Habitat use by nomadic ibis and spoonbills post-dispersal from breeding sites

Habitat use by nomadic ibis and spoonbills post-dispersal from breeding sites | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Habitat use by nomadic ibis and spoonbills post-dispersal from breeding sites Heather M. McGinness, Luke R. Lloyd-Jones, Freya Robinson, Art Langston, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4626784/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Nov, 2024 Read the published version in Landscape Ecology → Version 1 posted 10 You are reading this latest preprint version Abstract Context Nomadic waterbirds are highly mobile across a range of spatial and temporal scales, which makes it difficult to monitor, quantify, and predict their habitat use with traditional methods, especially between breeding events when individuals and flocks can move over vast areas. Objectives This study aimed to provide accurate information on habitat use to improve strategic conservation management of these species, particularly the provisioning of environmental water. Methods To overcome the challenges of distance and remoteness, we analysed a 7-year GPS satellite telemetry dataset from 141 individuals. We quantified habitat selection post-dispersal from breeding sites, and predicted habitat preference for two wading waterbird species of the Threskiornithidae family that frequently nest together at the same sites: straw-necked ibis ( Threskiornis spinicollis ) and royal spoonbill ( Platalea regia ). Results Both long-term and short-term landscape-scale habitat associations differed between species. Royal spoonbills used fewer and more restricted habitat types than straw-necked ibis. Spoonbills displayed strong preferences for reservoirs, marshes and permanent wetlands, while ibis used both aquatic and terrestrial habitat, including areas of intensive animal production, modified pasture, and woodlands. Analysis of nocturnal versus diurnal space use showed that roosting and foraging habitat requirements for both species are distinct. Conclusions Analysing over 1 million telemetry points revealed species-level variability in habitat use, informing resource allocation for environmental water management. Royal spoonbills are more vulnerable to habitat change due to water regime alterations, highlighting the need for focused conservation management. Differences in day and night habitat use indicate the necessity of considering roosting habitats alongside foraging habitats for effective conservation. This comprehensive understanding of waterbirds' spatiotemporal interactions with their environment is crucial for long-term management aimed at increasing waterbird numbers and maintaining diversity. environmental water satellite telemetry foraging nomadic habitat selection conservation management waterbirds Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Understanding the habitat use of a species across its life cycle is essential for effective management. Without this holistic knowledge, conservation efforts may prove ineffective in the long-term despite successful protection of habitat associated with an individual aspect of the species’ life cycle (Haig et al. 1998). For example, conserving a species’ breeding habitat might not increase recruitment into the population if there is post-natal dispersal into an unprotected area with high mortality (Price et al. 2018). This is particularly critical for species reliant on habitats that are susceptible to loss or alteration. One such group of species is waterbirds, which are highly reliant on wetlands that are being lost at increasing rates due to floodplain development, water abstraction, pollution and climate change (Davidson 2014; Haig et al. 1998; Kirby et al. 2008), leading to international agreements to protect them (e.g. Ramsar Convention on Wetlands). In Australia, waterbirds have declined significantly since European colonisation (Bino et al. 2021; Bino et al. 2015; Leslie 2001) and addressing these declines is a specified management goal for large management areas, for example the Murray Darling Basin (MDBA 2014). Supporting or recovering waterbird populations requires intensive conservation management because different species have different degrees of dependency on flooding to provide habitat; for example, aggregate-nesting wading waterbirds require extensive areas of shallow surface water for both foraging and breeding (Brandis et al. 2018; Chiau 2006; Lindsell et al. 2009; Serra et al. 2011; Wang et al. 2014). This is a particular challenge in inland areas of Australia that are naturally water poor but are subject to extensive agricultural water abstraction and regulation together with increasing drought frequency, duration and severity. Management can include the allocation of environmental water or ‘environmental flows’, to provide the quantity, timing, and quality of freshwater flows and water levels necessary (Arthington et al. 2018). Efficient application of environmental water by managers, including where to water, when to water, and for how long, requires knowledge of species habitat use and movements (Reid 2013). Many waterbird species are highly vagile, displaying a range of movement strategies at varying spatial and temporal scales across their lifecycle (Kingsford et al. 2010). However, understanding of waterbird habitat use across all stages of the lifecycle is often insufficient, particularly in remote inland areas, which hinders strategic management. Large aggregate-nesting wading waterbird breeding events and locations in Australia have been relatively well studied and the focus of significant management effort, due to the national and international importance of associated sites, their susceptibility to adverse effects of environmental change, and the large aggregations of waterbirds of multiple species that they attract ((Bino et al. 2015; Bino et al. 2014; Brandis et al. 2018; Brandis et al. 2024; Brandis et al. 2011; Kingsford and Johnson 1998; Kingsford 2000; Kingsford and Auld 2005). Comparatively, far less is understood about habitat or site use of these species after dispersal or between breeding events (Caley et al. 2022); yet critical population processes occur outside of breeding events with the potential to shape demography and population viability (Calvert et al. 2009; Carrick 1962). For waterbirds in the Threskiornithidae family, it can take 3-4 years for juveniles to mature and recruit into the adult breeding population, during which time high rates of attrition occur (Robinson and McGinness 2024). Likewise, adults disperse until the next breeding event; events which are irregular and can be years apart. These life stages involve movements at scales of hundreds to thousands of kilometres, frequently in remote areas (Kingsford and Norman 2002, Nicol et al. 2024). While banding studies have been valuable sources of information on waterbird movement, major knowledge gaps remain, especially concerning movement and habitat use in remote areas (Smith and Munro 2011). In this context, continuous GPS telemetry monitoring movements and habitat use over multiple years and individuals provides invaluable data for quantifying habitat selection across the full life cycle. These data can then inform where and when management actions and resources might be best targeted to improve habitat conditions and therefore long-term waterbird population outcomes. Here we use a GPS satellite telemetry dataset of 141 individuals spanning seven years to quantify non-breeding habitat requirements for two wading waterbird species of the Threskiornithidae family: straw-necked ibis Threskiornis spinicollis (SNI) and royal spoonbill Platalea regia (RSB). We hypothesised that topography, land-use, vegetation, and water indices would predict habitat selection, and that habitat selection and habitat availability would differ between SNI and RSB. Methods Transmitter deployment We tracked 141 birds (SNI = 89, RSB = 52) using satellite telemetry with tracking durations ranging from 30 – 2,000 days per bird and spanning October 2016 – January 2024. Individuals were captured and fitted with transmitters in eight wetlands of the Murray-Darling Basin (MDB), which spans approximately 14% of Australia’s landmass and is subject to intensive water and wetland management (Figure 1). We fitted birds with solar-powered GPS transmitter units from three sources: Druid (Druid Technology Co., Ltd., Chengdu, China), Geotrak (Geotrak Inc. North Carolina, USA), or Ornitela (Ornitela, UAB Vilnius. Lithuania). Transmitters weighed 12–40 g and ranged from < 1% to 5% of bird bodyweight. GPS fix resolution was 15–26 m and fix frequency ranged from one minute to 6 hours (depending on transmitter type and programmed duty cycle). Fix frequency was considered in analyses, with interpolation or down-scaling applied when appropriate. The data were transmitted via either the Argos satellite network (for Geotrak units) or the 3G network (for Ornitela and Druid units). We captured birds either by hand, or with leg-nooses, or by using a net launcher. We attached transmitters as a ‘backpack’ using harnesses made of Teflon ribbon or Spectra ribbon (Bally Ribbon Mills TM ), fitted either as wing-loops with a join at the keel (SNI and some RSB), or as leg-loops (RSB). Harness design was based on designs used in other species (Jirinec et al. 2021; Karl and Clout 1986; Roshier and Asmus 2009; Thaxter et al. 2014) and modified and improved over time. Data pre-processing Data were filtered based on dilution of precision, removal of outliers from visual inspection of tracks, and removal of points based on extreme outliers on distance travelled per unit time and precise vertical or horizontal movements. To reduce temporal autocorrelation and produce a standardised temporal fix frequency and timing, we interpolated and thinned the data to one fix every six hours. This reduced the initial data set of ~ 1.6 million presence points to ~ 123,000, with movements spanning ~ 4 million km 2 in eastern Australia (Figure 2). Most of the fix removals were for individuals tracked at very high frequency with Druid units. We removed data tracking movements from both nesting birds and points from pre-dispersal from capture site, since this study focused on habitat selection between breeding events, not during nesting or associated with capture sites (frequently nest sites for adults). This left 109,034 presence points across the two species. Background data We conducted data analysis and modelling using R Statistical Software version 4.4.0 (R Core Team 2024). To compare habitat selection as indicated by species presence (hereafter ‘presence’ data) to background habitat availability (i,e., pseudo-absence; hereafter ‘background’ data), we generated utilisation distribution grids for each individual. We fitted the ‘kernelUD’ function in the adehabitatHR package to create utilisation distribution from all presence points for each individual (Calenge & Fortmann-Roe, 2023). We then used the ‘getverticeshr’ function to extract the 99% contour of the utilisation distribution and sampled points on a regular grid such that the ratio of background points to presence points was approximately 10:1. The total dataset after initial quality control consisted of 84,425/ 900,149 and 24,609/ 284,909 presence/background locations for SNI and RSB respectively. We annotated presence and background data with several environmental covariates, including: Water Observations from Space (WOfS; (GEOSCIENCE AUSTRALIA, 2015; Mueller et al., 2016); Multi-Resolution Valley Bottom Flatness (MrVBF; Gallant et al., 2012); the Australian National Aquatic Ecosystem classification (ANAE; (Brooks 2021)); the Catchment Scale Land Use of Australia (CLUM; ABARES (2021)); and, the National Vegetation Information System (NVIS; NVIS technical Working Group, 2013). These datasets are extensively described in their respective sources, but briefly: WOfS summarises 30 years of Landsat imagery to provide a long-term understanding of the recurrence of water in the landscape (25 m resolution); MrVBF is a topographic index that uses digital elevation models to classify degrees of valley bottom flatness (30 m resolution); the ANAE integrates the best available mapping data from multiple sources combined with simple rules to define aquatic ecosystem types using a number of relevant attributes (e.g. water regime, water source, salinity, landform and dominant vegetation); CLUM summarises a single dominant land use for a defined area based on the objective of the land manager as identified by state and territory agencies and classified according to a three-tiered hierarchical structure based on potential degree of modification and the impact on a putative ‘natural state’ (50 m resolution); and, NVIS summarises the extent and distribution of vegetation types in the landscape (100 m resolution). For WOfS, we used the recurrence frequency of water as a percentage of the total number of clear observations for each grid cell. We rescaled WOfS to have a mean of 0 and a variance of 1 to assist with model convergence. To improve computational efficiency and facilitate interpretation, we chose the highest level of classification for each of the multilevel covariates which left 9 levels for MrVBF, 32 levels for CLUM and 30 levels for NVIS. To retain as many complete cases as possible, we imputed a small proportion (~1%) of points that reported no value for each covariate with either a new dummy variable (for MrVBF) or the unknown/no data variables for NVIS and CLUM. A summary of proportions of presence and background points stratified by the levels of each of the variables was initially performed as reference for model outputs. Habitat selection modelling For each species, we fitted an infinitely weighted logistic regression mixed model (IWGLMM) to model habitat selection with the environmental covariates as predictors using presence and background data with the glmmTMB package (Brooks et al. 2017). IWGLMM performs similarly to inhomogeneous Poisson point process and maximum entropy (maxent) for habitat selection modelling (Fithian and Hastie 2013). WOfS was continuous while the other four predictors (MrVBF, ANAE, CLUM, and NVIS) were categorical. Quadratic terms (fixed and random) were included for WOfS to model potential nonlinearity; linear terms were used for the categorical variables. Models were also explored within species split by age (juvenile or adult), season (summer = December – February, autumn = March – May, winter = June – August, spring = September – November), and time of day (diurnal = points from 12pm; nocturnal = points from 12am) fitted