Department for Environment and Water

Abstract Baited remote underwater stereo‐video systems (stereo‐BRUVs) are a popular tool to sample demersal fish assemblages and gather data on their relative abundance and body size structure in a robust, cost‐effective and non‐invasive manner. Given the rapid uptake of the method, subtle differences have emerged in the way stereo‐BRUVs are deployed and how the resulting imagery is annotated. These disparities limit the interoperability of datasets obtained across studies, preventing broadscale insights into the dynamics of ecological systems. We provide the first globally accepted guide for using stereo‐BRUVs to survey demersal fish assemblages and associated benthic habitats. Information on stereo‐BRUVs design, camera settings, field operations and image annotation are outlined. Additionally, we provide links to protocols for data validation, archiving and sharing. Globally, the use of stereo‐BRUVs is spreading rapidly. We provide a standardized protocol that will reduce methodological variation among researchers and encourage the use of Findable, Accessible, Interoperable and Reusable workflows to increase the ability to synthesize global datasets and answer a broad suite of ecological questions.

Context Many Australian mammal species are highly susceptible to predation by introduced domestic cats (Felis catus) and European red foxes (Vulpes vulpes). These predators have caused many extinctions and have driven large distributional and population declines for many more species. The serendipitous occurrence of, and deliberate translocations of mammals to, ‘havens’ (cat- and fox-free offshore islands, and mainland fenced exclosures capable of excluding cats and foxes) has helped avoid further extinction. Aims The aim of this study was to conduct a stocktake of current island and fenced havens in Australia and assess the extent of their protection for threatened mammal taxa that are most susceptible to cat and fox predation. Methods Information was collated from diverse sources to document (1) the locations of havens and (2) the occurrence of populations of predator-susceptible threatened mammals (naturally occurring or translocated) in those havens. The list of predator-susceptible taxa (67 taxa, 52 species) was based on consensus opinion from >25 mammal experts. Key results Seventeen fenced and 101 island havens contain 188 populations of 38 predator-susceptible threatened mammal taxa (32 species). Island havens cover a larger cumulative area than fenced havens (2152 km2 versus 346 km2), and reach larger sizes (largest island 325 km2, with another island of 628 km2 becoming available from 2018; largest fence: 123 km2). Islands and fenced havens contain similar numbers of taxa (27 each), because fenced havens usually contain more taxa per haven. Populations within fences are mostly translocated (43 of 49; 88%). Islands contain translocated populations (30 of 139; 22%); but also protect in situ (109) threatened mammal populations. Conclusions Havens are used increasingly to safeguard threatened predator-susceptible mammals. However, 15 such taxa occur in only one or two havens, and 29 such taxa (43%) are not represented in any havens. The taxon at greatest risk of extinction from predation, and in greatest need of a haven, is the central rock-rat (Zyzomys pedunculatus). Implications Future investment in havens should focus on locations that favour taxa with no (or low) existing haven representation. Although havens can be critical for avoiding extinctions in the short term, they cover a minute proportion of species’ former ranges. Improved options for controlling the impacts of cats and foxes at landscape scales must be developed and implemented.

Accurate ground-based estimation of the carbon stored in terrestrial ecosystems is critical to quantifying the global carbon budget. Allometric models provide cost-effective methods for biomass prediction. But do such models vary with ecoregion or plant functional type? We compiled 15 054 measurements of individual tree or shrub biomass from across Australia to examine the generality of allometric models for above-ground biomass prediction. This provided a robust case study because Australia includes ecoregions ranging from arid shrublands to tropical rainforests, and has a rich history of biomass research, particularly in planted forests. Regardless of ecoregion, for five broad categories of plant functional type (shrubs; multistemmed trees; trees of the genus Eucalyptus and closely related genera; other trees of high wood density; and other trees of low wood density), relationships between biomass and stem diameter were generic. Simple power-law models explained 84-95% of the variation in biomass, with little improvement in model performance when other plant variables (height, bole wood density), or site characteristics (climate, age, management) were included. Predictions of stand-based biomass from allometric models of varying levels of generalization (species-specific, plant functional type) were validated using whole-plot harvest data from 17 contrasting stands (range: 9-356 Mg ha(-1) ). Losses in efficiency of prediction were <1% if generalized models were used in place of species-specific models. Furthermore, application of generalized multispecies models did not introduce significant bias in biomass prediction in 92% of the 53 species tested. Further, overall efficiency of stand-level biomass prediction was 99%, with a mean absolute prediction error of only 13%. Hence, for cost-effective prediction of biomass across a wide range of stands, we recommend use of generic allometric models based on plant functional types. Development of new species-specific models is only warranted when gains in accuracy of stand-based predictions are relatively high (e.g. high-value monocultures).

