Persistent pools in non-perennial rivers (NPRs) are critical ecological refuges that sustain biodiversity and ecosystem function during dry periods. However, analysing their long-term dynamics is challenging; existing methods are often field-intensive and unscalable, especially in remote, data-scarce regions. To fill this gap, this study introduces a scalable, data-driven framework that integrates freely available remote sensing, rainfall, and discharge data to investigate multi-decadal surface water and persistent pool dynamics across large scales. The four-step framework (scope definition, data collection, processing, and analysis) is used to derive ecohydrological metrics, identify long-term trends, and determine the key environmental factors influencing pool dynamics. We demonstrate its utility through a case study on a 100 km section of the Gilbert River, Australia (1986-2023). The analysis revealed a dynamic mosaic of 29 persistent pools, with no single pool remaining permanent throughout the study period. Long-term trends showed statistically significant increases in rainfall and discharge, resulting in more numerous, larger, and less fragmented pools. The study confirmed strong correlations between hydrometeorological factors, such as wet-season rainfall and zero-flow duration, and pool morphology and persistence, highlighting the sensitivity of these refuges to seasonal variability and climatic shifts. Spatially, pools were clustered, with off-channel pools being larger and more persistent than their inchannel counterparts, suggesting local hydrogeomorphic controls. By combining standardised ecohydrological metrics with dynamic hydrological years, the approach enables reproducible, cross-system comparison while remaining sensor-agnostic and cost-efficient. This framework provides a transferable pathway for researchers and agencies to support targeted conservation and water-allocation decisions using open data and transparent analytics.

Gape size is a key functional trait that influences feeding performance and ecological fitness across animal taxa; yet, its role within species remains underexplored, particularly in dynamic environments. We investigated the relationship between relative gape size and condition factor (Kn) across 15 species of tropical freshwater fish from northern Australia, spanning four trophic groups: predators, micro-carnivores, omnivores, and herbivores/detritivores. Using linear mixed-effects modelling, we assessed how this relationship varies across trophic groups and seasonal hydrology. Our results revealed a strong positive association where larger gape sizes were associated with enhanced body condition in predators and, to a lesser extent, micro-carnivores—especially during the mid- and late-dry seasons when resource availability is reduced. In contrast, this relationship was weak or absent in omnivores and herbivores/detritivores, whose diets are less likely to be gape-limited. These findings suggest that the functional benefits of larger gape sizes are context-dependent, conferring greater fitness advantages to carnivorous species under seasonal resource limitation than to other trophic groups. Our study highlights the need to consider both trophic ecology and seasonal variability in resource availability regimes when linking traits to performance and provides empirical support for the context-dependent utility of gape size as a functional trait in ecological research.

Context Floodplain habitats support ecological processes in riverine ecosystems, including fish recruitment. In rivers with brief and variable river-floodplain connectivity, the association of young-of-year fish with floodplain pools is poorly understood.Aims We examined whether young-of-year fish associate with floodplain pools in the Fitzroy River, Western Australia, and whether mesohabitat associations (floodplain v. main channel) aligned with expectations from life-history theory.Methods Fish assemblages were sampled at 60 pools over 4 years. Individuals were classified as young-of-year or older using species-specific size thresholds. Mesohabitat associations were assessed using indicator species analysis applied to detection or non-detection data.Key results Young-of-year from 20 of 21 species were detected in floodplain pools, and 16 species either showed a significant floodplain association or no difference between mesohabitats. Ontogenetic shifts in habitat association were common, with adults of several species primarily associated with main channel pools. Young-of-year mesohabitat associations were not well aligned with life-history theory.Conclusions Floodplain pools are widely occupied by young-of-year fish across life-history strategies, despite brief and variable river-floodplain connectivity.Implications Protecting flows that sustain floodplain pools and reconnect them to the main channel is likely to support fish during early life in this highly variable river system.

