness as hypersaline habitats (H4–H6) expand. Reducing species richness and the decline of several taxonomic groups will impact the food web, affecting trophic transfer efficiency. Increasing picophytoplankton dominance will increase turbidity, reducing submerged macrophyte habitat with further consequences for the Coorong’s biodiversity. Increasing frequency and intensity of drought increases the vulnerability of the Coorong to toxin-producing Dinoflagellate HABs. Without sufficient freshwater inflows, the system risks declining to a permanent degraded state that resembles the end of the Millennium Drought (2007–2010). To maintain a healthy phytoplankton community and Coorong ecosystem, the following recommendations are outlined: 1. Maintain and enhance freshwater flows to preserve low‑salinity habitats critical for phytoplankton diversity, HAB mitigation and hydrological flushing to limit picophytoplankton dominance and maintain water clarity, 2. Implement regular and continuing phytoplankton monitoring to track changes in phytoplankton communities across varying flow periods, detect early HAB signals and improve predictive modelling of phytoplankton dynamics. It is recommended that this monitoring includes testing algal-produced toxin levels, 3. Incorporate phytoplankton community responses to coupled hydrodynamic–biogeochemical-ecological models predicting salinity, water level, temperature and nutrient inputs to capture the spatiotemporal dynamics of the Coorong, 4. Refine machine learning models predicting HABs by further refining model hyperparameters and incorporating additional variables and species‑specific predictors to improve bloom detection, particularly under extreme abundances, 5. Investigate species interactions and grazing dynamics, including co‑occurrence networks, which strongly influence phytoplankton community composition, an Strengthen HAB risk frameworks by developing species‑specific thresholds for the Coorong and integrating outputs into management response protocols.

To maintain viable waterbird and shorebird populations in the CLLMM region, it is important to understand how species from different functional groups select habitats for roosting, breeding and foraging, and how management interventions and climate change might impact those habitats. At a larger spatial scale, bird movement data can identify key movement routes and environmental triggers for bird breeding and dispersal, and inform landscape-scale water management decisions under climate change. This project aimed to use GPS telemetry to identify the habitats being selected by key waterbird and shorebird species within the CLLMM region, as well as population connectivity between the CLLMM populations and the broader wetland network (including the contribution from restored wetlands). We focused this research on four species representing different functional groups: (1) the fairy tern/Tuwit (Sternula nereis), a diving, piscivorous, visual forager that is listed as Endangered in South Australia; (2) the sharp-tailed sandpiper/Nemineri (Calidris acuminata), a migratory shorebird that is federally listed as Vulnerable; (3) the chestnut teal/Ngra:ki (Anas castanea), a dabbling herbivorous duck that is relatively sedentary and favours coastal areas; and (4) the royal spoonbill (Platalea regia), a colonial-nesting tactile forager that breeds in the Lower Lakes. From late 2024, we attached GPS tracking devices to a total of 35 waterbirds (11 fairy terns, 2 sharp-tailed sandpipers, 12 chestnut teal and 10 royal spoonbills) and monitored their movements remotely. Although fairy terns bred and regularly roosted on islands within the Coorong South Lagoon, they also regularly travelled long distances (at times > 30 km) to forage in the southern portions of the Coorong North Lagoon. Although most GPS-tagged fairy tern concentrated their activity within the Coorong lagoons, three individuals engaged in long-distance flights to the South East of South Australia region, with two adult birds tagged in 2026 relocating to Lake Bonney SE, a large coastal wetland, in April. These results suggest substantial mixing of fairy terns through South East region, but it is currently unclear whether this represents permanent emigration from the Coorong subpopulation or temporary emigration during the non-breeding season. GPS tracking of two sharp-tailed sandpipers provided important insights into how this migratory species responds to the South East landscape. For several weeks, both birds continued to roost and forage on the restored wetland where they were captured, one at Lake Hawdon and the other at Tolderol Game Reserve. After Lake Hawdon dried, one bird apparently scoped the coastal wetlands down to Lake Bonney SE in search of suitable foraging habitat. The second bird engaged in a c. 300 km exploratory flight from Tolderol Game Reserve, visiting Lake Hawdon North briefly about 4 weeks after this wetland had dried. Together, these results suggest that sharp-tailed sandpiper will use substantial searching behaviour and will likely locate restored wetlands if they are managed to hold water. Chestnut teal, tracked for 15 to 108 days, remained within the Coorong region, with most staying within the