Movement is a trait of fundamental importance in ecosystems subject to frequent disturbances, such as fire-prone ecosystems. Despite this, the role of movement in facilitating responses to fire has received little attention. Herein, we consider how animal movement interacts with fire history to shape species distributions. We consider how fire affects movement between habitat patches of differing fire histories that occur across a range of spatial and temporal scales, from daily foraging bouts to infrequent dispersal events, and annual migrations. We review animal movements in response to the immediate and abrupt impacts of fire, and the longer-term successional changes that fires set in train. We discuss how the novel threats of altered fire regimes, landscape fragmentation, and invasive species result in suboptimal movements that drive populations downwards. We then outline the types of data needed to study animal movements in relation to fire and novel threats, to hasten the integration of movement ecology and fire ecology. We conclude by outlining a research agenda for the integration of movement ecology and fire ecology by identifying key research questions that emerge from our synthesis of animal movements in fire-prone ecosystems.

Abstract Aim Maintaining genetic diversity and evolutionary processes are important goals in plant conservation. Genetic studies are increasingly undertaken but results from such studies are still rarely implemented as management actions in the field. We address this ‘research‐implementation gap’ by developing a plain‐language genetic assessment approach for population‐level conservation prioritization based on measurement of key genetic parameters. Our aim was to improve understanding between conservation researchers and practitioners, enabling practitioners to incorporate genetic information into conservation actions and conservation genetic researchers to address research explicitly resulting in conservation action. Location Applicable globally. Methods We derived a decision‐making framework that identifies appropriate management strategies for threatened populations based on the level of genetic differentiation ( F ST ), genetic diversity (expected heterozygosity, H E ) and inbreeding ( F IS ), characterized as ‘high’ or ‘low’ in comparison with a reference benchmark. We demonstrate the application of the framework in two case studies of threatened plants and more broadly from the literature. Results Applying the decision framework, we found that for Prostanthera eurybioides, the population of conservation concern does not currently require specialized genetic management and mitigation of ecological threats should be prioritized instead. For Allocasuarina robusta , we found connectivity was high and strategies should be put in place to maintain gene flow. In both cases, genetic information was important for designing restocking strategies accounting for the genetic structure and genetic diversity of source and recipient populations. From the literature, key examples of species types that fit each of the genetic management scenarios are given. Main conclusions We find that the application of our simplified genetic assessment framework helps to clarify management actions based on conservation genetic information for threatened flora, and should assist in bridging the gap between researchers and conservation practitioners for integrated conservation outcomes. Our framework could equally apply to fauna conservation with appropriate consideration of animal‐specific management issues.

This paper attempts to identify and explain patterns in the biogeography of Australia’s indigenous terrestrial mammals at the time of European settlement (before modern extinctions), and also compares species’ pre-European and current status by region. From subfossil, historical and contemporary sources, we compiled data on the past geographic range and present status of mammals for Australia’s 85 biogeographic regions. Of the 305 indigenous species originally present, 91 have disappeared from at least half of the bioregions in which they occurred before European settlement. Thirty-nine extant species ‘persist’ in less than 25% of their original bioregions; 28 of these are marsupials and 11 are rodents. Twenty-two of the original 305 species are extinct, a further eight became restricted to continental islands, and 100 have become extinct in at least one bioregion. Over the same period, 26 species of exotic mammals established wild populations and now occupy from one to 85 bioregions. When we classified the bioregions in terms of their original species composition, the 3-group level in the dendrogram approximated the Torresian, Eyrean and Bassian subregions proposed by Spencer in 1898, while the 4-group level separated southern semiarid Eyrean bioregions, including those in south-west Australia, from the arid Eyrean bioregions. The classification dendrogram showed geographically (and statistically) discrete clustering down to the 19-group level, suggesting that all four subregions can be further divided on the basis of their mammal faunas. Variation partitioning showed 66% of the biogeographical pattern can be explained by environmental factors (related to temperature and precipitation), the spatial position of each bioregion (a third-order polynomial of latitude and longitude), the area of each bioregion, and the richness of species in each bioregion. In addition to the marked distributional changes that indigenous mammals have experienced over the last 200 years, the 49% of variation explainable by temperature and precipitation implies further shifts with global climate change.