by partitioning the data by these variables. Night fixes for SNI represented roosting locations because they are known to roost in one place from sunset to sunrise; conversely, RSB are less predictable, often travelling or feeding at night, and either feeding or resting during the day. We used a generalized linear mixed model (GLMM) framework with random intercepts and random slopes stratified by individual to account for between individual variation and reduce pseudo-replication. We fixed the variance at 10 3 rather than shrinking individual-specific intercepts towards the overall mean, following Muff et al. (2020). We set the weight of background points as W = 10 5 , in line with previous IWGLMM applications (Muff et al. 2020; Renner et al. 2015). Initial analyses showed poor convergence when fitting all covariates in a single model due to complete or near-complete separation of factor levels and highly correlated variables. Highly correlated between-covariate (e.g. WOfS and water land use variables) water-based variables were also difficult to interpret in a single model. To overcome this numerical instability, we created a ‘low count’ factor level for all factor variables if at least one cell count of the two-way table of the factor level versus presence-availability status was less than 50. To further improve numerical stability and aid interpretation, we fitted each of the environmental covariates separately. For each species fitted with the IWGLMM we summarised fixed and random parameters and performed multiple comparison correction for fixed effect parameters following Benjamini (2001). Habitat predictions We constructed maps of relative probability of occurrence using a regular grid of points for the MDB (long. min = 138.57, long. max = 152.49, lat. min -37.68, lat. max = -24.59) at ~ 500 m resolution (~ 4 million total pixels). We constrained the predictions to the MDB because waterbird capture sites were targeted to this area and results are likely more representative of preferences for birds originating from this area. We matched environmental covariates to these points and used the ‘predict.glmmTMB’ from the glmmTMB package function along with the fitted GLMM models to produce population-level predictions (i.e., setting all random effects to zero). Model predictions were made separately for each of the environmental covariates (WOfS, MRVBF, ANAE, CLUM, NVIS) and combined into a single prediction using a weighted average of predictions, with the weights taken as the squared correlation Spearman's rank from five-fold cross-validation. Model validation We compared predictive performance of the fitted model using five-fold cross-validation to assess out-of-sample predictive error and model weights of each of the fitted models. Each fold consisted of an approximately equal number of individual birds. We trained models iteratively on four of five folds and made predictions with the fifth. We investigated overlap of presence points with the predicted relative habitat preference at a population scale by binning the predicted relative occurrence output over all presence absence points into deciles and computing the frequency of observed presence points within each decile. We then estimated the Spearman's rank correlation coefficient between frequency of the validation presence points within each bin and the bin rank (1-low to 10-high use), also referred to as the ‘Boyce Index’ (Hirzel et al. 2006). A strong positive correlation indicates the model assigns presence points well to preferred habitat (McCabe et al. 2021; Peters et al. 2015). Bi-monthly surface water relationships To explore potential relationships between short-term surface water presence and habitat selection, we intersected telemetry points with the Multi Index Method (MIM) surface water dataset for the MDB. This dataset represents a spatial time series of estimated water depth (in mm, error ± 500 mm) and maximum surface water extent at a single point in time every two months, based on multi-index surface water extents, a digital elevation model, and gauged water depths (Teng et al. 2023). After matching the six-hourly gridded telemetry data used for the prior analyses to the spatial and temporal boundaries of the MIM dataset, 64,220 telemetry points remained for analysis from 97 individuals (SNI, N = 60; RSB = 30). Given that the MIM data are time varying, we generated control points differently to the static variables used in the static habitat selection analyses. We sampled points uniformly at a ratio of 5:1 in a circular buffer around each point with a radius equal to the distance to the next point. For each bird, we intersected both observed and control telemetry points with the MIM datasets. For each point, we produced water presence/absence and depth (mm) datasets. We also produced water depth summary statistics (mean, median, max) and calculated the proportion of wet points for 2 km and 20 km radii around each point. We then used the mgcv package (Wood 2017) to produce generalised additive models (GAMs) that were used to explore relationships between tracked bird presence and: 1) water presence/absence and depth at the point/pixel; 2) water depth statistics within a) 2km and b) 20 km radii around each point; and, 3) the proportion of points wet within a) 2km and b) 20 km radii around each point. Results Habitat selection and prediction – straw-necked ibis The spatially explicit SNI habitat preference predictions (Fig. 3 ) were highly robust, with strong discrimination on held-out data and a cross-validated Spearman rank correlation of 0.86 from the weighted predictor. We also found a non-linear relationship between SNI habitat preference and WOfS (i.e., the frequency of surface water) with an optimal range of 15–20% surface water presence frequency over time, and a substantial decrease in suitability for values above 30%, where surface water was present more frequently (Fig. 4 ). We found positive and negative associations with the categorical environmental variables (Fig. 5 , Supplementary Table 1 and Supplementary Figs. 1–4). We note that predictive performance is MDB specific whereas habitat preferences are for the whole of the data set, which included information from outside the MDB. For the MRVBF topographic index classes we found a strong positive association with large depositional basins and a negative association with small hillside deposits. Positive associations with ANAE aquatic ecosystem classes included permanent and temporary lowland streams, woodland riparian zone or floodplain, temporary tall emergent marsh, clay pan and temporary shrub swamps while there was a strong negative association with permanent forb marsh (Fig. 5 , Supplementary Table 1). A wide range of CLUM catchment-scale land use classes were associated with SNI, including river, reservoir/dam, intensive animal production, marsh/wetland, residential and farm infrastructure, services, grazing irrigated modified pastures, grazing native vegetation, channel/aqueduct, irrigated cropping, transport and communication, and grazing modified pastures. There was also a wide range of NVIS vegetation classes positively associated with SNI, including inland aquatic areas (freshwater, salt lakes, lagoons), eucalypt woodlands, cleared non-native vegetation and buildings, grasslands, herblands, sedgelands and rushlands, eucalypt open forests and open woodlands, shrublands, including chenopod shrublands, samphire shrublands and forblands, modified native vegetation, and tussock grasslands. Habitat associations were similar between adult and juvenile SNI with few categories of preferred vegetation for juveniles and a preference for narrow valley floors that adults do not share (Supplementary Figs. 5 and 6). We found nocturnal roosting habitats differed from diurnal foraging habitats. Nocturnal roosting habitats were positively associated with river, stream, wetland, channel and reservoir environments, usually with trees; whereas diurnal foraging habitats usually occurred in more open areas, both natural and agricultural (Supplementary Figs. 7 and 8). We found seasonal differences in SNI habitat preferences in relation to MRVBF topographic values and ANAE aquatic ecosystem and NVIS vegetation classes, but CLUM land use class preferences were similar year-round (Supplementary Figs. 9–12). SNI showed a distinct preference for large depositional basins in summer (and to some extent spring), but a wider range of MRVBF zones used in autumn and winter, including valley floor environments and small depositional basins. There was more use of non-ANAE (i.e., terrestrial) areas in autumn compared to other seasons. Fewer ANAE classes were used in winter, with a preference for permanent and temporary lowland streams and woodland riparian zones or floodplains. The range of NVIS classes used in summer was greater than in all other seasons, with a greater preference for shrublands. NVIS class preferences were similar in autumn and winter, with SNI favouring eucalypt woodlands, cleared non-native vegetation and buildings, inland aquatic habitats (freshwater, salt lakes, lagoons), grasslands, herblands, sedgelands and rushlands, and eucalypt open woodlands and forests. A subset of these was preferred in spring, including eucalypt woodlands, cleared non-native vegetation and buildings, and inland aquatic habitats (freshwater, salt lakes, lagoons). Habitat selection and prediction – royal spoonbill The spatially explicit RSB habitat preference predictions were also highly robust, with predictions showing strong discrimination on held-out data with a cross-validated Spearman rank correlation of 0.95 from the weighted predictor (Fig. 6 ). There was a non-linear relationship between RSB habitat preference and WOfS, with a higher optimal frequency of surface water presence than SNI of 20–45%, and a decrease in suitability for areas with water presence frequencies above 60% (representing semi-permanent and permanent water; Fig. 4 ). As for SNI, we found evidence for positive and negative associations for RSB with the categorical environmental variables (Fig. 7 , Supplementary Table 2 and Supplementary Figs. 13–16). For the MRVBF topographic index, we found a positive association with depositional basins. Positive associations were observed with the following ANAE classes: permanent wetland, permanent lowland stream, temporary lowland stream, temporary tall emergent marsh and temporary shrub swamp. We found RSB selected for a range of CLUM (i.e., land uses) and NVIS (i.e., vegetation) classes; however, this range was smaller for RSB compared with SNI. Preferred CLUM classes were reservoir/dam, marsh/wetland, rivers, lakes, nature conservation and grazing native vegetation. Preferred NVIS classes were eucalypt tall open forests, inland aquatic (freshwater, salt lakes, lagoons), and eucalypt open forests. We found circadian differences in RSB habitat preferences between the ANAE classes (Supplementary Figs. 17 and 18). At night, RSB preferred temporary tall emergent marsh, whereas diurnal preferences were for permanent lowland stream and temporary lowland stream habitats. We did not find any differences between day and night habitat selection for the other environmental covariates. There were insufficient data for age group or seasonal comparisons for RSB. Bi-monthly surface water relationships There were significant relationships between tracked bird presence and average water depth at both the point/pixel and 2km radii scales for both species (Supplementary Tables 3–6). There were also relationships between the proportion of water in the landscape and species presence at the 2km radii scale. However, at the 20 km scale, there were no relationships between tracked bird presence and either water depth statistics or the proportion of points wet (Supplementary Tables 7–8). At point/pixel scale, the GAM smooth model suggested that SNI most preferentially selected sites with water depths of 12–14 m (Supplementary Fig. 19). However, there were relatively few data points in this range and ~ 85% of observations were at depths between 0-0.5 m (Supplementary Fig. 20), so the size of this preference should be interpreted with caution. When split by age group, this ‘deep water’ model peak appeared to be strongly related to the juvenile SNI model, with adult models indicating bimodal preferences for water depths of ~ 2 m and ~ 7.5m (Supplementary Fig. 19). RSB GAM smooth models at the point/pixel scale were all strongly bimodal, suggesting distinct preferences for locations with shallow water ~ 1 m deep, as well as locations with water ~ 5 m deep (Supplementary Fig. 19). Similar to SNI, ~ 85% of observations were actually at depths 0-0.5 m, with relatively few observations at greater depths. Overall model patterns were also strongly influenced by juvenile data, with the second peak in preferences for adults at ~ 4 m rather than at 5 m as seen for juveniles (Supplementary Fig. 21). At the 2 km radii landscape scale, models suggested that SNI presence was greatest where water depths averaged 0.5–2.5 m, with a rapid decline in presence as average depths increased beyond this average (Supplementary Fig. 22). SNI presence was also high where water depths were zero. When split by age group, the pattern was similar for SNI adults, however models suggested juvenile presence peaked where water depths were < 0.5 m and declined rapidly beyond this depth. The proportion of the landscape inundated within the 2km radii was also significant (Supplementary Fig. 22), with SNI presence highest where 50–75% and 10–20% of the landscape was inundated and declining after > 75% of the area was inundated (Supplementary Fig. 22). Again, SNI presence was also high in areas with no inundation. Adult patterns for SNI were similar to overall patterns, but juvenile presence was highest at 50–70% water coverage and declined after > 60% of the area was inundated (Supplementary Fig. 22). At the 2km radii scale, RSB presence peaked slightly at shallow average depths of 0.5–1 m but was consistent at a range of average depths up to 2.5 m. In contrast to SNI, RSB presence was very low where the proportion of the area inundated was low, and after peaking at 25% of the 2 km radius inundated, RSB presence remained relatively high with increasing area inundated all the way to 100% inundation (Supplementary Fig. 23). Discussion Effective management and policy to support waterbirds across broad geographic areas requires a comprehensive understanding of habitat requirements, including seasonal change in preferences. This has historically been difficult to quantify for nomadic species, particularly those in remote areas and