main lagoons, and a few also using small inland water bodies. The trapping site at Parnka Point was a hotspot for chestnut teal activity and most individuals did not move far from or regularly returned to this area. The water depths occupied by chestnut teal varied over the diurnal cycle and differed between Coorong locations, with the regions around Parnka Point and Long Point likely used for foraging in water depths of up to 20 cm. Male chestnut teal moved more than females on average and spent substantially more time in water depths suitable for foraging. Water-level management strategies to support chestnut teal could therefore consider these between-sex differences in water depth preferences, and thus suitable foods available at these depths. GPS tracking of juvenile royal spoonbills suggested the CLLMM region provides suitable roosting and foraging habitat early in their development. Key roosting sites identified lay along the fringes of Lake Alexandrina and Lake Albert, particularly on islands created by vegetation consisting of common reeds and bulrushes. Royal spoonbills typically occupied roosting sites during the day and, consistent with their tactile foraging strategy, travelled at dusk to foraging locations up to 30 km away (including sites in the Coorong North Lagoon), before returning to roost again near dawn. Further research is needed to understand whether royal spoonbills within the CLLMM region form a self-sustaining subpopulation and/or whether juveniles recruit into the broader population of south-eastern Australia.

Freshwater mussels are integral to freshwater biodiversity and provide critical ecosystem services. As filter feeders, they are ideal indicators of water quality and ecosystems health. Globally, human modification of aquatic ecosystems threatens freshwater mussel populations. The Coorong, Lower Lakes and Murray Mouth (CLLMM) region is home to the floodplain mussel (Velesunio ambiguus). Freshwater mussels are of cultural importance to First Nations in Australia, including Ngarrindjeri people of the CLLMM region who know freshwater mussel as Lokeri (Loker-engk for two mussels; Lokerar for three or more mussels). Forgotten caches of Lokeri dated to over 8000 years old have been found on Ngarrindjeri Country thereby demonstrating its use as a food source. A greater understanding of Lokeri on Ngarrindjeri Country will provide essential knowledge to address cultural considerations and the ecological management of the species. The citizen science ‘muscles in the mud’ project was sparked by community interest in the contemporary status of Lokeri in the Lower Lakes, where it was very abundant up until the early 2000s but disappeared by 2009 during the Millennium Drought. The first aim of the project was to connect with community to discuss project objectives, identify potential survey sites and increase awareness of Lokeri. The second aim was to conduct a widespread survey of Lake Alexandrina and Lake Albert to identify if and where Lokeri may be present. The third aim of the project was to examine shells of Lokeri collected from the Lower Lakes to compare three approaches to determining age and growth (counting growth increments from the umbo section and from the full cross-section, and counting external growth interruption lines). In November 2024, more than 50 members of the Ngarrindjeri community attended two yarning circles held in Raukkan and at Point Sturt. The lead researcher presented the idea for a Lokeri project that included a survey where community members could be employed to participate in field and laboratory work. The events provided a successful way to inform community of the intention to conduct a survey on Ngarrindjeri Country and the opportunity for people to participate. After the yarns, Community members entered the water, searched for Lokerar and shared knowledge across generations. This work supported both the collection of information and the continuation of cultural practice. Many not having been in the water for years and feeling connected to the water again after being disconnected for so long. Many reporting how their mi:wi (spirit) felt happy again. In December 2024, 34 sites were surveyed in the Lower Lakes in lake-edge habitats known to be preferred by Lokeri. Lokeri presence and abundance at each site was measured using three adjacent 10 ×10 m quadrats in a randomly chosen location that were accompanied by a wider general search up to 50 m from shore and approximately 50 m either side of the quadrats. Numbers of live Lokeri and empty shells were recorded and each live Lokeri was measured for a range of morphometric parameters including length, height and width. A total of 76 live floodplain mussels and two river mussels (Alathyria jacksoni) were recorded from 9 and 1 of the 30 sites in Lake Alexandrina, respectively, but neither species was detected in Lake Albert. The quantitative surveying revealed low abundances across all sites, with temporal comparison from surveys conducted 45 years prior at Point Sturt, Lake Alexandrina, indicating the current population is