Funding for species conservation is insufficient to meet the current challenges facing global biodiversity, yet many programs use expensive single-species recovery actions and neglect broader management that addresses threatening processes. Arid Australia has the world's worst modern mammalian extinction record, largely attributable to competition from introduced herbivores, particularly European rabbits (Oryctolagus cuniculus) and predation by feral cats (Felis catus) and foxes (Vulpes vulpes). The biological control agent rabbit hemorrhagic disease virus (RHDV) was introduced to Australia in 1995 and resulted in dramatic, widespread rabbit suppression. We compared the area of occupancy and extent of occurrence of 4 extant species of small mammals before and after RHDV outbreak, relative to rainfall, sampling effort, and rabbit and predator populations. Despite low rainfall during the first 14 years after RHDV, 2 native rodents listed by the International Union for Conservation of Nature (IUCN), the dusky hopping-mouse (Notomys fuscus) and plains mouse (Pseudomys australis), increased their extent of occurrence by 241-365%. A threatened marsupial micropredator, the crest-tailed mulgara (Dasycercus cristicauda), underwent a 70-fold increase in extent of occurrence and a 20-fold increase in area of occupancy. Both bottom-up and top-down trophic effects were attributed to RHDV, namely decreased competition for food resources and declines in rabbit-dependent predators. Based on these sustained increases, these 3 previously threatened species now qualify for threat-category downgrading on the IUCN Red List. These recoveries are on a scale rarely documented in mammals and give impetus to programs aimed at targeted use of RHDV in Australia, rather than simply employing top-down threat-based management of arid ecosystems. Conservation programs that take big-picture approaches to addressing threatening processes over large spatial scales should be prioritized to maximize return from scarce conservation funding. Further, these should be coupled with long-term ecological monitoring, a critical tool in detecting and understanding complex ecosystem change.

The uptake and acceptance of Geographic Information Systems (GIS) technology has increased since the early 1990s and public health applications are rapidly expanding. In this paper, we summarize the common uses of GIS technology in the public health sector, emphasizing applications related to mapping and understanding of parasitic diseases. We also present some of the success stories, and discuss the challenges that still prevent a full scope application of GIS technology in the public health context. Geographical analysis has allowed researchers to interlink health, population and environmental data, thus enabling them to evaluate and quantify relationships between health-related variables and environmental risk factors at different geographical scales. The ability to access, share and utilize satellite and remote-sensing data has made possible even wider understanding of disease processes and of their links to the environment, an important consideration in the study of parasitic diseases. For example, disease prevention and control strategies resulting from investigations conducted in a GIS environment have been applied in many areas, particularly in Africa. However, there remain several challenges to a more widespread use of GIS technology, such as: limited access to GIS infrastructure, inadequate technical and analytical skills, and uneven data availability. Opportunities exist for international collaboration to address these limitations through knowledge sharing and governance.

Fear of predation is a universal motivator. Because predators hunt using stealth and surprise, there is a widespread ability among prey to assess risk from chemical information - scents - in their environment. Consequently, scents often act as particularly strong modulators of memory and emotions. Recent advances in ecological research and analytical technology are leading to novel ways to use this chemical information to create effective attractants, repellents and anti-anxiolytic compounds for wildlife managers, conservation biologists and health practitioners. However, there is extensive variation in the design, results, and interpretation of studies of olfactory-based risk discrimination. To understand the highly variable literature in this area, we adopt a multi-disciplinary approach and synthesize the latest findings from neurobiology, chemical ecology, and ethology to propose a contemporary framework that accounts for such disparate factors as the time-limited stability of chemicals, highly canalized mechanisms that influence prey responses, and the context within which these scents are detected (e.g. availability of alternative resources, perceived shelter, and ambient physical parameters). This framework helps to account for the wide range of reported responses by prey to predator scents, and explains, paradoxically, how the same individual predator scent can be interpreted as either safe or dangerous to a prey animal depending on how, when and where the cue was deposited. We provide a hypothetical example to illustrate the most common factors that influence how a predator scent (from dingoes, Canis dingo) may both attract and repel the same target organism (kangaroos, Macropus spp.). This framework identifies the catalysts that enable dynamic scents, odours or odorants to be used as attractants as well as deterrents. Because effective scent tools often relate to traumatic memories (fear and/or anxiety) that cause future avoidance, this information may also guide the development of appeasement, enrichment and anti-anxiolytic compounds, and help explain the observed variation in post-traumatic-related behaviours (including post-traumatic stress disorder, PTSD) among diverse terrestrial taxa, including humans.