outside of known breeding locations when individuals disperse hundreds to thousands of kilometres. Advances in satellite tracking have enabled data collection over vast distances and in remote areas. Consequently, we have been able to characterise non-breeding habitat preferences for two such nomadic waterbirds. This knowledge has potential to inform identification and prioritisation of potential sites for environmental watering and other management, to support a) juvenile survival to breeding age (recruitment), b) adult recovery from breeding and c) adult survival between breeding events. While ibis and spoonbill species in Australia share breeding sites and nest in mixed aggregations, their non-breeding habitats differ. RSB have primarily aquatic diets and adaptations focused on feeding in surface water (‘obligate wetland foragers’) whereas SNI have mixed terrestrial and aquatic diets and foraging strategies (‘non-obligate wetland foragers’). We found strong evidence that these species consequently require different foraging-habitat provision and management. RSB habitat use is more strongly related to both long-term and short-term inundation frequency, extent and depth than SNI habitat use. RSB displayed strong preferences for reservoirs, marshes and permanent wetlands, while SNI used both aquatic and terrestrial habitat, including areas of intensive animal production, modified pasture, and woodlands. Furthermore, we found RSB use fewer and less available habitat types compared to SNI. This leaves RSB more vulnerable to habitat change and loss and increases the relative importance of management for this and similar species such as egrets, where water resource developments, climate change impacts or other threats are present. Our analysis of nocturnal versus diurnal habitat use showed that roosting habitat requirements are clearly different to foraging habitat requirements for both species. Other observational field studies have suggested that both species prefer to roost in trees next to or surrounded by water, whether at rivers, streams, lakes, or reservoirs/dams (Carrick 1962 ; Marchant and Higgins 1990 ; McKilligan 1975 ; McKilligan 1979 ). The bimodal patterns in local and landscape water depth seen for these species in our modelling with MIM inundation data reflect this bi-modal split in habitat needs, with selection of sites adjacent to deep-water habitats for roosting, and selection of shallow-water habitats for foraging. The MIM modelling also indicated that these species are selecting for short-term water characteristics at relatively fine spatial scales, with water depth and inundation relationships being significant at pixel/point and 2km radii scales but not at the 20km radii scale. The implication of this split at local scales is that conservation or provision of foraging habitats without simultaneous consideration of available roosting habitats nearby is likely to be problematic for both species and may result in a reduced response by these species to management efforts. Provision of foraging habitats without suitable roosting habitats nearby may also result in increased energy use or wasted energy expenditure by birds that choose to travel from distant roosting habitats and may therefore not be optimal use of resources. The relative availability and modifications of habitats by agriculture and water resource use will also influence habitat use, most likely to a greater extent for straw-necked ibis than for royal spoonbills. This is because ibis tend to use more terrestrial (‘non-ANAE’) habitats than spoonbills, and this together with their generalist diet and adaptability makes them more able to use agricultural land uses for foraging than spoonbills. The regular cycles of shallow irrigation, tillage, and sometimes burning used in intensive agriculture flush prey such as crickets, spiders and frogs from soil and attract other prey types. Ibis are well known to flock to sites receiving water to take advantage of the temporary abundance of food. However these resources are typically temporary, and management support for foraging habitats may be particularly important in areas where irrigated agriculture and other water sources effectively ‘dry up’ seasonally, as well as in areas where the effects of climate change are severe (Perez-Moreno et al. 2016 ). Spoonbills are more dependent on permanent surface water habitats and marsh/wetland habitats that provide a range of aquatic food types than ibis species and are therefore more likely to be affected by changes in these habitats either through water regime change or due to other pressures or threats. Support for foraging habitats when agricultural or other temporary resources are not available, particularly during winter, is likely to be critical for juvenile survival for these and similar species (Jelena et al. 2012 ). Conclusion Beyond breeding sites, waterbirds need suitable habitat for critical processes within the life cycle. These include juvenile survival to breeding age (i.e. recruitment, which for some ibis and spoonbill species may take up to four years), adult recovery from breeding efforts, and adult survival between breeding events. Knowledge of habitat use during these periods is important for efficient and effective management and policy for waterbirds, particularly for species that are nomadic and use remote sites across wide geographic areas. Here, we characterised non-breeding habitat preferences for RSB and SNI and found clear intra-species differences. These differences have implications for management, and mapping of predicted habitat availability at basin-wide scales, providing context for prioritisation and application of resources. Such increased knowledge of the spatio-temporal interactions of waterbirds with their environment across complete life cycles is essential for informing management aimed at increasing waterbird numbers or maintaining diversity in the long term. Declarations Declaration of competing interests The authors declare no competing interests. Animal ethics statement All research protocols were approved by an authorized Animal Care and Ethics Committee, according to the Australian code of practice for the care and use of animals for scientific purposes. On-ground fieldwork activities were conducted under New South Wales and Victoria Scientific Licences 102180 and 10010534. Funding The original research that formed the basis of this article was co-funded by the Commonwealth Environmental Water Holder’s Office (CEWH/CEWO) and the Commonwealth Scientific and Industrial Research Organisation (CSIRO) through the CEWH Monitoring, Evaluation and Research project (2019–2024) and the CEWO Environmental Watering Knowledge and Research project (2015–2018), administered through the Commonwealth Environmental Water Office within the Department of Climate Change, Energy, the Environment and Water and its precursors. The research also benefited from co-investment by the Lake Cowal Conservation Centre, and from in-kind support from the Royal Botanic Garden Sydney (John Martin), NSW Department of Planning and Environment and its precursors, and the Goulburn-Broken Catchment Management Authority (Keith Ward). Author Contribution HM initiated and led the research and obtained funding. HM, FR, LON, SR, MP, MD, JH and JM conducted fieldwork and data collection. LLJ, AL, HM, FR, and JH processed the data for analysis. LLJ performed the analysis of the data with guidance from HM. HM and LLJ wrote the manuscript. RK, KB, VD and RM provided research direction and design advice at the beginning of the project. Acknowledgement The authors express their gratitude for the assistance of colleagues, collaborators and volunteers with fieldwork, and the support of program leaders. Data Availability Code is available via the CSIRO Data Access Portal, https://data.csiro.au/, including all scripts used in the analysis. This repository contains the necessary files required to run the scripts and recreate the analyses. Raw data are available from the corresponding author upon reasonable request and will be uploaded to the CSIRO Data Access Portal at the time of publication. References ABARES (2021) Catchment scale land use of australia – update December 2020. In '.' (Ed. ABARES): Canberra, Australia) Arthington, A.H., Bhaduri, A., Bunn, S.E., Jackson, S.E., Tharme, R.E., Tickner, D. , et al. (2018) The Brisbane Declaration and Global Action Agenda on Environmental Flows (2018). 6 (45). [In English] Benjamini, Y., and Yekutieli, D. (2001) The control of the false discovery rate in multiple testing under dependency. Annals of Statistics 29 , 1165-1188. Bino, G., Brandis, K., Kingsford, R.T., and Porter, J. (2021) Shifting Goalposts: Setting Restoration Targets for Waterbirds in the Murray-Darling Basin Under Climate Change. Frontiers in Environmental Science 9 . [In English] Bino, G., Kingsford, R.T., and Porter, J. (2015) Prioritizing Wetlands for Waterbirds in a Boom and Bust System: Waterbird Refugia and Breeding in the Murray-Darling Basin. Plos One 10 (7). Bino, G., Steinfeld, C., and Kingsford, R.T. (2014) Maximizing colonial waterbirds' breeding events using identified ecological thresholds. and environmental flow management. Ecological Applications 24 (1), 142-157. Brandis, K.J., Bino, G., Spencer, J.A., Ramp, D., and Kingsford, R.T. (2018) Decline in colonial waterbird breeding highlights loss of Ramsar wetland function. Biological Conservation 225 , 22-30. Brandis, K.J., Francis, R.J., and Bino, G. (2024) Vegetation and inundation characteristics of waterbird breeding sites in the Murray–Darling Basin, Australia. Marine and Freshwater Research 75 (7), -. Brandis, K.J., Kingsford, R.T., Ren, S., and Ramp, D. (2011) Crisis water management and ibis breeding at Narran Lakes in arid Australia. Environ Manage 48 (3), 489-98. Brooks, M.E., Kristensen, K., van Benthem, K.J., Magnusson, A., Berg, C.W., Nielsen, A. , et al. (2017) glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. The R journal 9 , 400. [In en] Brooks, S. (2021) ANAE Classification of the Murray-Darling Basin Technical Report, Revision 3.0, March 2021. Technical report for the Commonwealth Environmental Water Office, Department of Agriculture, Water and the Environment, Australia., Canberra, Australia. Caley, P., Reid, J.R.W., Colloff, M.J., and Barry, S.C. (2022) On inferring population trends of mobile waterbirds from aerial transect surveys in variable environments. Environmental and Ecological Statistics 29 (1), 3-31. Calvert, A.M., Walde, S.J., and Taylor, P.D. (2009) Nonbreeding-Season Drivers of Population Dynamics in Seasonal Migrants: Conservation Parallels Across Taxa. Avian Conservation and Ecology 4 (2), 5[online]. Carrick, R. (1962) Breeding, movements and conservation of Ibises (Threskiornithidae) in Australia. CSIRO Wildlife Research 7 (1), 71-88. Chiau, W.Y. (2006) Lagoon Management and Black-faced Spoonbill Conservation: Issues and Challenges in theGreater Chiku Area, Taiwan. Journal of Coastal Research 2 (Winter 2006) (Special Issue No. 39. Proceedings of the 8th International Coastal Symposium (ICS 2004)), 912-914. Davidson, N.C. (2014) How much wetland has the world lost? Long-term and recent trends in global wetland area. Marine and Freshwater Research 65 (10), 934-941. Fithian, W., and Hastie, T. (2013) Finite-sample equivalence in statistical models for presence-only data. The Annals of Applied Statistics 7 (4), 1917. Haig, S.M., Mehlman, D.W., and Oring, L.W. (1998) Avian Movements and Wetland Connectivity in Landscape Conservation. Conservation Biology 12 (4), 749-758. Hirzel, A.H., Le Lay, G., Helfer, V., Randin, C., and Guisan, A. (2006) Evaluating the ability of habitat suitability models to predict species presences. Ecological Modelling 199 (2), 142-152. Jelena, K., Antun, Z., Tibor, M., and Otto, O. (2012) Movements of Immature Eurasian Spoonbills Platalea leucorodia from the Breeding Grounds of the Eastern Metapopulation in the Pannonian Basin. Waterbirds 35 (2), 239-247. Jirinec, V., Rodrigues, P.F., and Amaral, B. (2021) Adjustable leg harness for attaching tags to small and medium-sized birds. Journal of Field Ornithology 92 (1), 77-87. Karl, B.J., and Clout, M.N. (1986) An improved radio transmitter harness with a weak link to prevent snagging. Journal of Field Ornithology 58 (1), 73-77. Kingsford, R., and Johnson, W. (1998) Impact of water diversions on colonially-nesting waterbirds in the Macquarie Marshes of arid Australia. Colonial Waterbirds 21 (2), 159-170. Kingsford, R., and Norman, F. (2002) Australian waterbirds—products of the continent's ecology. Emu 102 (1), 47-69. Kingsford, R.T. (2000) Ecological impacts of dams, water diversions and river management on floodplain wetlands in Australia. Austral Ecology 25 (2), 109-127. Kingsford, R.T., and Auld, K.M. (2005) Waterbird breeding and environmental flow management in the Macquarie Marshes, arid Australia. River Research and Applications 21 (2-3), 187-200. Kingsford, R.T., Roshier, D.A., and Porter, J.L. (2010) Australian waterbirds - time and space travellers in dynamic desert landscapes. Marine and Freshwater Research 61 (8), 875-884. Kirby, J.S., Stattersfield, A.J., Butchart, S.H.M., Evans, M.I., Grimmett, R.F.A., Jones, V.R. , et al. (2008) Key conservation issues for migratory land- and waterbird species on the world's major flyways. Bird Conservation International 18 (S1), S49-S73. Leslie, D.J. (2001) Effect of river management on colonially-nesting waterbirds in the Barmah-Millewa forest, south-eastern Australia Regulated Rivers: Research & Management 17 (1), 21-36. Lindsell, J.A., Serra, G., Peške, L., Abdullah, M.S., al Qaim, G., Kanani, A. , et al. (2009) Satellite tracking reveals the migration route and wintering area of the Middle East population of Critically Endangered northern bald ibis Geronticus eremita. Oryx 43 (3), 329-335. Marchant, S., and Higgins, P.J. (1990) 'Handbook of Australian, New Zealand and Antarctic Birds. Volume 1, Ratites to ducks.' (Oxford University Press, Melbourne) McCabe, J.D., Clare, J.D., Miller, T.A., Katzner, T.E., Cooper, J., Somershoe, S. , et al. (2021) Resource selection functions based on hierarchical generalized additive models provide new insights into individual animal variation and species distributions. Ecography 44 (12), 1756-1768. McKilligan, N. (1975) Breeding and movements of the straw-necked ibis in Australia. Emu 75 (4), 199-212. McKilligan, N. (1979) The ecology of the straw-necked ibis in winter at Toowoomba, South-east Queensland. Sunbird: Journal of the Queensland Ornithological Society, The 10 (3/4), 49. MDBA (2014) Basin-wide environmental watering strategy Murray–Darling Basin Authority for and on behalf of the Commonwealth of Australia, No. MDBA Publication No 20/14, Canberra, ACT, Australia. Muff, S., Signer, J., and Fieberg, J. (2020) Accounting for individual-specific variation in habitat-selection studies: Efficient estimation of mixed-effects models using Bayesian or frequentist computation. 