The rapid decline of freshwater ecosystems has prompted significant investment in riparian restoration globally. Demonstrating positive ecological outcomes from this investment requires adequate monitoring against stated objectives. This study aimed to assess whether ecological outcomes were achieved for revegetation projects undertaken between 2006 and 2023 in the Canning River catchment, Western Australia. To achieve this, we systematically reviewed the funding applications and project reporting for 64 riparian revegetation projects that represented a AUD$3.5 million investment to determine: (1) if the ecological objectives of projects were met, (2) how grant funders can improve the collection and reporting of adequate monitoring data, and (3) if longer-term outcomes could be determined using a remote sensing method. Of the 64 projects assessed, 20 did not have final reports or acquittals. We found there was insufficient monitoring data to evaluate whether projects met ecological objectives of the application or the funding program. A major limiting factor was that most projects were funded for a short duration (12 months) which is insufficient to collect monitoring data and evaluate project outcomes. We identified a range of barriers to adequate collection and reporting of monitoring data, including that 22% of projects did not require applicants to describe a monitoring or data analysis approach, and that of the 75% of projects that did collect data, none could be used to quantitatively assess ecological objectives. We found that remote sensing was limited in its usefulness as a low-cost rapid assessment tool because the spatial analysis required extensive manual correction using specialised expertise to process and analyse the data. We recommend that future restoration programs require proponents include SMART (i.e., Specific, Measurable, Achievable, Relevant and Time-bound) objectives, a clear description of the methods and analysis of data to reflect the objectives, and the use of standardised monitoring methods and collection of baseline data. Quantitative monitoring is a central part of adaptive management and therefore requires adequate time and resourcing.

1. Research is integral to protected area (PA) management, but to be effective, it should be responsive to local needs, priorities and expertise. Recent research has highlighted the risks of 'parks parachuting', where research is done by external researchers without collaboration with, or even acknowledgment of, the national park agency (NPA). 2. Indigenous peoples are increasingly recognised for their role in conservation and PA management and should be an integral party in research, particularly where this occurs in jointly managed PAs. 3. We use Kakadu National Park (northern Australia) as a case study, and we used an analysis of authorship patterns of peer-reviewed publications to demonstrate the need for more inclusive research collaborations with Indigenous people. 4. We then describe an Indigenous-led collaborative research project that developed a strategy to improve collaboration and we identify key decisions and actions that demonstrate the positive role that external researchers played in empowering Indigenous people in research. 5. Synthesis and applications. We make 10 general recommendations to foster equitable and inclusive research collaborations centred on Indigenous-led partnerships, applicable across a range of contexts.

Intact riparian vegetation plays an important role in the energetics of freshwater ecosystems. Plant and animal biomass from the riparian zone may be directly consumed by freshwater organisms without intermediary processing by microbial or metazoan communities. This study examined the extent and importance of such direct consumption by Australian freshwater fishes. We assembled a data set that spanned the Australian continent and examined the extent of terrestrial contributions to freshwater fish diets for 137 species (i.e., approximately one half of Australia's fish fauna). Using a range of multivariate methods we examined variation in reliance on terrestrial material (leaves, fruit, invertebrates and vertebrates) between species and families and between different regions of Australia, particularly focussing on differences between temperate and tropical Australia. We also examined whether particular aspects of morphology were related to increased consumption of terrestrially-derived material. Overall, terrestrial invertebrates contributed 10% or more of the diet of about one quarter of all species and families examined. Most families contained at least one species for which consumption of terrestrial invertebrates was high (> 20%) but only two families, Galaxiidae and Melanotaeniidae, exhibited consistently high consumption across most species. Terrestrial vegetation, principally fruit, contributed 10% or more of the diet in three species from two families only (Terapontidae and Ariidae). There was little regional variation in reliance on terrestrial invertebrates as a food source, but species in south-western Western Australia consistently consumed more than did species elsewhere. Frugivory, in contrast, was significantly more common in the wet-dry tropical northern Australia as reported elsewhere for other tropical regions. Australian freshwater fishes appear to rely more on terrestrial invertebrates than do freshwater fishes elsewhere although data is largely lacking to make broader comparisons with other continents. Morphological correlates with the consumption of terrestrial material were individually and collectively limited, but species with a pelagic habit and an upturned (supraterminal) mouth consumed significantly more such material than did species within other habit/mouth orientation combinations. The present study further confirms the importance of the riparian zone to lotic ecosystem function with direct contributions of terrestrially derived material contributing significantly to the diet of many species and of some families in particular. Although data is lacking for large-scale (i.e., continental) assessments of the importance of terrestrial subsidies to aquatic food webs elsewhere, riparian inputs appear especially important in the Australian context and especially in regions characterised by low primary productivity. Loss or degradation of riparian vegetation can pose a threat to freshwater fishes at a variety of spatial scales. The present study provides a baseline for further studies of the importance of terrestrially-derived material at large spatial scales.