The extent of increasing anthropogenic impacts on large marine vertebrates partly depends on the animals' movement patterns. Effective conservation requires identification of the key drivers of movement including intrinsic properties and extrinsic constraints associated with the dynamic nature of the environments the animals inhabit. However, the relative importance of intrinsic versus extrinsic factors remains elusive. We analyze a global dataset of ∼2.8 million locations from >2,600 tracked individuals across 50 marine vertebrates evolutionarily separated by millions of years and using different locomotion modes (fly, swim, walk/paddle). Strikingly, movement patterns show a remarkable convergence, being strongly conserved across species and independent of body length and mass, despite these traits ranging over 10 orders of magnitude among the species studied. This represents a fundamental difference between marine and terrestrial vertebrates not previously identified, likely linked to the reduced costs of locomotion in water. Movement patterns were primarily explained by the interaction between species-specific traits and the habitat(s) they move through, resulting in complex movement patterns when moving close to coasts compared with more predictable patterns when moving in open oceans. This distinct difference may be associated with greater complexity within coastal microhabitats, highlighting a critical role of preferred habitat in shaping marine vertebrate global movements. Efforts to develop understanding of the characteristics of vertebrate movement should consider the habitat(s) through which they move to identify how movement patterns will alter with forecasted severe ocean changes, such as reduced Arctic sea ice cover, sea level rise, and declining oxygen content.

Restoration of coastal wetlands has the potential to deliver both climate change mitigation, called blue carbon, and adaptation benefits to coastal communities, as well as supporting biodiversity and providing additional ecosystem services. Valuing carbon sequestration may incentivize restoration projects; however, it requires development of rigorous methods for quantifying blue carbon sequestered during coastal wetland restoration. We describe the development of a blue carbon accounting model (BlueCAM) used within the Tidal Restoration of Blue Carbon Ecosystems Methodology Determination 2022 of the Emissions Reduction Fund (ERF), which is Australia's voluntary carbon market scheme. The new BlueCAM uses Australian data to estimate abatement from carbon and greenhouse gas sources and sinks arising from coastal wetland restoration (via tidal restoration) and aligns with the Intergovernmental Panel for Climate Change guidelines for national greenhouse gas inventories. BlueCAM includes carbon sequestered in soils and biomass and avoided emissions from alternative land uses. A conservative modeled approach was used to provide estimates of abatement (as opposed to on‐ground measurements); and in doing so, this will reduce the costs associated with monitoring and verification for ERF projects and may increase participation in blue carbon projects by Australian landholders. BlueCAM encompasses multiple climate regions and plant communities and therefore may be useful to others outside Australia seeking to value blue carbon benefits from coastal wetland restoration.

Abstract Each year, terrestrial ecosystems absorb more than a quarter of the anthropogenic carbon emissions, termed as land carbon sink. An exceptionally large land carbon sink anomaly was recorded in 2011, of which more than half was attributed to Australia. However, the persistence and spatially attribution of this carbon sink remain largely unknown. Here we conducted an observation-based study to characterize the Australian land carbon sink through the novel coupling of satellite retrievals of atmospheric CO 2 and photosynthesis and in-situ flux tower measures. We show the 2010–11 carbon sink was primarily ascribed to savannas and grasslands. When all biomes were normalized by rainfall, shrublands however, were most efficient in absorbing carbon. We found the 2010–11 net CO 2 uptake was highly transient with rapid dissipation through drought. The size of the 2010–11 carbon sink over Australia (0.97 Pg) was reduced to 0.48 Pg in 2011–12, and was nearly eliminated in 2012–13 (0.08 Pg). We further report evidence of an earlier 2000–01 large net CO 2 uptake, demonstrating a repetitive nature of this land carbon sink. Given a significant increasing trend in extreme wet year precipitation over Australia, we suggest that carbon sink episodes will exert greater future impacts on global carbon cycle.

were identified to affect the photodegradation process significantly, and the possible photocatalytic mechanism was proposed. The as-prepared sample was verified to possess good stability and reusability, suggesting its potential application prospect in the treatment of fluoroquinolone antibiotics.