89 (1), 80-92. Nicol, S., Lloyd-Jones, L., & McGinness, H. M. (2024). A method to predict connectivity for nomadic waterbird species from tracking data. Landscape Ecology , 39 (2), 1-15. Perez-Moreno, H., Martinez-Meyer, E., Mainero, J.S., and Rojas-Soto, O. (2016) Climatic patterns in the establishment of wintering areas by North American migratory birds. Ecology and Evolution 6 (7), 2022-2033. Peters, W., Hebblewhite, M., Cavedon, M., Pedrotti, L., Mustoni, A., Zibordi, F. , et al. (2015) Resource selection and connectivity reveal conservation challenges for reintroduced brown bears in the Italian Alps. Biological Conservation 186 , 123-133. Price, C.J., Morris, A., Staines, G., Payne, R., and Smith, J. (2018) Leaving home but nowhere to go: lessons learnt from almost two decades of Bush Stone-curlew Burhinus grallarius monitoring on the Central Coast of NSW. Australian Zoologist 39 (4), 769-783. R Core Team (2024) R: A language and environment for statistical computing. In '.' (R Foundation for Statistical Computing: Vienna, Austria) Reid, J.R.W., Colloff, M.J., Arthur, A.D., McGinness, H.M. (2013) Influence of catchment condition and water resource development on waterbird assemblages in the Murray-Darling Basin, Australia. Biological Conservation 165 , 25-34. Renner, I.W., Elith, J., Baddeley, A., Fithian, W., Hastie, T., Phillips, S.J. , et al. (2015) Point process models for presence-only analysis. 6 (4), 366-379. Robinson, F. and McGinness, H.M. (2024). Mortality causes in ibis and spoonbill (Threskiornithidae) species and life stages: a global review. Under Review. Roshier, D.A., and Asmus, M.W. (2009) Use of satellite telemetry on small-bodied waterfowl in Australia. Marine and Freshwater Research 60 (4), 299-305. Serra, G., Bruschini, C., Lindsell, J.A., Peske, L., and Kanani, A. (2011) Breeding range of the last eastern colony of Critically Endangered Northern Bald Ibis Geronticus eremita in the Syrian steppe: a threatened area. Bird Conservation International 21 (3), 284-295. Smith, A., and Munro, U. (2011) Local and regional movements of the Australian white ibis threskiornis molucca in eastern Australia. Corella 35 , 89-94. Teng, J., Penton, D., Ticehurst, C., Sengupta, A., Freebairn, A., Marvanek, S. , et al. (2023) Two-monthly Maximum Flood Water Depth Spatial Timeseries for the MDB v20 CSIRO, https://doi.org/10.25919/c5ab-h019. Thaxter, C.B., Ross-Smith, V.H., Clark, J.A., Clark, N.A., Conway, G.J., Marsh, M. , et al. (2014) A trial of three harness attachment methods and their suitability for long-term use on Lesser Black-backed Gulls and Great Skuas. Ringing & Migration 29 (2), 65-76. Wang, C., Liu, D.-P., Qing, B.-P., Ding, H.-H., Cut, Y.-Y., Ye, Y.-X. , et al. (2014) The Current Population and Distribution of Wild Crested Ibis Nipponia nippon. Chinese Journal of Zoology 49 (5), 666-671. Wood, S.N. (2017) 'Generalized Additive Models: An Introduction with R.' (Chapman and Hall/CRC) Additional Declarations No competing interests reported. Supplementary Files Waterbirdshabitatusesupplementaryinformation.docx Cite Share Download PDF Status: Published Journal Publication published 05 Nov, 2024 Read the published version in Landscape Ecology → Version 1 posted Editorial decision: Revision requested 24 Aug, 2024 Reviews received at journal 29 Jul, 2024 Reviews received at journal 27 Jul, 2024 Reviewers agreed at journal 19 Jul, 2024 Reviewers agreed at journal 17 Jul, 2024 Reviewers agreed at journal 06 Jul, 2024 Reviewers invited by journal 06 Jul, 2024 Editor assigned by journal 25 Jun, 2024 Submission checks completed at journal 25 Jun, 2024 First submitted to journal 23 Jun, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4626784","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":327912427,"identity":"ed60e071-785a-4011-9bf6-d6e4ed3d4ab6","order_by":0,"name":"Heather M. McGinness","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDElEQVRIie3QMWrDMBSA4WcM6fJMV0PBvoK8GAqmvopCQZOHhC4dOhgE7hUSegmPHToIBNGiAxgKbYqgWROydGorNWknO3jsoH8SD388SwA+3z+MCIDwcAjX4md0VgPQ0yTg7nApJuRAUIwkpcDjKD71vSOKb8zs9gVypQM5fyzKdGnM9u2pgPNFvy31KuMLfQO5rkAuNZu2zyyP6TuDuOsnpKMZjxoKeYczGTWSkotqAlRIgCHyutnx6JNai8SSL/tjOtw6kg5uQbul/iMiqDu0L2AJGSK6mj/giiLRzJHraavdXQTDTK8HXky1e7yjCVHS7KPmqkzvpdl9iCJJVP+W33DExOfz+Xyj+wY6WWpA0UJWVgAAAABJRU5ErkJggg==","orcid":"","institution":"CSIRO Environment","correspondingAuthor":true,"prefix":"","firstName":"Heather","middleName":"M.","lastName":"McGinness","suffix":""},{"id":327912429,"identity":"a41eb1b3-931e-4c7b-b403-dcbfd0c132be","order_by":1,"name":"Luke R. Lloyd-Jones","email":"","orcid":"","institution":"CSIRO Data61","correspondingAuthor":false,"prefix":"","firstName":"Luke","middleName":"R.","lastName":"Lloyd-Jones","suffix":""},{"id":327912434,"identity":"fd6123e3-fd65-48eb-8b61-a6cf9d9f97cf","order_by":2,"name":"Freya Robinson","email":"","orcid":"","institution":"CSIRO Health and Biosecurity","correspondingAuthor":false,"prefix":"","firstName":"Freya","middleName":"","lastName":"Robinson","suffix":""},{"id":327912435,"identity":"ff895924-0ad5-45cf-91c5-26234246651c","order_by":3,"name":"Art Langston","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Art","middleName":"","lastName":"Langston","suffix":""},{"id":327912437,"identity":"82575661-2047-4f1a-8565-de715c1b16e2","order_by":4,"name":"Louis G. O’Neill","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Louis","middleName":"G.","lastName":"O’Neill","suffix":""},{"id":327912438,"identity":"97ac0e59-c4f5-4938-b08b-0e6e69c6b980","order_by":5,"name":"Shoshana Rapley","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Shoshana","middleName":"","lastName":"Rapley","suffix":""},{"id":327912441,"identity":"d41630ce-176e-4a20-ae08-a9ddb2c5f5c8","order_by":6,"name":"Micha V. Jackson","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Micha","middleName":"V.","lastName":"Jackson","suffix":""},{"id":327912442,"identity":"519bfd69-e074-4340-9cbf-6ad9d241e7eb","order_by":7,"name":"Jessica Hodgson","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Jessica","middleName":"","lastName":"Hodgson","suffix":""},{"id":327912444,"identity":"22fcdee3-357b-4949-948a-5bb2a792b3f2","order_by":8,"name":"Melissa Piper","email":"","orcid":"","institution":"CSIRO Agriculture","correspondingAuthor":false,"prefix":"","firstName":"Melissa","middleName":"","lastName":"Piper","suffix":""},{"id":327912446,"identity":"627334e3-35a8-423d-ba6a-ec23447d410a","order_by":9,"name":"Micah Davies","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Micah","middleName":"","lastName":"Davies","suffix":""},{"id":327912447,"identity":"ede02b5b-ab56-4e2e-872c-c22c64e79fe0","order_by":10,"name":"John M. Martin","email":"","orcid":"","institution":"Royal Botanic Gardens Sydney","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"M.","lastName":"Martin","suffix":""},{"id":327912449,"identity":"dedd6408-f781-454d-bca5-5add96ef0157","order_by":11,"name":"Richard Kingsford","email":"","orcid":"","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Richard","middleName":"","lastName":"Kingsford","suffix":""},{"id":327912451,"identity":"2dab0139-b202-4f61-900e-d390405d6ba9","order_by":12,"name":"Kate Brandis","email":"","orcid":"","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Kate","middleName":"","lastName":"Brandis","suffix":""},{"id":327912452,"identity":"ab11cb56-b6a5-4817-85a4-b3620110d99a","order_by":13,"name":"Veronica Doerr","email":"","orcid":"","institution":"CSIRO Environment","correspondingAuthor":false,"prefix":"","firstName":"Veronica","middleName":"","lastName":"Doerr","suffix":""},{"id":327912455,"identity":"0f68b9a3-1bf9-48ad-8eb9-34188a157bf7","order_by":14,"name":"Ralph Mac Nally","email":"","orcid":"","institution":"University of Canberra","correspondingAuthor":false,"prefix":"","firstName":"Ralph","middleName":"Mac","lastName":"Nally","suffix":""}],"badges":[],"createdAt":"2024-06-24 01:18:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4626784/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4626784/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10980-024-01982-1","type":"published","date":"2024-11-05T15:57:01+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":60630555,"identity":"7d680b69-7382-4ccb-a807-2ad92d7628a6","added_by":"auto","created_at":"2024-07-19 00:46:11","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":160787,"visible":true,"origin":"","legend":"\u003cp\u003eBreeding sites at which straw-necked ibis and royal spoonbills were fitted with GPS satellite tracking transmitters, Australia.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/8dee1fcaa78343096d10b2d5.png"},{"id":60630561,"identity":"07ee2c3d-2526-420e-9504-fe0919797661","added_by":"auto","created_at":"2024-07-19 00:46:12","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":11134063,"visible":true,"origin":"","legend":"\u003cp\u003eMap of eastern Australian showing raw telemetry data for straw-necked ibis (SNI; purple; n = 89) and royal spoonbill (RSB; pink; n = 52). Background is a true colour Sentinel image. Inset shows location within the Australian continent. Grey internal outline shows the Murray-Darling Basin boundary.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/e5cddb2022f4a99718b6fe40.png"},{"id":60630558,"identity":"2ce40c1a-f8e3-4790-b27f-599dc397460e","added_by":"auto","created_at":"2024-07-19 00:46:11","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2699423,"visible":true,"origin":"","legend":"\u003cp\u003ePredicted spatially explicit habitat for straw-necked ibis within the Murray-Darling Basin, with lighter colours indicating higher suitability (weighted averages). Left: without satellite tracking data; Right: with satellite tracking data (gold).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/b82de4b1c9e0ddb16c481ff6.png"},{"id":60630559,"identity":"a1fa728d-d183-4bbe-8351-c9c976abbb92","added_by":"auto","created_at":"2024-07-19 00:46:11","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":800759,"visible":true,"origin":"","legend":"\u003cp\u003eSummary of predicted relative preference from a quadratic model of Water Observations from Space at population levels for straw-necked ibis (SNI) and royal spoonbill (RBS). Predictions included a global intercept estimate holding all other variables at zero.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/4ba966852d4c1bae61e5387a.png"},{"id":60630980,"identity":"6d3424f6-352a-46e3-a157-367fd975bdbc","added_by":"auto","created_at":"2024-07-19 00:54:12","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":800273,"visible":true,"origin":"","legend":"\u003cp\u003eModelled habitat selection coefficients for satellite-tracked straw-necked ibis from random slope and intercept by individual models. Bar width indicates effect size and whisker lines the 95% interval on the effect.\u003c/p\u003e\n\u003cp\u003eBold whisker lines indicate covariate levels significance after multiple testing correction.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/f6b2cdcc86240d168337253d.png"},{"id":60630557,"identity":"222673db-baf7-4e7c-b1c5-4421c4a8ab2f","added_by":"auto","created_at":"2024-07-19 00:46:11","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2083630,"visible":true,"origin":"","legend":"\u003cp\u003ePredicted spatially explicit habitat for royal spoonbill within the Murray-Darling Basin, with lighter colours indicating higher suitability (weighted averages). Left: without satellite tracking data; Right: with satellite tracking data (gold).\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/edeacc43a8a57baa14b8f999.png"},{"id":60630981,"identity":"27603f51-7e09-4190-b5d5-5df32bf00ed5","added_by":"auto","created_at":"2024-07-19 00:54:12","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":696500,"visible":true,"origin":"","legend":"\u003cp\u003eModelled habitat selection coefficients for satellite-tracked straw-necked ibis from random slope and intercept by individual models. Bar width indicates effect size and whisker lines the 95% interval on the effect. Bold whisker lines indicate covariate levels significance after multiple testing correction.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/693ea86cf947bba16f51edeb.png"},{"id":68749957,"identity":"65012ad7-7c94-4b21-baac-4b1d786646e1","added_by":"auto","created_at":"2024-11-11 16:08:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":19466782,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/fef43e59-3459-4423-a011-72521ea97965.pdf"},{"id":60630563,"identity":"77562508-8a2d-4474-a069-68ac75b8b2a6","added_by":"auto","created_at":"2024-07-19 00:46:12","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":20430735,"visible":true,"origin":"","legend":"","description":"","filename":"Waterbirdshabitatusesupplementaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-4626784/v1/62971d3d9365db5d204457f9.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Habitat use by nomadic ibis and spoonbills post-dispersal from breeding sites","fulltext":[{"header":"Introduction","content":"\u003cp\u003eUnderstanding the habitat use of a species across its life cycle is essential for effective management. Without this holistic knowledge, conservation efforts may prove ineffective in the long-term despite successful protection of habitat associated with an individual aspect of the species\u0026rsquo; life cycle\u0026nbsp;(Haig\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 1998). For example, conserving a species\u0026rsquo; breeding habitat might not increase recruitment into the population if there is post-natal dispersal into an unprotected area with high mortality\u0026nbsp;(Price\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2018). This is particularly critical for species reliant on habitats that are susceptible to loss or alteration. One such group of species is waterbirds, which are highly reliant on wetlands that are being lost at increasing rates due to floodplain development, water abstraction, pollution and climate change\u0026nbsp;(Davidson 2014; Haig\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 1998; Kirby\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2008), leading to international agreements to protect them (e.g. Ramsar Convention on Wetlands). In Australia, waterbirds have declined significantly since European colonisation\u0026nbsp;(Bino\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2021; Bino\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2015; Leslie 2001)\u0026nbsp;and addressing these declines is a specified management goal for large management areas, for example the Murray Darling Basin\u0026nbsp;(MDBA 2014).