Decomposition of plant litter is a key ecological process in streams, whose contribution to the global carbon cycle is large relative to their extent on Earth. We examined the mechanisms underlying the temperature sensitivity (TS) of instream decomposition and forecast effects of climate warming on this process. Comparing data from 41 globally distributed sites, we assessed the TS of microbial and total decomposition using litter of nine plant species combined in six mixtures. Microbial decomposition conformed to the metabolic theory of ecology and its TS was consistently higher than that of total decomposition, which was higher than found previously. Litter quality influenced the difference between microbial and total decomposition, with total decomposition of more recalcitrant litter being more sensitive to temperature. Our projections suggest that (i) warming will enhance the microbial contribution to decomposition, increasing CO2 outgassing and intensifying the warming trend, especially in colder regions; and (ii) riparian species composition will have a major influence on this process.

While it is widely recognized that reduced river-floodplain connectivity has contributed to the decline of biodiversity in floodplain rivers, surprisingly few studies have quantified the relationship between connectivity, pool persistence, and fish assemblage structure to the level required to generate measurable targets for management. The task is further complicated by the inherent complexity of accurately describing fish assemblages. We maximized our capacity to describe unbiased hydrology-fish relationships by sampling fish assemblages in floodplain pools with a variety of connection histories (60 sampling events), and by using a hierarchical multispecies occupancy model that accounts for changes in sampling design and species detection. Our study was conducted in a tropical wet-dry river threatened by water resource development and elevated temperatures associated with climate change, the Fitzroy River (Western Australia). Our results revealed that wet season (river-floodplain connectivity) and dry season (pool persistence) components of the hydrological cycle influenced fish occurrence in floodplain pools. Pools that were connected to the river by short distances were substantially more species rich than distal pools. This effect was strong at distances <2000 m but negligible at distances greater than 3000 m. Species richness in floodplain pools increased when wet season connection to the river lasted more than 25 days, and when river stage height exceeded 6 m. Prolonged connection to the river (up to 90 days) during overbank flooding (river stage height >11 m) maximized fish species richness in floodplain pools. Dry season components of the hydrological cycle also influenced fish assemblage structure, with pools that persisted during the preceding dry season twice as species rich as those that dried. Our model revealed that sampling gear influenced species detectability, indicating that accounting for variable detection is critical when assessing fish assemblage structure. Given that large flood events are less likely to be impacted by water take, we recommend that managers seeking to maintain floodplain fish diversity ensure that water resource development does not negatively impact pool persistence during the dry season.

Rivers and societies coevolve and emerge through continued engagement over space and time, acting in conjunction to produce dynamic riverscapes. Research on flow regimes has advanced understanding of the temporal dimensions of ecohydrological relationships but has largely neglected the social structures and human experience of time spent living in relation with rivers. We highlight seasonal calendars as a means of creatively organizing and visually representing the temporal dimensions of rivers, which, in both research and practical conservation actions, tend to receive less attention than spatial dimensions. Through 3 case studies relating to major rivers in Australia and South America, we show how seasonal calendars articulate conceptualizations of river time by riparian human communities. Seasonal calendars produced with the participation of local communities can enhance transdisciplinary knowledge production and promote collaboration among different communities of practice-including researchers, conservation practitioners, policymakers, Indigenous peoples, and local communities. Our cases demonstrate that calendars are a tool for articulating and analyzing the many ways that Indigenous people and local communities understand, value, utilize, and shape their biodiverse riverscapes.