Abstract Aim The koala is a widely distributed Australian marsupial with regional populations that are in rapid decline, are stable or have increased in size. This study examined whether it is possible to use expert elicitation to estimate abundance and trends of populations of this species. Diverse opinions exist about estimates of abundance and, consequently, the status of populations. Location Eastern and south‐eastern Australia Methods Using a structured, four‐step question format, a panel of 15 experts estimated population sizes of koalas and changes in those sizes for bioregions within four states. They provided their lowest plausible estimate, highest plausible estimate, best estimate and their degree of confidence that the true values were contained within these upper and lower estimates. We derived estimates of the mean population size of koalas and associated uncertainties for each bioregion and state. Results On the basis of estimates of mean population sizes for each bioregion and state, we estimated that the total number of koalas for Australia is 329,000 (range 144,000–605,000) with an estimated average decline of 24% over the past three generations and the next three generations. Estimated percentage of loss in Queensland, New South Wales, Victoria and South Australia was 53%, 26%, 14% and 3%, respectively. Main conclusions It was not necessary to achieve high levels of certainty or consensus among experts before making informed estimates. A quantitative, scientific method for deriving estimates of koala populations and trends was possible, in the absence of empirical data on abundances.

Although recent molecular studies have indicated the presence of a number of distinct species within the C aulerpa racemosa–peltata complex, due to the difficulties presented by high levels of phenotypic plasticity and the large number of synonyms, infra‐specific taxa, and names of uncertain affinity, taxonomic proposals are yet to be made. In this study, we aimed to resolve the taxonomy of the complex and provide an example of how historical nomenclature can best be integrated into molecular based taxonomies. We accomplished this by first determining the number of genetic species within our globally sampled data set through a combination of phylogenetic and species‐delimitation approaches of partial elongation factor TU and RUBISCO large subunit gene sequences. Guided by these results, comparative morphological examinations were then undertaken to gauge the extent of phenotypic plasticity within each species, as well as any morphological overlap between them. Our results revealed the presence of 11 distinct species within the complex, five of which showed high levels of phenotypic plasticity and partial overlap with other species. On the basis of observations of a large number of specimens, including type specimens/descriptions, and geographic inferences, we were able to confidently designate names for the lineages. Caulerpa peltata , C . imbricata and C . racemosa vars. laetevirens, occidentalis and turbinata were found to represent environmentally induced forms of a single species, for which the earlier‐described C . chemnitzia , previously regarded as a synonym of C . racemosa var. turbinata , is reinstated. C . cylindracea , C . lamourouxii, C . macrodisca , C . nummularia and C . oligophylla are also reinstated and two new species, C . macra stat. nov. and C . megadisca sp. nov., are proposed.

The utility of species distribution models for applications in invasion and global change biology is critically dependent on their transferability between regions or points in time, respectively. We introduce two methods that aim to improve the transferability of presence-only models: density-based occurrence thinning and performance-based predictor selection. We evaluate the effect of these methods along with the impact of the choice of model complexity and geographic background on the transferability of a species distribution model between geographic regions. Our multifactorial experiment focuses on the notorious invasive seaweed Caulerpa cylindracea (previously Caulerpa racemosa var. cylindracea) and uses Maxent, a commonly used presence-only modeling technique. We show that model transferability is markedly improved by appropriate predictor selection, with occurrence thinning, model complexity and background choice having relatively minor effects. The data shows that, if available, occurrence records from the native and invaded regions should be combined as this leads to models with high predictive power while reducing the sensitivity to choices made in the modeling process. The inferred distribution model of Caulerpa cylindracea shows the potential for this species to further spread along the coasts of Western Europe, western Africa and the south coast of Australia.

Green Infrastructure is the network of green spaces and water systems that delivers multiple environmental, social and economic values and services to urban communities. This living network strengthens the resilience of urban environments to respond to the major current and future challenges of climate change, growth, health and biodiversity loss, as well as water, energy and food security. With the rapid expansion of towns and cities around the world, the far-reaching value of Green Infrastructure is increasingly recognised by scientific, planning and design communities. A key strategy area of the South Australian Green Infrastructure Project has been development of a sound and credible evidence base to demonstrate the multiple benefits and make the case for investment in Green Infrastructure. A review of local and global literature, with an emphasis on the most recent peer-reviewed research, was carried out between 2012 and 2014. The body of evidence firmly establishes the many and diverse benefits of Green Infrastructure. These include modification of temperatures and climatic conditions, improved human health and well-being, enhanced community liveability, more effective water management, increased economic prosperity, greater opportunity for biodiversity conservation and more extensive urban food production. The primary focus of this paper is the contribution of Green Infrastructure to climate adaptation and protection.