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSupporting or recovering waterbird populations requires intensive conservation management because different species have different degrees of dependency on flooding to provide habitat; for example, aggregate-nesting wading waterbirds require extensive areas of shallow surface water for both foraging and breeding\u0026nbsp;(Brandis\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2018; Chiau 2006; Lindsell\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2009; Serra\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2011; Wang\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2014). This is a particular challenge in inland areas of Australia that are naturally water poor but are subject to extensive agricultural water abstraction and regulation together with increasing drought frequency, duration and severity. Management can include the allocation of environmental water or \u0026lsquo;environmental flows\u0026rsquo;, to provide the quantity, timing, and quality of freshwater flows and water levels necessary\u0026nbsp;(Arthington\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2018). Efficient application of environmental water by managers, including where to water, when to water, and for how long, requires knowledge of species habitat use and movements\u0026nbsp;(Reid 2013). Many waterbird species are highly vagile, displaying a range of movement strategies at varying spatial and temporal scales across their lifecycle\u0026nbsp;(Kingsford\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2010). However, understanding of waterbird habitat use across all stages of the lifecycle is often insufficient, particularly in remote inland areas, which hinders strategic management.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLarge aggregate-nesting wading waterbird breeding events and locations in Australia have been relatively well studied and the focus of significant management effort, due to the national and international importance of associated sites, their susceptibility to adverse effects of environmental change, and the large aggregations of waterbirds of multiple species that they attract ((Bino\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2015; Bino\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2014; Brandis\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2018; Brandis\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2024; Brandis\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2011; Kingsford and Johnson 1998; Kingsford 2000; Kingsford and Auld 2005). Comparatively, far less is understood about habitat or site use of these species after dispersal or between breeding events\u0026nbsp;(Caley\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2022); yet critical population processes occur outside of breeding events with the potential to shape demography and population viability\u0026nbsp;(Calvert\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2009; Carrick 1962). For waterbirds in the Threskiornithidae family, it can take 3-4 years for juveniles to mature and recruit into the adult breeding population, during which time high rates of attrition occur (Robinson and McGinness 2024). Likewise, adults disperse until the next breeding event; events which are irregular and can be years apart. These life stages involve movements at scales of hundreds to thousands of kilometres, frequently in remote areas\u0026nbsp;(Kingsford and Norman 2002, Nicol et al. 2024). While banding studies have been valuable sources of information on waterbird movement, major knowledge gaps remain, especially concerning movement and habitat use in remote areas\u0026nbsp;(Smith and Munro 2011). In this context, continuous GPS telemetry monitoring movements and habitat use over multiple years and individuals provides invaluable data for quantifying habitat selection across the full life cycle. These data can then inform where and when management actions and resources might be best targeted to improve habitat conditions and therefore long-term waterbird population outcomes.\u003c/p\u003e\n\u003cp\u003eHere we use a GPS satellite telemetry dataset of 141 individuals spanning seven years to quantify non-breeding habitat requirements for two wading waterbird species of the Threskiornithidae family: straw-necked ibis \u003cem\u003eThreskiornis spinicollis\u003c/em\u003e (SNI) and royal spoonbill \u003cem\u003ePlatalea regia\u003c/em\u003e (RSB). We hypothesised that topography, land-use, vegetation, and water indices would predict habitat selection, and that habitat selection and habitat availability would differ between SNI and RSB.\u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eTransmitter deployment\u003c/h2\u003e\n\u003cp\u003eWe tracked 141 birds (SNI = 89, RSB = 52) using satellite telemetry with tracking durations ranging from 30 \u0026ndash; 2,000 days per bird and spanning October 2016 \u0026ndash; January 2024. Individuals were captured and fitted with transmitters in eight wetlands of the Murray-Darling Basin (MDB), which spans approximately 14% of Australia\u0026rsquo;s landmass and is subject to intensive water and wetland management (Figure 1). We fitted birds with solar-powered GPS transmitter units from three sources: Druid (Druid Technology Co., Ltd., Chengdu, China), Geotrak (Geotrak Inc. North Carolina, USA), or Ornitela (Ornitela, UAB Vilnius. Lithuania). Transmitters weighed 12\u0026ndash;40 g and ranged from \u0026lt; 1% to 5% of bird bodyweight. GPS fix resolution was 15\u0026ndash;26 m and fix frequency ranged from one minute to 6 hours (depending on transmitter type and programmed duty cycle). Fix frequency was considered in analyses, with interpolation or down-scaling applied when appropriate. The data were transmitted via either the Argos satellite network (for Geotrak units) or the 3G network (for Ornitela and Druid units).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe captured birds either by hand, or with leg-nooses, or by using a net launcher. We attached transmitters as a \u0026lsquo;backpack\u0026rsquo; using harnesses made of Teflon ribbon or Spectra ribbon (Bally Ribbon Mills\u003csup\u003eTM\u003c/sup\u003e), fitted either as wing-loops with a join at the keel (SNI and some RSB), or as leg-loops (RSB). Harness design was based on designs used in other species\u0026nbsp;(Jirinec\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2021; Karl and Clout 1986; Roshier and Asmus 2009; Thaxter\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2014) and modified and improved over time.\u003c/p\u003e\n\u003ch2\u003eData pre-processing\u003c/h2\u003e\n\u003cp\u003eData were filtered based on dilution of precision, removal of outliers from visual inspection of tracks, and removal of points based on extreme outliers on distance travelled per unit time and precise vertical or horizontal movements. To reduce temporal autocorrelation and produce a standardised temporal fix frequency and timing, we interpolated and thinned the data to one fix every six hours. This reduced the initial data set of ~ 1.6 million presence points to ~ 123,000, with movements spanning ~ 4 million km\u003csup\u003e2\u003c/sup\u003e in eastern Australia (Figure 2). Most of the fix removals were for individuals tracked at very high frequency with Druid units. We removed data tracking movements from both nesting birds and points from pre-dispersal from capture site, since this study focused on habitat selection between breeding events, not during nesting or associated with capture sites (frequently nest sites for adults). This left 109,034 presence points across the two species.\u003c/p\u003e\n\u003ch3\u003eBackground data\u003c/h3\u003e\n\u003cp\u003eWe conducted data analysis and modelling using R Statistical Software version 4.4.0\u0026nbsp;(R Core Team 2024). To compare habitat selection as indicated by species presence (hereafter \u0026lsquo;presence\u0026rsquo; data) to background habitat availability (i,e., pseudo-absence; hereafter \u0026lsquo;background\u0026rsquo; data), we generated utilisation distribution grids for each individual. We fitted the \u0026lsquo;kernelUD\u0026rsquo; function in the \u003cem\u003eadehabitatHR\u003c/em\u003e package to create utilisation distribution from all presence points for each individual\u0026nbsp;(Calenge \u0026amp; Fortmann-Roe, 2023).\u0026nbsp;\u0026nbsp;We then used the \u0026lsquo;getverticeshr\u0026rsquo; function to extract the 99% contour of the utilisation distribution and sampled points on a regular grid such that the ratio of background points to presence points was approximately 10:1. The total dataset after initial quality control consisted of 84,425/ 900,149 and 24,609/ 284,909 presence/background locations for SNI and RSB respectively.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe annotated presence and background data with several environmental covariates, including: Water Observations from Space (WOfS;\u0026nbsp;(GEOSCIENCE AUSTRALIA, 2015; Mueller et al., 2016); Multi-Resolution Valley Bottom Flatness (MrVBF;\u0026nbsp;Gallant et al., 2012);\u0026nbsp;the Australian National Aquatic Ecosystem classification (ANAE;\u0026nbsp;(Brooks 2021)); the Catchment Scale Land Use of Australia (CLUM;\u0026nbsp;ABARES (2021)); and,\u0026nbsp;the\u0026nbsp;National Vegetation Information System (NVIS;\u0026nbsp;NVIS technical Working Group, 2013). These datasets are extensively described in their respective sources, but briefly: WOfS summarises 30 years of Landsat imagery to provide a long-term understanding of the recurrence of water in the landscape (25 m resolution); MrVBF is a topographic index that uses digital elevation models to classify degrees of valley bottom flatness (30 m resolution); the ANAE\u0026nbsp;integrates the best available mapping data from multiple sources combined with simple rules to define aquatic ecosystem types using a number of relevant attributes (e.g. water regime, water source, salinity, landform and dominant vegetation);\u0026nbsp;CLUM summarises a single dominant land use for a defined area based on the objective of the land manager as identified by state and territory agencies and classified according to a\u0026nbsp;three-tiered hierarchical structure based on potential degree of modification and the impact on a putative \u0026lsquo;natural state\u0026rsquo;\u0026nbsp;(50 m resolution);\u0026nbsp;and,\u0026nbsp;NVIS summarises the extent and distribution of vegetation types in the landscape (100 m resolution). For WOfS, we used the recurrence frequency of water as a percentage of the total number of clear observations for each grid cell. We rescaled WOfS to have a mean of 0 and a variance of 1 to assist with model convergence.\u003c/p\u003e\n\u003cp\u003eTo improve computational efficiency and facilitate interpretation, we chose the highest level of classification for each of the multilevel covariates which left 9 levels for MrVBF, 32 levels for CLUM and 30 levels for NVIS. To retain as many complete cases as possible, we imputed a small proportion (~1%) of points that reported no value for each covariate with either a new dummy variable (for MrVBF) or the unknown/no data variables for NVIS and CLUM. A summary of proportions of presence and background points stratified by the levels of each of the variables was initially performed as reference for model outputs.\u003c/p\u003e\n\u003ch3\u003eHabitat selection modelling\u003c/h3\u003e\n\u003cp\u003eFor each species, we fitted an infinitely weighted logistic regression mixed model (IWGLMM) to model habitat selection with the environmental covariates as predictors using presence and background data with the \u003cem\u003eglmmTMB\u003c/em\u003e package\u0026nbsp;(Brooks\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2017). IWGLMM performs similarly to inhomogeneous Poisson point process and maximum entropy (maxent) for habitat selection modelling\u0026nbsp;(Fithian and Hastie 2013).\u0026nbsp;WOfS was continuous\u0026nbsp;while the other four predictors (MrVBF, ANAE, CLUM, and NVIS) were categorical.\u0026nbsp;Quadratic terms (fixed and random) were included for WOfS to model potential nonlinearity; linear terms were used for the categorical variables.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eModels were also explored within species split by age (juvenile or adult), season (summer = December \u0026ndash; February, autumn = March \u0026ndash; May, winter = June \u0026ndash; August, spring = September \u0026ndash; November), and time of day (diurnal = points from 12pm; nocturnal = points from 12am) fitted by partitioning the data by these variables. Night fixes for SNI represented roosting locations because they are known to roost in one place from sunset to sunrise; conversely, RSB are less predictable, often travelling or feeding at night, and either feeding or resting during the day. We used a generalized linear mixed model (GLMM) framework with random intercepts and random slopes stratified by individual to account for between individual variation and reduce pseudo-replication. We fixed the variance at 10\u003csup\u003e3\u003c/sup\u003e rather than shrinking individual-specific intercepts towards the overall mean, following\u0026nbsp;Muff\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e (2020). We set the weight of background points as \u003cem\u003eW\u0026nbsp;\u003c/em\u003e=\u0026nbsp;10\u003csup\u003e5\u003c/sup\u003e, in line with previous IWGLMM applications\u0026nbsp;(Muff\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2020; Renner\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2015).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eInitial analyses showed poor convergence when fitting all covariates in a single model due to complete or near-complete separation of factor levels and highly correlated variables. Highly correlated between-covariate (e.g. WOfS and water land use variables) water-based variables were also difficult to interpret in a single model. To overcome this numerical instability, we created a \u0026lsquo;low count\u0026rsquo; factor level for all factor variables if at least one cell count of the two-way table of the factor level versus presence-availability status was less than 50. To further improve numerical stability and aid interpretation, we fitted each of the environmental covariates separately.