Riverine fish assemblages are strongly influenced by attributes of the flow regime. Tropical savannah river systems have distinct and predictable hydrologic seasonality, reflecting the wet-dry climate, but can vary substantially in terms of dry season flow permanency and wet season flow-pulse characteristics. Understanding how flow permanence and variability influence fish assemblages, and whether these factors can be used to predict responses to future hydrological change, are key knowledge gaps that impede effective management. We examined the influence of hydrological variability on the structure and diversity of freshwater fish assemblages across rivers of the wet-dry tropics of northern Australia. We found distinct fish assemblages that varied predictably across three hydrological river types: Intermittent, Perennial Stable and Perennial Flashy flow regimes. This distinction emerged despite a common species pool across the region. Species richness was greatest in rivers with Perennial Stable flow regimes, whereas beta-diversity was greatest in Intermittent rivers. However, life history strategies of constituent species were generally poor predictors of species abundances within each hydrological river type. The distinct fish assemblages evident among hydrological classes may provide some cautious ability to both predict potential fish assemblage changes with future hydrological changes (e.g. if perennial streams became more flashy or intermittent), and to predict fish assemblages expected in unsampled rivers with particular hydrological characteristics. Our findings provide further support for the importance of maintaining regional flow-habitat heterogeneity and the connectivity between hydrological river types, and their essential role for conserving tropical fish species diversity into an uncertain hydrological future.

ABSTRACTMonitoring alien species is critical to their management. However, early detection of invading alien freshwater fish can be challenging due to the difficulty of observing fish in low abundance. Environmental DNA (eDNA) has emerged as a new and potentially more sensitive method for sampling invasive species as compared to conventional methods, but the comparative financial cost is not often assessed. Adoption of eDNA by managers requires studies that showcase its cost‐effectiveness relative to conventional approaches. Here we use eDNA to assist in the management of an aggressive alien fish, the pearl cichlid (Geophagus brasiliensis), that is invading an urban river in south‐western Australia. We applied an occupancy model to survey data collected 6 years apart (2015, 2021) to assess how the species' distribution had changed and to evaluate whether an instream barrier had the potential to limit upstream invasion. To understand the effectiveness of eDNA, we used our model to quantify the relative efficiency (capture probability) of two eDNA sampling methods (active eDNA and passive eDNA) and fyke netting, as well as the number of replicate samples required per site to deliver >95% detection. We coupled the number of replicates needed with the cost per replicate to determine the cost‐efficiency of each method. We found that G. brasiliensis abundance was higher in downstream reaches in both survey years, and there was no evidence that its distribution had changed through time. However, G. brasiliensis was present above the instream barrier. Active eDNA sampling was considerably better at detecting G. brasiliensis than the other methods, making it the most cost‐effective method. Fyke nets came in a close second, and passive eDNA was a very distant third. Our results directly inform management in the study river and broadly highlight the cost‐effectiveness of active eDNA as a freshwater biosecurity tool.

Context Floodplain inundation creates a diversity of aquatic habitats that diverge in their physical, chemical and biological characteristics through space and time, influencing site-scale ecological processes, with implications for local and landscape-scale ecosystem functioning. Aims In this study, we characterise conditions of pools on the floodplain and pools in the main channel of the Fitzroy River, north-western Australia. Methods We used linear models to investigate the spatial and temporal dynamics of top–down and bottom–up forces acting on phytoplankton and zooplankton. Key results Floodplain pools showed considerable heterogeneity compared with the main channel and were shallow and turbid with high nutrient loads, whereas main-channel pools were deep, clear and nutrient-limited. Phytoplankton and zooplankton biomass (mass per unit volume) were considerably greater and more variable in floodplain pools than in the main channel, where both were largely absent. On the floodplain, bottom–up processes drove water-column productivity (e.g. nutrients to phytoplankton to zooplankton) to a degree not observed in the main channel, providing a valuable resource pathway supporting consumer populations locally and catchment wide. We detected no top–down effects in floodplain pools and no top–down or bottom–up effects in the main channel. Conclusion Maintaining flows that inundate the floodplain and promote habitat heterogeneity in the Fitzroy River is crucial for preserving local and landscape-scale ecosystem functioning. Implications Water managers should take into account the important contribution of floodplain pools to the wider riverine ecosystem and ensure these habitats are not unduely affected by water resource development.