The Coorong Lower Lakes and Murray Mouth (CLLMM) region supports more than 35 freshwater fish species, including several threatened small-bodied species. Already at risk, small-bodied fishes were severely impacted by the prolonged Millennium Drought (2001–2010) and will be vulnerable to the impacts of environmental change in the region. Some of the most affected fish species were the southern pygmy perch and Yarra pygmy perch. A range of conservation actions have been implemented in the attempt to recover these two species. Among these actions are improved water level management, habitat restoration, alien species control, and translocations to reestablish resilient, connected populations to help secure the long-term survival of the species. Despite 17 years of interventions, both species remain precarious, and key questions remain about how to achieve self‑sustaining populations. To help inform future management of wild and ex situ pygmy perch populations, reintroduction ecology, and the use of water for the environment, this project built on ongoing collaborations and conservation actions, and used 15+ years of data on southern pygmy perch from the CLLMM and Eastern Mount Lofty Ranges (EMLR) regions, modelling, otolith aging, and field experiments to investigate how life history processes (e.g. reproduction) and population demographics are associated with biotic and abiotic variables. Population modelling aimed to determine trends (changes) in southern pygmy perch abundance and recruitment over time (phase 1), and the drivers of these trends (phase 2). Broadly, relative abundance of southern pygmy perch declined in the EMLR and increased in the Lower Lakes over the study period (2007–2025). These findings indicate that the effects of environmental change may be evident in the EMLR population, with fish abundance declining in a continuing drying climate and limited water resource system. Although the Lower Lakes population underwent considerable disturbance during and after the Millennium Drought, the continuing positive trajectory of this population indicates this region may confer resilience to this species in the face of environmental change. In the EMLR, however, the southern pygmy perch population is likely to continue to decline and urgent intervention is required. The second phase of characterising trends in southern pygmy perch populations in the EMLR and Lower Lakes was to determine which environmental variables may be driving the population trends using empirical and model methods. In the EMLR, at the regional scale, modelling of the relative abundance of southern pygmy perch, in relation to potential regional environmental drivers, revealed that the Antarctic Oscillation (AAO), with a time lag of two years, was the most influential driver, with empirical analysis showing a positive AAO drove a decrease in southern pygmy perch and that this relationship strengthened after the Millennium Drought (2009). This result highlights the compromised long-term sustainability of southern pygmy perch, and other small-bodied fish species in the EMLR under current and future changing climates where positive AAO conditions will be more common. In the Lower Lakes, modelling of the relative abundance of southern pygmy perch in relation to regional environmental driver data revealed that mean lake level, with a lag time of six months (previous September), was found to be the most influential driver with a positive relationship, where elevated lake water levels promote increased abundance. As the key driver of relative abundance of southern pygmy perch in the Lower Lakes, the importance of lake water level management, via River Murray flow and barrage operations, is paramount. The optimal conditions to promote recruitment in Lower Lakes southern pygmy perch populations is approximately a 0.8 m Australian Height Datum (AHD) lake water level in September. In both the EMLR and Lower Lakes, the density of aquatic macrophytes present was found to be the most influential site level, local driver of southern pygmy perch relative abundance compared to the other suite of drivers tested. A strong positive relationship was found where increasing macrophyte density drove an increase in southern pygmy perch abundance. These modelling results, of site-scale drivers of southern pygmy perch abundance, led to a field experiment to investigate how aquatic macrophyte density may influence the recruitment and abundance of southern pygmy perch in the Lower Lakes. Although Southern pygmy perch were not recaptured during field experiments, results revealed early-stage juveniles from other native fish species were only found in high vegetation enclosures, and that most adult native fishes were found in high vegetation enclosures. No significant differences were observed between water parameters in high and low vegetation enclosures; thus, water physico-chemistry may not be a contributing factor to variability in fish abundances between treatments. The use of otoliths to gain knowledge about southern pygmy perch can help inform how hydrological manipulations of lake water levels will influence the timing of breeding and the outcomes of recruitment (e.g. growth rates). This study is the first to provide credible age estimates of the Vulnerable MDB population of southern pygmy perch. Daily increments were found to be interpretable