\u003c/p\u003e\n\u003cp\u003eFor each species fitted with the IWGLMM we summarised fixed and random parameters and performed multiple comparison correction for fixed effect parameters following\u0026nbsp;Benjamini (2001).\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003eHabitat predictions\u003c/h3\u003e\n\u003cp\u003eWe constructed maps of relative probability of occurrence using a regular grid of points for the MDB (long. min = 138.57, long. max = 152.49, lat. min -37.68, lat. max = -24.59) at ~ 500 m resolution (~ 4 million total pixels). We constrained the predictions to the MDB because waterbird capture sites were targeted to this area and results are likely more representative of preferences for birds originating from this area. We matched environmental covariates to these points and used the \u0026lsquo;predict.glmmTMB\u0026rsquo; from the \u003cem\u003eglmmTMB\u0026nbsp;\u003c/em\u003epackage function along with the fitted GLMM models to produce population-level predictions (i.e., setting all random effects to zero). Model predictions were made separately for each of the environmental covariates (WOfS, MRVBF, ANAE, CLUM, NVIS) and combined into a single prediction using a weighted average of predictions, with the weights taken as the squared correlation Spearman\u0026apos;s rank from five-fold cross-validation.\u003c/p\u003e\n\u003ch3\u003eModel validation\u003c/h3\u003e\n\u003cp\u003eWe compared predictive performance of the fitted model using five-fold cross-validation to assess out-of-sample predictive error and model weights of each of the fitted models. Each fold consisted of an approximately equal number of individual birds. We trained models iteratively on four of five folds and made predictions with the fifth. We investigated overlap of presence points with the predicted relative habitat preference at a population scale by binning the predicted relative occurrence output over all presence absence points into deciles and computing the frequency of observed presence points within each decile. We then estimated the Spearman\u0026apos;s rank correlation coefficient between frequency of the validation presence points within each bin and the bin rank (1-low to 10-high use), also referred to as the \u0026lsquo;Boyce Index\u0026rsquo;\u0026nbsp;(Hirzel\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2006). A strong positive correlation indicates the model assigns presence points well to preferred habitat\u0026nbsp;(McCabe\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2021; Peters\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2015).\u003c/p\u003e\n\u003ch3\u003eBi-monthly surface water relationships\u003c/h3\u003e\n\u003cp\u003eTo explore potential relationships between short-term surface water presence and habitat selection, we intersected telemetry points with the Multi Index Method (MIM) surface water dataset for the MDB. This dataset represents a spatial time series of estimated water depth (in mm, error \u0026plusmn; 500 mm) and maximum surface water extent at a single point in time every two months, based on multi-index surface water extents, a digital elevation model, and gauged water depths\u0026nbsp;(Teng\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e 2023).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter matching the six-hourly gridded telemetry data used for the prior analyses to the spatial and temporal boundaries of the MIM dataset,\u0026nbsp;64,220 telemetry points remained for analysis from 97 individuals (SNI, N = 60; RSB = 30). Given that the MIM data are time varying, we generated control points differently to the static variables used in the static habitat selection analyses. We sampled points uniformly at a ratio of 5:1 in a circular buffer around each point with a radius equal to the distance to the next point. For each bird, we intersected both observed and control telemetry points with the MIM datasets. For each point, we produced water presence/absence and depth (mm) datasets. We also produced water depth summary statistics (mean, median, max) and calculated the proportion of wet points for 2 km and 20 km radii around each point. We then used the \u003cem\u003emgcv\u003c/em\u003e package (Wood 2017) to produce generalised additive models (GAMs) that were used to explore relationships between tracked bird presence and: 1) water presence/absence and depth at the point/pixel; 2) water depth statistics within a) 2km and b) 20 km radii around each point; and, 3) the proportion of points wet within a) 2km and b) 20 km radii around each point.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eHabitat selection and prediction \u0026ndash; straw-necked ibis\u003c/h2\u003e \u003cp\u003eThe spatially explicit SNI habitat preference predictions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) were highly robust, with strong discrimination on held-out data and a cross-validated Spearman rank correlation of 0.86 from the weighted predictor. We also found a non-linear relationship between SNI habitat preference and WOfS (i.e., the frequency of surface water) with an optimal range of 15\u0026ndash;20% surface water presence frequency over time, and a substantial decrease in suitability for values above 30%, where surface water was present more frequently (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). We found positive and negative associations with the categorical environmental variables (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Supplementary Table\u0026nbsp;1 and Supplementary Figs.\u0026nbsp;1\u0026ndash;4). We note that predictive performance is MDB specific whereas habitat preferences are for the whole of the data set, which included information from outside the MDB.\u003c/p\u003e \u003cp\u003eFor the MRVBF topographic index classes we found a strong positive association with large depositional basins and a negative association with small hillside deposits. Positive associations with ANAE aquatic ecosystem classes included permanent and temporary lowland streams, woodland riparian zone or floodplain, temporary tall emergent marsh, clay pan and temporary shrub swamps while there was a strong negative association with permanent forb marsh (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Supplementary Table\u0026nbsp;1). A wide range of CLUM catchment-scale land use classes were associated with SNI, including river, reservoir/dam, intensive animal production, marsh/wetland, residential and farm infrastructure, services, grazing irrigated modified pastures, grazing native vegetation, channel/aqueduct, irrigated cropping, transport and communication, and grazing modified pastures. There was also a wide range of NVIS vegetation classes positively associated with SNI, including inland aquatic areas (freshwater, salt lakes, lagoons), eucalypt woodlands, cleared non-native vegetation and buildings, grasslands, herblands, sedgelands and rushlands, eucalypt open forests and open woodlands, shrublands, including chenopod shrublands, samphire shrublands and forblands, modified native vegetation, and tussock grasslands. Habitat associations were similar between adult and juvenile SNI with few categories of preferred vegetation for juveniles and a preference for narrow valley floors that adults do not share (Supplementary Figs.\u0026nbsp;5 and 6).\u003c/p\u003e \u003cp\u003eWe found nocturnal roosting habitats differed from diurnal foraging habitats. Nocturnal roosting habitats were positively associated with river, stream, wetland, channel and reservoir environments, usually with trees; whereas diurnal foraging habitats usually occurred in more open areas, both natural and agricultural (Supplementary Figs.\u0026nbsp;7 and 8).\u003c/p\u003e \u003cp\u003eWe found seasonal differences in SNI habitat preferences in relation to MRVBF topographic values and ANAE aquatic ecosystem and NVIS vegetation classes, but CLUM land use class preferences were similar year-round (Supplementary Figs.\u0026nbsp;9\u0026ndash;12). SNI showed a distinct preference for large depositional basins in summer (and to some extent spring), but a wider range of MRVBF zones used in autumn and winter, including valley floor environments and small depositional basins. There was more use of non-ANAE (i.e., terrestrial) areas in autumn compared to other seasons. Fewer ANAE classes were used in winter, with a preference for permanent and temporary lowland streams and woodland riparian zones or floodplains. The range of NVIS classes used in summer was greater than in all other seasons, with a greater preference for shrublands. NVIS class preferences were similar in autumn and winter, with SNI favouring eucalypt woodlands, cleared non-native vegetation and buildings, inland aquatic habitats (freshwater, salt lakes, lagoons), grasslands, herblands, sedgelands and rushlands, and eucalypt open woodlands and forests. A subset of these was preferred in spring, including eucalypt woodlands, cleared non-native vegetation and buildings, and inland aquatic habitats (freshwater, salt lakes, lagoons).\u003c/p\u003e\u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eHabitat selection and prediction \u0026ndash; royal spoonbill\u003c/h2\u003e \u003cp\u003eThe spatially explicit RSB habitat preference predictions were also highly robust, with predictions showing strong discrimination on held-out data with a cross-validated Spearman rank correlation of 0.95 from the weighted predictor (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). There was a non-linear relationship between RSB habitat preference and WOfS, with a higher optimal frequency of surface water presence than SNI of 20\u0026ndash;45%, and a decrease in suitability for areas with water presence frequencies above 60% (representing semi-permanent and permanent water; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs for SNI, we found evidence for positive and negative associations for RSB with the categorical environmental variables (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, Supplementary Table\u0026nbsp;2 and Supplementary Figs.\u0026nbsp;13\u0026ndash;16). For the MRVBF topographic index, we found a positive association with depositional basins. Positive associations were observed with the following ANAE classes: permanent wetland, permanent lowland stream, temporary lowland stream, temporary tall emergent marsh and temporary shrub swamp. We found RSB selected for a range of CLUM (i.e., land uses) and NVIS (i.e., vegetation) classes; however, this range was smaller for RSB compared with SNI. Preferred CLUM classes were reservoir/dam, marsh/wetland, rivers, lakes, nature conservation and grazing native vegetation. Preferred NVIS classes were eucalypt tall open forests, inland aquatic (freshwater, salt lakes, lagoons), and eucalypt open forests.\u003c/p\u003e \u003cp\u003eWe found circadian differences in RSB habitat preferences between the ANAE classes (Supplementary Figs.\u0026nbsp;17 and 18). At night, RSB preferred temporary tall emergent marsh, whereas diurnal preferences were for permanent lowland stream and temporary lowland stream habitats. We did not find any differences between day and night habitat selection for the other environmental covariates. There were insufficient data for age group or seasonal comparisons for RSB.\u003c/p\u003e\u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eBi-monthly surface water relationships\u003c/h2\u003e \u003cp\u003eThere were significant relationships between tracked bird presence and average water depth at both the point/pixel and 2km radii scales for both species (Supplementary Tables\u0026nbsp;3\u0026ndash;6). There were also relationships between the proportion of water in the landscape and species presence at the 2km radii scale. However, at the 20 km scale, there were no relationships between tracked bird presence and either water depth statistics or the proportion of points wet (Supplementary Tables\u0026nbsp;7\u0026ndash;8).\u003c/p\u003e \u003cp\u003eAt point/pixel scale, the GAM smooth model suggested that SNI most preferentially selected sites with water depths of 12\u0026ndash;14 m (Supplementary Fig.\u0026nbsp;19). However, there were relatively few data points in this range and ~\u0026thinsp;85% of observations were at depths between 0-0.5 m (Supplementary Fig.\u0026nbsp;20), so the size of this preference should be interpreted with caution. When split by age group, this \u0026lsquo;deep water\u0026rsquo; model peak appeared to be strongly related to the juvenile SNI model, with adult models indicating bimodal preferences for water depths of ~\u0026thinsp;2 m and ~\u0026thinsp;7.5m (Supplementary Fig.\u0026nbsp;19).\u003c/p\u003e \u003cp\u003eRSB GAM smooth models at the point/pixel scale were all strongly bimodal, suggesting distinct preferences for locations with shallow water\u0026thinsp;~\u0026thinsp;1 m deep, as well as locations with water\u0026thinsp;~\u0026thinsp;5 m deep (Supplementary Fig.\u0026nbsp;19). Similar to SNI, ~ 85% of observations were actually at depths 0-0.5 m, with relatively few observations at greater depths. Overall model patterns were also strongly influenced by juvenile data, with the second peak in preferences for adults at ~\u0026thinsp;4 m rather than at 5 m as seen for juveniles (Supplementary Fig.\u0026nbsp;21).\u003c/p\u003e \u003cp\u003eAt the 2 km radii landscape scale, models suggested that SNI presence was greatest where water depths averaged 0.5\u0026ndash;2.5 m, with a rapid decline in presence as average depths increased beyond this average (Supplementary Fig.\u0026nbsp;22). SNI presence was also high where water depths were zero. When split by age group, the pattern was similar for SNI adults, however models suggested juvenile presence peaked where water depths were \u0026lt;\u0026thinsp;0.5 m and declined rapidly beyond this depth. The proportion of the landscape inundated within the 2km radii was also significant (Supplementary Fig.\u0026nbsp;22), with SNI presence highest where 50\u0026ndash;75% and 10\u0026ndash;20% of the landscape was inundated and declining after \u0026gt;\u0026thinsp;75% of the area was inundated (Supplementary Fig.\u0026nbsp;22). Again, SNI presence was also high in areas with no inundation. Adult patterns for SNI were similar to overall patterns, but juvenile presence was highest at 50\u0026ndash;70% water coverage and declined after \u0026gt;\u0026thinsp;60% of the area was inundated (Supplementary Fig.\u0026nbsp;22).\u003c/p\u003e \u003cp\u003eAt the 2km radii scale, RSB presence peaked slightly at shallow average depths of 0.5\u0026ndash;1 m but was consistent at a range of average depths up to 2.5 m. In contrast to SNI, RSB presence was very low where the proportion of the area inundated was low, and after peaking at 25% of the 2 km radius inundated, RSB presence remained relatively high with increasing area inundated all the way to 100% inundation (Supplementary Fig.