Otoliths are calcified structures in the inner ear of fish, the analysis of which can be used to derive important life-history characteristics. Otoliths can be used to age young fish by counting daily growth increments visible in the otolith cross-section; however, this is costly and time-consuming. Otolith weight is a potential surrogate for fish age in growth analysis, providing a rapid alternative. Bony bream (Nematalosa erebi) is Australia’s most widespread freshwater fish and an important component of riverine food webs, yet its life-history characteristics are informed by few publications. We investigated the relationship between assumed fish age derived from otolith increments and otolith weight in young-of-year bony bream. We also assessed the utility of otolith weight for use in relative growth rate analysis. Linear modelling showed a significant positive relationship between increment count and otolith weight. Otolith weight when paired with body length was a reliable alternative to increment count, and thus age, for use in relative growth studies. This method can facilitate research into the factors shaping the life history of this ecologically significant species.

River systems once safeguarded from water development are being developed. This includes intermittent rivers that annually dry to a series of pools. Describing fish species relationships between abundance and pool depth can help managers set water-take rules that protect fish in dry-season pools. We sampled fish in main-channel and floodplain pools that spanned a gradient of depths and overcame sampling challenges by accounting for interacting effects of species mean length, environmental attributes, and sampling attributes on fish capture probabilities. Fish abundance-depth relationships varied systematically with species mean length, mesohabitat type (main channel, floodplain), water turbidity, and structural complexity, highlighting system complexity and the potential generality of abundance-depth relationships. Similarly, fish length moderated the effects of environmental attributes on capture probability for all sampling methods. We evaluated impacts of hypothetical water-take regulations on fish species' distributions. Results suggested that water-take rules prohibiting draining of main-channel pools below 1.65 m and reducing floodplain pools by no more than 14% minimises impacts to species' distributions, promoting conservation of the fish community. Additionally, our approach demonstrates the capacity of species length for predicting distributional and sampling patterns of fish species.

Water development threatens rivers and their biodiversity. Amphidromous shrimp are particularly vulnerable as they require migration between freshwater and estuaries to complete their life cycle. The Fitzroy River is a large tropical intermittent river undergoing water development that is home to the amphidromous shrimp Macrobrachium spinipes (cherabin), yet little is known about its habitat use and flow-ecology making it difficult to inform sustainable water-take. We investigated habitat associations, distributional patterns suggestive of amphidromy, and the influence of water availability by sampling main channel and floodplain pools along a 350-km river length during 2 contrasting flow years. Applying a size-specific abundance model, we estimated abundance per size class, site, and year. We then predicted abundance at the landscape scale with remotely sensed water to reveal the impact of water availability on the meta-population. Our model revealed that juveniles were in greatest abundance in downstream main channel pools, whereas adults were in greatest abundance in upstream floodplain pools. Abundance varied by year with lower numbers predicted in the low-flow year. Longitudinal and habitat patterns remained when our pool-level results were scaled to the landscape, and the positive relationship of abundance to wet-season flow was strengthened. The predominance of smaller cherabin in the lower reaches of the river provides indirect support for an estuarine nursery and amphidromous life history; however, small individuals observed in landlocked pools, during late dry season suggests possible within-river recruitment. The importance of water development policies that protect wet-season flow and passage along the Fitzroy River is supported by this work. These types of policies are likely to be important for this and other amphidromous shrimp species across Australia, Southeast Asia and further afield. Further research detailing the species life history and describing flow-recruitment relationships will be important contributions to understanding this important taxonomic group and refining policies for current and future water resource development.