in most southern pygmy perch up to ~50 mm total length. Daily age estimated hatch dates were predominantly in spring and summer. September and December were the most frequent months of hatching, closely followed by January. Quantification of hatch dates, estimated from age, are potentially informative for the conservation of southern pygmy perch particularly in the Lower Lakes. For example, knowing the time of spawning can inform the precise timing of environmental watering interventions to maximise reproduction and recruitment. This type of watering intervention is frequently undertaken by the Department of Environment and Water (DEW). The modelling chapter of this report shows the optimal conditions to promote recruitment of Lower Lakes southern pygmy perch populations is approximately 0.8 m AHD lake water level in September. The hatch dates in the current study were most numerous in September, which suggests the optimal conditions revealed by the model are strongly related to spawning and recruitment. Together, these findings provide a clear evidence base to guide future conservation actions for threatened small‑bodied fish in the CLLMM region

Beach-nesting bird species reproduce on beaches and dunes and are vulnerable to threats from both the land and the sea. Their breeding habitat is increasingly threatened by human development and recreational use, invasive and native predators, and sea level rise and extreme weather events (including storm surges) driven by the overarching impacts of climate change. In the Coorong, Lower Lakes and Murray Mouth (CLLMM) region, community consultation identified that vehicle access to beaches and protected areas, and its impact on beach-nesting birds, were a key concern of community groups. Elsewhere, it has been shown that vehicles directly cause the mortality of birds and their eggs and chicks, but also have indirect impacts including habitat destruction and disturbance of nesting birds.In response to these community concerns, this project aimed to investigate nesting success and behaviour of beach-nesting birds along Coorong Ocean Beach, using the pied oystercatcher (Haematopus longirostris) as a model species. To understand the impact of vehicles and predators, we monitored pied oystercatcher nests and the surrounding environment over two breeding seasons, using a combination of observational site visits and installed trail cameras. Simultaneously, we also studied the movement and nesting behaviour of pied oystercatcher using GPS telemetry.Our monitoring revealed a low, steady vehicle traffic along Coorong Ocean Beach during the vehicle closure operating between 24 October and 24 December, reflecting commercial fishing activities. Vehicle traffic peaked at around 18 vehicle passes per day along Ocean Beach in late December and early January, likely reflecting access by additional recreational vehicles after the public driving restrictions were lifted. Although we documented no direct nest mortality due to vehicles, vehicle access likely displaces breeding birds from suitable beach-nesting habitat, particularly near the high-tide line. In our study, most pied oystercatcher nests located beyond the Murray Mouth region were positioned on the foredunes or within the dunes, which could reflect a habitat displacement effect if birds avoid nesting within the beach areas used by vehicles.Over two breeding seasons, we monitored 28 pied oystercatcher nests on Younghusband Peninsula and documented the outcome (successful hatching of at least one chick, or complete failure) in all but 2 cases. Of the 26 nests with known outcomes, 50% (13/26) failed completely. Predation by foxes was the greatest single cause of nest failure, causing 53.8% (7/13) of nest failures, or 87.5% (7/8) of failures in nests with a known failure cause. The daily nest mortality rate increased with nest age, which could reflect predator learning or increased nest detection by predators later in the incubation period (potentially due to the audible calls of embryos from within the eggs as they approach hatching). These results suggest more intensive or coordinated fox control is warranted along the peninsula. We recommend that the feasibility of constructing a fox-proof fence to protect the northern portion ofYounghusband Peninsula from fox incursions should be considered, as this could protect multiple beach-nesting bird species.During the day, GPS-tagged pied oystercatchers were more often at their nests, with nest faithfulness signficantly higher for females than males. Although we did not find strong impacts of temperature on nest attendance during day, at night there was a strong negative relationship between temperature and nest attendance by females, likely reflecting increased incubation at cool temperatures to reduce the risk of embryonic mortality or slowed development. Consistent with these results, female pied oystercatcher also ventured further from their nests (presumably to forage) on warm nights. While nesting, GPS-tagged birds of both sexes used the Ocean Beach and the Coorong North Lagoon, suggesting that maintaining the ecological condition of the Coorong is important for the species, in addition to beach-focused management strategies.