\u0026nbsp;23).\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eEffective management and policy to support waterbirds across broad geographic areas requires a comprehensive understanding of habitat requirements, including seasonal change in preferences. This has historically been difficult to quantify for nomadic species, particularly those in remote areas and outside of known breeding locations when individuals disperse hundreds to thousands of kilometres. Advances in satellite tracking have enabled data collection over vast distances and in remote areas. Consequently, we have been able to characterise non-breeding habitat preferences for two such nomadic waterbirds. This knowledge has potential to inform identification and prioritisation of potential sites for environmental watering and other management, to support a) juvenile survival to breeding age (recruitment), b) adult recovery from breeding and c) adult survival between breeding events.\u003c/p\u003e \u003cp\u003eWhile ibis and spoonbill species in Australia share breeding sites and nest in mixed aggregations, their non-breeding habitats differ. RSB have primarily aquatic diets and adaptations focused on feeding in surface water (\u0026lsquo;obligate wetland foragers\u0026rsquo;) whereas SNI have mixed terrestrial and aquatic diets and foraging strategies (\u0026lsquo;non-obligate wetland foragers\u0026rsquo;). We found strong evidence that these species consequently require different foraging-habitat provision and management. RSB habitat use is more strongly related to both long-term and short-term inundation frequency, extent and depth than SNI habitat use. RSB displayed strong preferences for reservoirs, marshes and permanent wetlands, while SNI used both aquatic and terrestrial habitat, including areas of intensive animal production, modified pasture, and woodlands. Furthermore, we found RSB use fewer and less available habitat types compared to SNI. This leaves RSB more vulnerable to habitat change and loss and increases the relative importance of management for this and similar species such as egrets, where water resource developments, climate change impacts or other threats are present.\u003c/p\u003e \u003cp\u003eOur analysis of nocturnal versus diurnal habitat use showed that roosting habitat requirements are clearly different to foraging habitat requirements for both species. Other observational field studies have suggested that both species prefer to roost in trees next to or surrounded by water, whether at rivers, streams, lakes, or reservoirs/dams (Carrick \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1962\u003c/span\u003e; Marchant and Higgins \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e1990\u003c/span\u003e; McKilligan \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e1975\u003c/span\u003e; McKilligan \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e1979\u003c/span\u003e). The bimodal patterns in local and landscape water depth seen for these species in our modelling with MIM inundation data reflect this bi-modal split in habitat needs, with selection of sites adjacent to deep-water habitats for roosting, and selection of shallow-water habitats for foraging. The MIM modelling also indicated that these species are selecting for short-term water characteristics at relatively fine spatial scales, with water depth and inundation relationships being significant at pixel/point and 2km radii scales but not at the 20km radii scale. The implication of this split at local scales is that conservation or provision of foraging habitats without simultaneous consideration of available roosting habitats nearby is likely to be problematic for both species and may result in a reduced response by these species to management efforts. Provision of foraging habitats without suitable roosting habitats nearby may also result in increased energy use or wasted energy expenditure by birds that choose to travel from distant roosting habitats and may therefore not be optimal use of resources.\u003c/p\u003e \u003cp\u003eThe relative availability and modifications of habitats by agriculture and water resource use will also influence habitat use, most likely to a greater extent for straw-necked ibis than for royal spoonbills. This is because ibis tend to use more terrestrial (\u0026lsquo;non-ANAE\u0026rsquo;) habitats than spoonbills, and this together with their generalist diet and adaptability makes them more able to use agricultural land uses for foraging than spoonbills. The regular cycles of shallow irrigation, tillage, and sometimes burning used in intensive agriculture flush prey such as crickets, spiders and frogs from soil and attract other prey types. Ibis are well known to flock to sites receiving water to take advantage of the temporary abundance of food. However these resources are typically temporary, and management support for foraging habitats may be particularly important in areas where irrigated agriculture and other water sources effectively \u0026lsquo;dry up\u0026rsquo; seasonally, as well as in areas where the effects of climate change are severe (Perez-Moreno et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Spoonbills are more dependent on permanent surface water habitats and marsh/wetland habitats that provide a range of aquatic food types than ibis species and are therefore more likely to be affected by changes in these habitats either through water regime change or due to other pressures or threats. Support for foraging habitats when agricultural or other temporary resources are not available, particularly during winter, is likely to be critical for juvenile survival for these and similar species (Jelena et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eBeyond breeding sites, waterbirds need suitable habitat for critical processes within the life cycle. These include juvenile survival to breeding age (i.e. recruitment, which for some ibis and spoonbill species may take up to four years), adult recovery from breeding efforts, and adult survival between breeding events. Knowledge of habitat use during these periods is important for efficient and effective management and policy for waterbirds, particularly for species that are nomadic and use remote sites across wide geographic areas. Here, we characterised non-breeding habitat preferences for RSB and SNI and found clear intra-species differences. These differences have implications for management, and mapping of predicted habitat availability at basin-wide scales, providing context for prioritisation and application of resources. Such increased knowledge of the spatio-temporal interactions of waterbirds with their environment across complete life cycles is essential for informing management aimed at increasing waterbird numbers or maintaining diversity in the long term.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eDeclaration of competing interests\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests. \u003c/p\u003e \u003ch2\u003eAnimal ethics statement\u003c/h2\u003e \u003cp\u003eAll research protocols were approved by an authorized Animal Care and Ethics Committee, according to the Australian code of practice for the care and use of animals for scientific purposes. On-ground fieldwork activities were conducted under New South Wales and Victoria Scientific Licences 102180 and 10010534.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe original research that formed the basis of this article was co-funded by the Commonwealth Environmental Water Holder\u0026rsquo;s Office (CEWH/CEWO) and the Commonwealth Scientific and Industrial Research Organisation (CSIRO) through the CEWH Monitoring, Evaluation and Research project (2019\u0026ndash;2024) and the CEWO Environmental Watering Knowledge and Research project (2015\u0026ndash;2018), administered through the Commonwealth Environmental Water Office within the Department of Climate Change, Energy, the Environment and Water and its precursors. The research also benefited from co-investment by the Lake Cowal Conservation Centre, and from in-kind support from the Royal Botanic Garden Sydney (John Martin), NSW Department of Planning and Environment and its precursors, and the Goulburn-Broken Catchment Management Authority (Keith Ward).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eHM initiated and led the research and obtained funding. HM, FR, LON, SR, MP, MD, JH and JM conducted fieldwork and data collection. LLJ, AL, HM, FR, and JH processed the data for analysis. LLJ performed the analysis of the data with guidance from HM. HM and LLJ wrote the manuscript. RK, KB, VD and RM provided research direction and design advice at the beginning of the project.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors express their gratitude for the assistance of colleagues, collaborators and volunteers with fieldwork, and the support of program leaders.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eCode is available via the CSIRO Data Access Portal, https://data.csiro.au/, including all scripts used in the analysis. This repository contains the necessary files required to run the scripts and recreate the analyses. Raw data are available from the corresponding author upon reasonable request and will be uploaded to the CSIRO Data Access Portal at the time of publication.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eABARES (2021) Catchment scale land use of australia \u0026ndash; update December 2020. In \u0026apos;.\u0026apos; (Ed. ABARES): Canberra, Australia) \u003c/li\u003e\n\u003cli\u003eArthington, A.H., Bhaduri, A., Bunn, S.E., Jackson, S.E., Tharme, R.E., Tickner, D.\u003cem\u003e, et al.\u003c/em\u003e (2018) The Brisbane Declaration and Global Action Agenda on Environmental Flows (2018). \u003cstrong\u003e6\u003c/strong\u003e(45). [In English]\u003c/li\u003e\n\u003cli\u003eBenjamini, Y., and Yekutieli, D. (2001) The control of the false discovery rate in multiple testing under dependency. \u003cem\u003eAnnals of Statistics\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1165-1188. \u003c/li\u003e\n\u003cli\u003eBino, G., Brandis, K., Kingsford, R.T., and Porter, J. (2021) Shifting Goalposts: Setting Restoration Targets for Waterbirds in the Murray-Darling Basin Under Climate Change. \u003cem\u003eFrontiers in Environmental Science\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e. [In English]\u003c/li\u003e\n\u003cli\u003eBino, G., Kingsford, R.T., and Porter, J. (2015) Prioritizing Wetlands for Waterbirds in a Boom and Bust System: Waterbird Refugia and Breeding in the Murray-Darling Basin. \u003cem\u003ePlos One\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e(7). \u003c/li\u003e\n\u003cli\u003eBino, G., Steinfeld, C., and Kingsford, R.T. (2014) Maximizing colonial waterbirds\u0026apos; breeding events using identified ecological thresholds. and environmental flow management. \u003cem\u003eEcological Applications\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e(1), 142-157. \u003c/li\u003e\n\u003cli\u003eBrandis, K.J., Bino, G., Spencer, J.A., Ramp, D., and Kingsford, R.T. (2018) Decline in colonial waterbird breeding highlights loss of Ramsar wetland function. \u003cem\u003eBiological Conservation\u003c/em\u003e \u003cstrong\u003e225\u003c/strong\u003e, 22-30. \u003c/li\u003e\n\u003cli\u003eBrandis, K.J., Francis, R.J., and Bino, G. (2024) Vegetation and inundation characteristics of waterbird breeding sites in the Murray\u0026ndash;Darling Basin, Australia. \u003cem\u003eMarine and Freshwater Research\u003c/em\u003e \u003cstrong\u003e75\u003c/strong\u003e(7), -. \u003c/li\u003e\n\u003cli\u003eBrandis, K.J., Kingsford, R.T., Ren, S., and Ramp, D. (2011) Crisis water management and ibis breeding at Narran Lakes in arid Australia. \u003cem\u003eEnviron Manage\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e(3), 489-98. \u003c/li\u003e\n\u003cli\u003eBrooks, M.E., Kristensen, K., van Benthem, K.J., Magnusson, A., Berg, C.W., Nielsen, A.\u003cem\u003e, et al.\u003c/em\u003e (2017) glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. \u003cem\u003eThe R journal\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 400. [In en]\u003c/li\u003e\n\u003cli\u003eBrooks, S. (2021) ANAE Classification of the Murray-Darling Basin Technical Report, Revision 3.0, March 2021. Technical report for the Commonwealth Environmental Water Office, Department of Agriculture, Water and the Environment, Australia., Canberra, Australia.