Accurately describing the hydrology of intermittent rivers is a critical step in improving our understanding and management of freshwater ecosystems. Traditional approaches such as using gauged discharge data provide little information once flow ceases and no insight into the location, morphology, or persistence of river pools. However, multispectral images can be used to describe surface water, characterize hydrology, and provide insight into ecological functioning. A multispectral approach is highly cost-effective and well suited to remote intermittent rivers with little or no gauging infrastructure. Here, we develop an algorithm to extract hydrological attributes (i.e., pool area, length, perimeter, and mean width) from multispectral imagery (Sentinel-2) and use these attributes to create a suite of ecologically relevant hydrological metrics. We describe changes in attributes and metrics in a large lowland intermittent river as it transitions from wet to dry over a four-year period. We also describe temporal changes in attributes and metrics among five river sections with contrasting hydrological persistence and fragmentation. Our algorithm successfully identified surface water in the main channel and the adjacent floodplain, the centerline of pools, and their upstream and downstream ends. Metrics proved effective at describing seasonal patterns in hydrology; revealing how the size, complexity, and elongation of surface water features (e.g., pools) decreased as the study river transitioned from wet to dry and how fragmentation increased. Metrics also successfully differentiated the river sections with varied hydrological persistence. Ecohydrological metrics derived from multispectral imagery have the potential to provide meaningful insights into riverine morphology, resilience, and ecological functioning. Our spatial approach represents a significant advancement in the ability to characterize and manage intermittent rivers, which are increasingly threatened by water resource development and a drying climate.

Riparian trees are critically important for maintaining the ecological function of freshwater ecosystems. Globally, anthropogenic changes to water regimes are impacting the health and distribution of riparian trees. Understanding the physiological constraints on the distribution of riparian tree species in relation to the water regime is essential for informed water resource management that seeks to limit impacts on riparian trees. To fill an identified knowledge gap for the Fitzroy River, a significant river in the wet–dry tropics of north‐western Australia, we used a trait‐based approach to characterise nine common riparian tree species in relation to their distribution along a hydrological gradient. We assessed key functional traits related to drought and flood flows. Leaf mass per unit area (LMA), leaf dry matter content (LDMC), foliar carbon content (% C) and the ratio of carbon to nitrogen (C:N) are broadly related to plant productivity and durability and may reflect resistance to fluvial stress. Traits related to water availability were stem specific density (SSD), mean xylem vessel diameter and xylem vessel density, as well as foliar δ 13 C which is related to water use efficiency, and leaf osmotic potential at full hydration ( π 100 ) as a measure of drought tolerance. We found that π 100 , δ 13 C and SSD values reflected species hydrological habitat preferences, with higher δ 13 C values and lower π 100 for species distributed in the drier floodplain habitats, compared with species constrained to the riverbank. Low SSD values for species close to the riverbank may be indicative of aerenchyma tissue in response to flooding. Differences in leaf trait values were primarily attributed to differences between evergreen Myrtaceous and deciduous non‐Myrtaceous species, rather than hydrological habitat preferences. LMA was greatest for the Myrtaceous study species, with deciduous non‐Myrtaceous species associated with wetter habitats having the lowest values. The physiology of riparian trees in northern Australia is not well described for the majority of species. Our study provides insight into plant functional strategies in response to both flooding and drought and is a critical step in understanding plant responses to future water‐take scenarios to support evidence‐based decision making.

Most of the planet's vital ecosystems are managed on lands owned by Indigenous peoples. Indigenous people face many challenges in managing these lands, including rapidly growing threats causing species extinctions and ecosystem losses. In response, many Indigenous groups are looking for ethical ways to use digital technology and data analytical tools to support their existing knowledge practices to solve complex environmental management problems. We draw on an action co-research project to show how a range of knowledge coproduction mechanisms were developed and applied to weave Indigenous knowledge, artificial intelligence (AI), and technical sources to monitor the health of Nardab, a culturally significant and Ramsar-listed wetland in Australia's World Heritage-listed Kakadu National Park. The coproduction mechanisms included: holistic assessments of the health of indicators; a dynamic and creative decision-support tool to adaptively manage a complex system; ongoing monitoring and testing of knowledge used for collaborative action; and Indigenous-led governance of research activities and impact at local and regional scales. It was important for local Bininj traditional owners to determine where and how multiple sources of evidence could or should be used and applied to direct and assess on-the-ground actions as part of this collaborative and cross-cultural knowledge sharing and coproduction process. At Nardab, this required negotiating the evidence from qualitative Indigenous-led assessments of significant sites and quantitative ecological information collected and analyzed from cameras and drone surveys. The coproduction mechanisms developed provided a practical and ethical means of empowering different sources of knowledge for adaptive decision making while respecting and protecting differences in how knowledge is generated, interpreted, and applied.