Ocean beaches in the Coorong, Lower Lakes and Murray Mouth (CLLMM) region are ecologically, socially, and economically important, supporting threatened species, significant fisheries, and a wide range of recreational activities. However, they have received comparatively little scientific attention compared to the adjacent freshwater and estuarine systems. As such, their current condition and how they are changing over time remains poorly understood. These beaches face a range of pressures, including overharvesting, human disturbance (e.g. vehicle use), invasive species, pollution, catchment impacts (i.e. altered land and water use), and climate change. Knowledge of these threats is variable as some pressures are widely recognised while others remain underrepresented. Previous research has examined aspects of beach ecology in the region, but this work is fragmented and has not been brought together in a way that can be used easily by managers to inform future management of this area. There is also limited insight into how different stakeholders use and value these environments, and how they perceive the threats affecting them. This project addresses these gaps by integrating existing scientific knowledge with spatial data on beach use and stakeholder perspectives, providing a more complete understanding of ocean beaches in the CLLMM region to support future management and planning.A literature review was undertaken using an ecosystem services framework (the direct and indirect benefits humans receive from nature) to clarify what is currently known about the ecology of ocean beaches in the CLLMM region, including the benefits they provide and the pressures affecting them. The review highlighted that a moderate body of research exists but has largely concentrated on a limited set of ecosystem services, particularly those that are most visible or directly linked to economic and ecological outcomes, such as fisheries (e.g. pipi and mulloway) or maintaining habitats for species (e.g. shorebirds). The most commonly identified threats were overharvesting and human disturbance, especially from activities such as off-road vehicle use. These pressures were often linked to impacts on species and habitats, particularly shorebirds and fishery resources. However, broader ecological processes, including how beaches change over time and their role in protecting the coast from extreme events were rarely considered despite their importance for long-term ecosystem function. Cultural values beyond recreation, including deeper social and cultural connections to beaches, were also underrepresented. In addition, the literature generally considered threats in isolation, with limited attention given to how multiple pressures interact. Together, these findings indicate that while a moderate body of knowledge exists, it does not yet capture the full range of values and processes associated with ocean beaches in the CLLMM region.To complement the literature review, a Public Participation Geographic Information System (PPGIS) was used to capture how people use and value ocean beaches across the CLLMM region, and how they perceive key threats. The results show that beaches are used by both local residents and visitors from outside the region, with use peaking in summer but remaining steady across seasons. A wide range of activities were reported, including walking, swimming, fishing, four-wheel driving, and wildlife observation, with use concentrated around key access points, particularly near Goolwa. Participants generally expressed a preference for balancing recreation with environmental protection, with many supporting management actions such as limiting vehicle access and improving responsible use, either through penalties or incentives. Concern about environmental impacts was widespread, particularly in relation to vehicle use, pollution and human disturbance. Participants expressed moderate concern about climate change, with greater concern related to this issue in the longer-term, rather than the near future. Areas of highest concern closely aligned with the most heavily used locations across the coastline. Overall, the findings highlight the importance of ocean beaches for recreation and wellbeing, while also indicating growing awareness of environmental pressures and support for improved management.To further explore these patterns, semi-structured interviews were conducted with stakeholders to understand how different groups use, value, and perceive threats to ocean beaches in the CLLMM region. Participants consistently emphasised the importance of beaches for recreation, wellbeing, and connection to nature, with activities such as swimming, walking, fishing, and socialising frequently highlighted. A strong and recurring theme was the value of the beaches’ relatively undeveloped and uncrowded character, with many participants describing a sense of solitude, natural beauty, and emotional connection to the landscape. Concerns about the current and future condition of beaches were widespread. Human activities, particularly four-wheel driving, were most frequently identified as a key pressure on the environment. Participants also expressed concern about pollution, coastal erosion, population growth, and the effects of climate change. While climate change was widely recognised as a significant long-term threat, it was