\u003c/li\u003e\n\u003cli\u003eCaley, P., Reid, J.R.W., Colloff, M.J., and Barry, S.C. (2022) On inferring population trends of mobile waterbirds from aerial transect surveys in variable environments. \u003cem\u003eEnvironmental and Ecological Statistics\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e(1), 3-31. \u003c/li\u003e\n\u003cli\u003eCalvert, A.M., Walde, S.J., and Taylor, P.D. (2009) Nonbreeding-Season Drivers of Population Dynamics in Seasonal Migrants: Conservation Parallels Across Taxa. \u003cem\u003eAvian Conservation and Ecology\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e(2), 5[online]. \u003c/li\u003e\n\u003cli\u003eCarrick, R. (1962) Breeding, movements and conservation of Ibises (Threskiornithidae) in Australia. \u003cem\u003eCSIRO Wildlife Research\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e(1), 71-88. \u003c/li\u003e\n\u003cli\u003eChiau, W.Y. (2006) Lagoon Management and Black-faced Spoonbill Conservation: Issues and Challenges in theGreater Chiku Area, Taiwan. \u003cem\u003eJournal of Coastal Research\u003c/em\u003e \u003cstrong\u003e2 (Winter 2006)\u003c/strong\u003e(Special Issue No. 39. Proceedings of the 8th International Coastal Symposium (ICS 2004)), 912-914. \u003c/li\u003e\n\u003cli\u003eDavidson, N.C. (2014) How much wetland has the world lost? Long-term and recent trends in global wetland area. \u003cem\u003eMarine and Freshwater Research\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e(10), 934-941. \u003c/li\u003e\n\u003cli\u003eFithian, W., and Hastie, T. (2013) Finite-sample equivalence in statistical models for presence-only data. \u003cem\u003eThe Annals of Applied Statistics\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e(4), 1917. \u003c/li\u003e\n\u003cli\u003eHaig, S.M., Mehlman, D.W., and Oring, L.W. (1998) Avian Movements and Wetland Connectivity in Landscape Conservation. \u003cem\u003eConservation Biology\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e(4), 749-758. \u003c/li\u003e\n\u003cli\u003eHirzel, A.H., Le Lay, G., Helfer, V., Randin, C., and Guisan, A. (2006) Evaluating the ability of habitat suitability models to predict species presences. \u003cem\u003eEcological Modelling\u003c/em\u003e \u003cstrong\u003e199\u003c/strong\u003e(2), 142-152. \u003c/li\u003e\n\u003cli\u003eJelena, K., Antun, Z., Tibor, M., and Otto, O. (2012) Movements of Immature Eurasian Spoonbills Platalea leucorodia from the Breeding Grounds of the Eastern Metapopulation in the Pannonian Basin. \u003cem\u003eWaterbirds\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e(2), 239-247. \u003c/li\u003e\n\u003cli\u003eJirinec, V., Rodrigues, P.F., and Amaral, B. (2021) Adjustable leg harness for attaching tags to small and medium-sized birds. \u003cem\u003eJournal of Field Ornithology\u003c/em\u003e \u003cstrong\u003e92\u003c/strong\u003e(1), 77-87. \u003c/li\u003e\n\u003cli\u003eKarl, B.J., and Clout, M.N. (1986) An improved radio transmitter harness with a weak link to prevent snagging. \u003cem\u003eJournal of Field Ornithology\u003c/em\u003e \u003cstrong\u003e58\u003c/strong\u003e(1), 73-77. \u003c/li\u003e\n\u003cli\u003eKingsford, R., and Johnson, W. (1998) Impact of water diversions on colonially-nesting waterbirds in the Macquarie Marshes of arid Australia. \u003cem\u003eColonial Waterbirds\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e(2), 159-170. \u003c/li\u003e\n\u003cli\u003eKingsford, R., and Norman, F. (2002) Australian waterbirds\u0026mdash;products of the continent\u0026apos;s ecology. \u003cem\u003eEmu\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e(1), 47-69. \u003c/li\u003e\n\u003cli\u003eKingsford, R.T. (2000) Ecological impacts of dams, water diversions and river management on floodplain wetlands in Australia. \u003cem\u003eAustral Ecology\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e(2), 109-127. \u003c/li\u003e\n\u003cli\u003eKingsford, R.T., and Auld, K.M. (2005) Waterbird breeding and environmental flow management in the Macquarie Marshes, arid Australia. \u003cem\u003eRiver Research and Applications\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e(2-3), 187-200. \u003c/li\u003e\n\u003cli\u003eKingsford, R.T., Roshier, D.A., and Porter, J.L. (2010) Australian waterbirds - time and space travellers in dynamic desert landscapes. \u003cem\u003eMarine and Freshwater Research\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e(8), 875-884. \u003c/li\u003e\n\u003cli\u003eKirby, J.S., Stattersfield, A.J., Butchart, S.H.M., Evans, M.I., Grimmett, R.F.A., Jones, V.R.\u003cem\u003e, et al.\u003c/em\u003e (2008) Key conservation issues for migratory land- and waterbird species on the world\u0026apos;s major flyways. \u003cem\u003eBird Conservation International\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e(S1), S49-S73. \u003c/li\u003e\n\u003cli\u003eLeslie, D.J. (2001) Effect of river management on colonially-nesting waterbirds in the Barmah-Millewa forest, south-eastern Australia \u003cem\u003eRegulated Rivers: Research \u0026amp; Management\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e(1), 21-36. \u003c/li\u003e\n\u003cli\u003eLindsell, J.A., Serra, G., Pe\u0026scaron;ke, L., Abdullah, M.S., al Qaim, G., Kanani, A.\u003cem\u003e, et al.\u003c/em\u003e (2009) Satellite tracking reveals the migration route and wintering area of the Middle East population of Critically Endangered northern bald ibis Geronticus eremita. \u003cem\u003eOryx\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e(3), 329-335. \u003c/li\u003e\n\u003cli\u003eMarchant, S., and Higgins, P.J. (1990) \u0026apos;Handbook of Australian, New Zealand and Antarctic Birds. Volume 1, Ratites to ducks.\u0026apos; (Oxford University Press, Melbourne) \u003c/li\u003e\n\u003cli\u003eMcCabe, J.D., Clare, J.D., Miller, T.A., Katzner, T.E., Cooper, J., Somershoe, S.\u003cem\u003e, et al.\u003c/em\u003e (2021) Resource selection functions based on hierarchical generalized additive models provide new insights into individual animal variation and species distributions. \u003cem\u003eEcography\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e(12), 1756-1768. \u003c/li\u003e\n\u003cli\u003eMcKilligan, N. (1975) Breeding and movements of the straw-necked ibis in Australia. \u003cem\u003eEmu\u003c/em\u003e \u003cstrong\u003e75\u003c/strong\u003e(4), 199-212. \u003c/li\u003e\n\u003cli\u003eMcKilligan, N. (1979) The ecology of the straw-necked ibis in winter at Toowoomba, South-east Queensland. \u003cem\u003eSunbird: Journal of the Queensland Ornithological Society, The\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e(3/4), 49. \u003c/li\u003e\n\u003cli\u003eMDBA (2014) Basin-wide environmental watering strategy Murray\u0026ndash;Darling Basin Authority for and on behalf of the Commonwealth of Australia, No. MDBA Publication No 20/14, Canberra, ACT, Australia.\u003c/li\u003e\n\u003cli\u003eMuff, S., Signer, J., and Fieberg, J. (2020) Accounting for individual-specific variation in habitat-selection studies: Efficient estimation of mixed-effects models using Bayesian or frequentist computation. \u003cstrong\u003e89\u003c/strong\u003e(1), 80-92. \u003c/li\u003e\n\u003cli\u003eNicol, S., Lloyd-Jones, L., \u0026amp; McGinness, H. M. (2024). A method to predict connectivity for nomadic waterbird species from tracking data. \u003cem\u003eLandscape Ecology\u003c/em\u003e, \u003cstrong\u003e\u003cem\u003e39\u003c/em\u003e\u003c/strong\u003e(2), 1-15.\u003c/li\u003e\n\u003cli\u003ePerez-Moreno, H., Martinez-Meyer, E., Mainero, J.S., and Rojas-Soto, O. (2016) Climatic patterns in the establishment of wintering areas by North American migratory birds. \u003cem\u003eEcology and Evolution\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e(7), 2022-2033. \u003c/li\u003e\n\u003cli\u003ePeters, W., Hebblewhite, M., Cavedon, M., Pedrotti, L., Mustoni, A., Zibordi, F.\u003cem\u003e, et al.\u003c/em\u003e (2015) Resource selection and connectivity reveal conservation challenges for reintroduced brown bears in the Italian Alps. \u003cem\u003eBiological Conservation\u003c/em\u003e \u003cstrong\u003e186\u003c/strong\u003e, 123-133. \u003c/li\u003e\n\u003cli\u003ePrice, C.J., Morris, A., Staines, G., Payne, R., and Smith, J. (2018) Leaving home but nowhere to go: lessons learnt from almost two decades of Bush Stone-curlew Burhinus grallarius monitoring on the Central Coast of NSW. \u003cem\u003eAustralian Zoologist\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e(4), 769-783. \u003c/li\u003e\n\u003cli\u003eR Core Team (2024) R: A language and environment for statistical computing. In \u0026apos;.\u0026apos; (R Foundation for Statistical Computing: Vienna, Austria) \u003c/li\u003e\n\u003cli\u003eReid, J.R.W., Colloff, M.J., Arthur, A.D., McGinness, H.M. (2013) Influence of catchment condition and water resource development on waterbird assemblages in the Murray-Darling Basin, Australia. \u003cem\u003eBiological Conservation\u003c/em\u003e \u003cstrong\u003e165\u003c/strong\u003e, 25-34. \u003c/li\u003e\n\u003cli\u003eRenner, I.W., Elith, J., Baddeley, A., Fithian, W., Hastie, T., Phillips, S.J.\u003cem\u003e, et al.\u003c/em\u003e (2015) Point process models for presence-only analysis. \u003cstrong\u003e6\u003c/strong\u003e(4), 366-379. \u003c/li\u003e\n\u003cli\u003eRobinson, F. and McGinness, H.M. (2024). Mortality causes in ibis and spoonbill (Threskiornithidae) species and life stages: a global review. Under Review.\u003c/li\u003e\n\u003cli\u003eRoshier, D.A., and Asmus, M.W. (2009) Use of satellite telemetry on small-bodied waterfowl in Australia. \u003cem\u003eMarine and Freshwater Research\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e(4), 299-305. \u003c/li\u003e\n\u003cli\u003eSerra, G., Bruschini, C., Lindsell, J.A., Peske, L., and Kanani, A. (2011) Breeding range of the last eastern colony of Critically Endangered Northern Bald Ibis Geronticus eremita in the Syrian steppe: a threatened area. \u003cem\u003eBird Conservation International\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e(3), 284-295. \u003c/li\u003e\n\u003cli\u003eSmith, A., and Munro, U. (2011) Local and regional movements of the Australian white ibis threskiornis molucca in eastern Australia. \u003cem\u003eCorella\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 89-94. \u003c/li\u003e\n\u003cli\u003eTeng, J., Penton, D., Ticehurst, C., Sengupta, A., Freebairn, A., Marvanek, S.\u003cem\u003e, et al.\u003c/em\u003e (2023) Two-monthly Maximum Flood Water Depth Spatial Timeseries for the MDB v20 CSIRO, https://doi.org/10.25919/c5ab-h019.\u003c/li\u003e\n\u003cli\u003eThaxter, C.B., Ross-Smith, V.H., Clark, J.A., Clark, N.A., Conway, G.J., Marsh, M.\u003cem\u003e, et al.\u003c/em\u003e (2014) A trial of three harness attachment methods and their suitability for long-term use on Lesser Black-backed Gulls and Great Skuas. \u003cem\u003eRinging \u0026amp; Migration\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e(2), 65-76. \u003c/li\u003e\n\u003cli\u003eWang, C., Liu, D.-P., Qing, B.-P., Ding, H.-H., Cut, Y.-Y., Ye, Y.-X.\u003cem\u003e, et al.\u003c/em\u003e (2014) The Current Population and Distribution of Wild Crested Ibis Nipponia nippon. \u003cem\u003eChinese Journal of Zoology\u003c/em\u003e \u003cstrong\u003e49\u003c/strong\u003e(5), 666-671. \u003c/li\u003e\n\u003cli\u003eWood, S.N. (2017) \u0026apos;Generalized Additive Models: An Introduction with R.\u0026apos; (Chapman and Hall/CRC) \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"landscape-ecology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"land","sideBox":"Learn more about [Landscape Ecology](https://www.springer.com/journal/10980)","snPcode":"10980","submissionUrl":"https://submission.nature.com/new-submission/10980/3","title":"Landscape Ecology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"environmental water, satellite telemetry, foraging, nomadic, habitat selection, conservation management, waterbirds","lastPublishedDoi":"10.21203/rs.3.rs-4626784/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4626784/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eContext\u003c/h2\u003e \u003cp\u003eNomadic waterbirds are highly mobile across a range of spatial and temporal scales, which makes it difficult to monitor, quantify, and predict their habitat use with traditional methods, especially between breeding events when individuals and flocks can move over vast areas.\u003c/p\u003e\u003ch2\u003eObjectives\u003c/h2\u003e \u003cp\u003eThis study aimed to provide accurate information on habitat use to improve strategic conservation management of these species, particularly the provisioning of environmental water.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eTo overcome the challenges of distance and remoteness, we analysed a 7-year GPS satellite telemetry dataset from 141 individuals. We quantified habitat selection post-dispersal from breeding sites, and predicted habitat preference for two wading waterbird species of the Threskiornithidae family that frequently nest together at the same sites: straw-necked ibis (\u003cem\u003eThreskiornis spinicollis\u003c/em\u003e) and royal spoonbill (\u003cem\u003ePlatalea regia\u003c/em\u003e).\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eBoth long-term and short-term landscape-scale habitat associations differed between species. Royal spoonbills used fewer and more restricted habitat types than straw-necked ibis. Spoonbills displayed strong preferences for reservoirs, marshes and permanent wetlands, while ibis used both aquatic and terrestrial habitat, including areas of intensive animal production, modified pasture, and woodlands. Analysis of nocturnal versus diurnal space use showed that roosting and foraging habitat requirements for both species are distinct.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eAnalysing over 1\u0026nbsp;million telemetry points revealed species-level variability in habitat use, informing resource allocation for environmental water management. Royal spoonbills are more vulnerable to habitat change due to water regime alterations, highlighting the need for focused conservation management. Differences in day and night habitat use indicate the necessity of considering roosting habitats alongside foraging habitats for effective conservation. This comprehensive understanding of waterbirds' spatiotemporal interactions with their environment is crucial for long-term management aimed at increasing waterbird numbers and maintaining diversity.\u003c/p\u003e","manuscriptTitle":"Habitat use by nomadic ibis and spoonbills post-dispersal from breeding sites","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-19 00:46:07","doi":"10.21203/rs.3.rs-4626784/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-08-24T12:12:43+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-29T05:48:53+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-28T00:24:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"39503761045420040861111630855399300000","date":"2024-07-19T12:20:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"279142547436247375865089733240122118477","date":"2024-07-17T23:28:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"310218433166119729294732717129436003216","date":"2024-07-06T22:45:07+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-07-06T06:49:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-06-25T16:15:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-06-25T16:12:25+00:00","index":"","fulltext":""},{"type":"submitted","content":"Landscape Ecology","date":"2024-06-24T01:17:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"landscape-ecology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"land","sideBox":"Learn more about [Landscape Ecology](https://www.springer.com/journal/10980)","snPcode":"10980","submissionUrl":"https://submission.nature.com/new-submission/10980/3","title":"Landscape Ecology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"412583d3-8815-498b-a188-fc5be0417cd0","owner":[],"postedDate":"July 19th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-11-11T16:01:26+00:00","versionOfRecord":{"articleIdentity":"rs-4626784","link":"https://doi.org/10.1007/s10980-024-01982-1","journal":{"identity":"landscape-ecology","isVorOnly":false,"title":"Landscape Ecology"},"publishedOn":"2024-11-05 15:57:01","publishedOnDateReadable":"November 5th, 2024"},"versionCreatedAt":"2024-07-19 00:46:07","video":"","vorDoi":"10.1007/s10980-024-01982-1","vorDoiUrl":"https://doi.org/10.1007/s10980-024-01982-1","workflowStages":[]},"version":"v1","identity":"rs-4626784","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4626784","identity":"rs-4626784","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}