In addition to direct fire cues such as heat, smoke and charred wood, the passage of fire leads indirectly to changes in environmental conditions which may be able to break physical dormancy in hard-coated seeds. After a fire, the open canopy and the burnt material lying on the surface alter the thermal properties of the soil, resulting in elevated soil temperatures for long periods of time. We simulated daily temperature regimes experienced at different depths of soil profile after a summer fire. Our aim was to determine whether these temperature regimes and the duration of exposure (5, 15 and 30 days) play an important role breaking physical seed dormancy in six legumes from south-eastern Australia. Our results showed that simulated temperature regimes break seed dormancy. This effect is specially pronounced at temperatures that are expected to occur near the soil surface (0–2 cm depth). The duration of exposure interacts with temperature to break dormancy, with the highest germination rates reached after the longest duration and highest temperatures. However, the germination response varied among species. Therefore, this indirect post-fire cue could play a role in the regeneration of plant communities, and could stimulate seedling emergence independent of direct fire cues as well as in interaction with direct cues.

Context Over the last 230 years, the Australian terrestrial mammal fauna has suffered a very high rate of decline and extinction relative to other continents. Predation by the introduced red fox (Vulpes vulpes) and feral cat (Felis catus) is implicated in many of these extinctions, and in the ongoing decline of many extant species. Aims To assess the degree to which Australian terrestrial non-volant mammal species are susceptible at the population level to predation by the red fox and feral cat, and to allocate each species to a category of predator susceptibility. Methods We collated the available evidence and complemented this with expert opinion to categorise each Australian terrestrial non-volant mammal species (extinct and extant) into one of four classes of population-level susceptibility to introduced predators (i.e. ‘extreme’, ‘high’, ‘low’ or ‘not susceptible’). We then compared predator susceptibility with conservation status, body size and extent of arboreality; and assessed changes in the occurrence of species in different predator-susceptibility categories between 1788 and 2017. Key results Of 246 Australian terrestrial non-volant mammal species (including extinct species), we conclude that 37 species are (or were) extremely predator-susceptible; 52 species are highly predator-susceptible; 112 species are of low susceptibility; and 42 species are not susceptible to predators. Confidence in assigning species to predator-susceptibility categories was strongest for extant threatened mammal species and for extremely predator-susceptible species. Extinct and threatened mammal species are more likely to be predator-susceptible than Least Concern species; arboreal species are less predator-susceptible than ground-dwelling species; and medium-sized species (35 g–3.5 kg) are more predator-susceptible than smaller or larger species. Conclusions The effective control of foxes and cats over large areas is likely to assist the population-level recovery of ~63 species – the number of extant species with extreme or high predator susceptibility – which represents ~29% of the extant Australian terrestrial non-volant mammal fauna. Implications Categorisation of predator susceptibility is an important tool for conservation management, because the persistence of species with extreme susceptibility will require intensive management (e.g. predator-proof exclosures or predator-free islands), whereas species of lower predator susceptibility can be managed through effective landscape-level suppression of introduced predators.

The lower reaches of the expansive Murray–Darling Basin, Australia, are a hotspot for freshwater biodiversity. The regional ecosystem, however, has been significantly altered by river regulation, including local and catchment-wide water abstraction. Freshwater fishes have suffered from the resultant altered flow regime, together with other threats including habitat degradation and alien species. Impacts reached a critical point (imminent species extinction) during a prolonged drought (1997–2010) that lead to broad-scale habitat loss and drying of refuges during 2007–2010, and urgent conservation measures were subsequently instigated for five threatened small-bodied fish species. A critical response phase included ad hoc interventions that were later incorporated within a broader, coordinated multi-agency program (i.e. the Drought Action Plan and Critical Fish Habitat projects). On-ground actions included local translocation, alien species control, in situ habitat maintenance (e.g. earthworks, environmental water delivery), fish rescues, artificial refuge establishment and captive breeding. Improved river flows signalled an initial phase of recovery in 2011–2012 that included reintroductions. The present paper aims to document the actions undertaken in the Lower Murray, and review successes and lessons from practical examples that will help guide and inform management responses to conserve fish in modified systems subjected to severe water decline.