often perceived as less immediate or more difficult to address compared to local pressures such as recreational use. Overall, participants highlighted a need for improved management, including greater education, regulation, and shared responsibility for protecting the beach environments. These findings provide important context for understanding how communities experience and interpret environmental change, complementing both the spatial patterns identified through the PPGIS and the trends observed in the literature.Across all three components of this project, the literature review, the PPGIS analysis, and the stakeholder interviews, a consistent pattern emerges: while a moderate amount of information exists on ocean beaches in the CLLMM region, it remains fragmented, varied in focus, and not well connected to how people use and value these environments. This includes not only the literature, but also a wide range of existing monitoring and data collection activities that are currently dispersed across different organisations and programs. The literature emphasises a limited set of ecosystem services and associated threats, the PPGIS highlights that key values and pressures are concentrated near the access points and also the Murray Mouth, and stakeholder interviews provide insight into how these issues are experienced and understood by the community. Together, these findings indicate a need for a more coordinated and integrated approach to manage ocean beaches. Additionally, building visitor understanding of ocean beaches and promoting opportunities to engage in the protection of this important resource is fundamental to long term protection.Better alignment between scientific knowledge, spatial data, and community perspectives can support more informed decision-making. This includes improving the synthesis and accessibility of existing information, as well as strengthening monitoring efforts to capture both environmental change and patterns of human use. The strong community concern around issues such as beach driving, pollution, and environmental degradation also highlights the importance of targeted management responses, including improved education, clearer regulations, and increased compliance where needed.More broadly, the findings highlight the importance of moving beyond single-issue management towards approaches that recognise how multiple pressures interact across the coastline. Supporting shared responsibility for beach stewardship, including opportunities for community involvement in monitoring and management, can also be important for maintaining the ecological and social values of these environments. Together, these actions provide a pathway towards more adaptive and integrated management of ocean beaches in the CLLMM region.

The Coorong, Lower Lakes and Murray Mouth (CLLMM) region, at the terminus of the Murray–Darling Basin, has immense socio-cultural, economic and environmental importance. In the late 1930s, in response to expanding upstream irrigation and declining freshwater flow, the Murray Barrages were constructed to mitigate impacts of increasing saltwater ingress to the freshwater Lower Lakes. Whilst securing the upstream freshwater environment, the barrages, in conjunction with reduced freshwater flow, restrict the physical extent of the remnant downstream estuary and under low–average flows, form an abrupt interface, fragmenting freshwater, estuarine and marine habitats. Climate change is likely to exacerbate these impacts. This study investigated how a more hydrologically transparent operating regime at the Murray Barrages—allowing periods of bidirectional flow during “reverse flows” (when Coorong water levels exceed lake levels)—can improve connectivity and estuarine ecosystem function while managing risks to lake salinity and freshwater-dependent habitats. From June 2025 to March 2026, we undertook operational trials (in collaboration with SA Water and DEW) at Boundary Creek and Ewe Island barrages, intermittently keeping one barrage bay open to enable reverse flows. We collected water quality and hydrodynamic data (salinity, temperature, water level and velocity) upstream of the barrages. In concert, we assessed fish movement through the barrages using larval drift/tow nets, and open-bay and fishway trapping, and initiated ongoing investigations using high-frequency sonar and acoustic telemetry. We also refined and applied an integrated hydrodynamic model to evaluate salinity risks and explore alternative barrage operating scenarios. Key findings Controlled reverse flows produced short-lived, spatially constrained salinity increases upstream of the barrages. Salinity increased upstream of the barrages when gates were open during reverse flows (notably July–August 2025) but effects attenuated with distance from the barrages and receded once downstream discharge resumed. Reverse flows can facilitate the movement of larval and juvenile/adult fish across the barrage interface. Larval sampling captured 11 species of fish with catches dominated by the catadromous common galaxias (Galaxias maculatus) and marine-estuarine opportunist sandy sprat (Hyperlophus vitattus). Sampling of fish movement through the open barrage bays and fishways captured 13 species, with catches dominated by the freshwater-estuarine opportunist Australian smelt (Retropinna semoni). Our results indicate