Mapping surface water using remotely sensed optical imagery is a particular challenge in intermittent rivers because water contracts down to narrow linear features and isolated pools, which require accurate water detection methods and reliable image datasets. Of the many methods that use optical sensors to identify water, the Water Detect algorithm stands out as one of the best options due to its classification accuracy, open-source code, and because it does not require ancillary data. However, in the original study, the Water Detect algorithm was only tested with Sentinel-2 imagery. High-resolution and high-frequency imagery, such as Planetscope, combined with sharpening and band synthesizing techniques have the potential to improve the accuracy of surface water mapping, but their benefit to the Water Detect algorithm remains unknown. Uncertainty also exists about the extent to which different input parameters (i.e. maximum clustering and regularization) influence the accuracy of Water Detect. Practitioners seeking to map surface water in intermittent rivers need guidance on a best-practice approach to improve the accuracy of Water Detect. To meet this need, we automated an existing method for sharpening and synthesizing bands and applied it to a series of multispectral Sentinel-2 and Planetscope images. We then developed a sensitivity analysis algorithm that compared the accuracy for all possible combinations of input parameters in a given range for the water detection process - enabling optimal parameters to be identified. We applied this workflow to an 81 km stretch of the lower Fitzroy River (Western Australia) to periods when spatial water extent varied markedly, i.e. mid-wet (February), early-dry (June), and late-dry season (October), across three years with variable wet season flow. We found that the ability to accurately detect surface water using multispectral imagery was increased by using input parameters identified by the sensitivity analysis and using Visible + Near-infrared (VNIR) bands, with relatively little gained by image sharpening unless the area of interest was burnt or experienced considerable shading. Also, the regularization parameter exerted less influence on results than maximum clustering. Importantly, the accuracy of the Water Detect algorithm can vary drastically if input parameters are not calibrated to local conditions. Results also revealed that our approach was adept at detecting linear features in intermittent rivers. We recommend that practitioners using Water Detect to identify surface water undertake a workflow similar to that described here to improve the accuracy of the Water Detect algorithm. The automated routines provided by this study will significantly assist practitioners in doing so. Increasing the accuracy with which we detect and map water in intermittent rivers will improve our understanding and management of these important systems which are under increasing threat.

Riverine floodplains are highly productive habitats that often act as nurseries for fish but are threatened by flow regulation. The Fitzroy River in northern Australia is facing development, but uncertainty exists regarding the extent to which floodplain habitats deliver benefits to fish, particularly given the brevity of seasonal floodplain inundation. We investigated the growth rate of young-of-year bony bream (Nematalosa erebi) in main channel and ephemeral floodplain habitats using age derived from otolith daily increments. We also investigated potential mechanisms influencing growth and modelled the consequences of differential growth rate on survival. Our results revealed higher growth occurred exclusively on the floodplain and that zooplankton biomass was the best predictor of growth rate. Modelling indicated that elevated growth rate in high-growth floodplain pools (top 25th percentile) could translate into substantial increases in survivorship. The positive effect of zooplankton biomass on growth was moderated under highly turbid conditions. Temperature had a minor influence on growth, and only in floodplain habitats. Our results indicate ephemeral floodplain habitats can deliver substantial growth and survival benefits to young-of-year fish even when floodplain inundation is brief. This study highlights the need to ensure that water policy safeguards floodplain habitats due to their important ecological role.