the potential for substantial flow-assisted upstream movement of fish, particularly for species and life stages unable or unlikely to use fishways. The upgraded hydrodynamic model can represent key salinity and velocity dynamics associated with reverse flow events. Comparisons with logger data showed good agreement, with some over-/under-estimation likely due to the coarse representation of localised flow paths in the model and the timing/representation of gate closures. Scenario modelling suggests operational rules can increase connectivity while managing salinity risk. A “water level trigger” approach (closing gates when downstream levels exceed upstream by >0.05 m) reduced large salt ingress events relative to leaving gates open, while also supporting longer periods of openness (connectivity) than historical gate closure practices. Implications for management Our findings support the feasibility of using controlled, event-based barrage operation to expand the functional estuarine mixing zone and improve freshwater–estuarine connectivity, without implying permanent opening or removal of the barrages. Further research is required but initial management recommendations are apparent, including: 1) integrating consideration of open barrage bays and facilitating reverse flows into contemporary barrage operation; and 2) progressing further automation of barrage gates to allow remote operation in response to real-time water levels. Future modelling, monitoring and research Hydrodynamic modelling results highlight a need to review methods used to estimate barrage discharge, and a need for updated bathymetry data downstream of specific barrages (e.g. Boundary Creek). Expand the hydrodynamic modelling undertaken in this study to include a broader range of contemporary and future barrage operating and climate scenarios. Foster collaboration across SA Water, DEW, CEWH, MDBA, the community and researchers: workshop current and future barrage operating and climate scenarios, and explore opportunities to adopt the concept at additional locations across the barrage network. Continue to collect foundational physico-chemical and biological data in association with reverse flows at the Murray Barrages, particularly under alternate hydrological conditions and across seasons. The current project forms a foundation but was a preliminary exploration. Expand research to consider whole-of-aquatic ecosystem response to reverse flows, including the aquatic-terrestrial interface in regions influenced by potential changes in long-term salinity regimes (e.g. current freshwater environments immediately upstream of the barrages).

This project investigates the feasibility of using community-sourced photographs combined with automated machine learning to assess the body condition of migratory shorebirds in the Coorong, Lower Lakes, and Murray Mouth (CLLMM) region of South Australia. Body condition was assessed using the Abdominal Profile Index (API), scored on an ordinal scale of 1 (lean) to 5 (plump), with an additional class 6 for "fluffed" (non-assessable) birds. Body condition API was scored on a dataset of 697 images of Sharp-tailed Sandpiper (Calidris acuminata), Curlew Sandpiper (Calidris ferruginea), Red-necked Stint (Calidris ruficollis) and Common Greenshank (Tringa nebularia) collected prior to this study. A state-of-the-art, three-phase machine learning pipeline, trained by the expert-elicited body condition assessments, was developed to automate body condition scoring from photographs. The new pipeline represents a substantial methodological advance over previous approaches, offering greater accuracy, transparency, generalisability, and opportunity for scientific audit. Overall model accuracy was high, with the lowest for Class 1 (lean) birds, attributable to underrepresentation of this class in the training dataset, and confidence scores were consistently high for correctly classified images (median 0.5–0.8 for classes 2–5). The field survey and citizen science project encompasses two migratory seasons (2024–2025 and 2025–2026) and targets species including Sharp-tailed Sandpiper, Red-necked Stint, Curlew Sandpiper, and Common Greenshank. A community science programme was established, distributing educational materials to regional conservation groups and government environmental organisations across South Australia. Community submissions via BirdLife Australia's “Birdata” portal were modest. Data gaps were supplemented by 1,947 standardised images across 19 targeted surveys between September 2025 and March 2026, contributed by a professional nature photographer. Applying the trained model to 1,947 unlabelled images revealed meaningful spatiotemporal variation in body condition across species and regions of the CLLMM. The training dataset was class-imbalanced, particularly underrepresenting lean (Class 1) individuals, which contributes to boundary classifier uncertainty and potential central-tendency bias, limiting the ability to score shorebirds as lean. Training labels are currently standardised to a single expert, and we recommend expanding expert assessments to quantify inter-rater reliability. Priority recommendations include expanding the expert-labelled image database across all body condition classes and species, computing inter-rater agreement metrics, and evaluating fluffed-bird classifier performance.