Context Understanding habitat selection and home ranges of threatened animals is important to their effective conservation. The numbat (Myrmecobius fasciatus), an endangered marsupial that has experienced extreme range contraction and now only persists in two small indigenous populations, is a species where more nuanced information on habitat selection would greatly benefit its management. Aim Investigate the spatial ecology of the numbat in the largest of these two remaining populations: the Upper Warren region of south-western Australia. Methods We collected 18 GPS datasets from 15 numbats (n = 13 females, two males) over an 18-month period between October 2019 and March 2021, to assess habitat use at a macrohabitat (landscape) and microhabitat scale. Key results Female home-range size varied from 11.2 to 53.4 ha, with a median value of 29.5 ha, whereas male home-range size was considerably larger (80.2 and 278.7 ha). Factors that increased the likelihood of habitat use by numbats at a macrohabitat (landscape) scale included more recent timber harvesting, proximity to roads, and loamy gravel soil type, whereas more recent fire and proximity to nearest surface water feature decreased the likelihood of habitat use. The presence of hollow logs was positively related with high-use areas by numbats at the microhabitat scale. Conclusions This study has advanced our understanding of how numbats utilise forested habitats and showed that they have mixed responses to land management activities. Future research could focus on more detailed understanding of habitat preferences found in this study, including an investigation of numbat habitat use in relation to the broad range of fire regimes, and the importance of hollow logs. Implications We recommend integrating these findings into conservation management strategies for the numbat, to ensure continuing availability of high-quality habitat for the important Upper Warren population.
Context Protected areas (PAs) receive visitors pursuing environmental education and recreation objectives, but this can negatively affect the conservation of their biodiversity. Quantification of this impact is scarce, especially in developing countries. The Anthropause generated by the COVID-19 lockdown constituted a remarkable change in human activities and a good opportunity to investigate wild biota in the absence of people. Aims The objective of this study was to quantify the impact of visitors on the richness and abundance of birds on the paths of an Argentine protected area with high visitor traffic (~1 million people per year). Methods Birds perched on the paths (ground and surrounding vegetation) were recorded weekly during 12 days in the years 2020 (closed to visitors) and 2021 (open to visitors). Each day, 10 points were sampled and each point was recorded for 15 min. Key results Both the time of the survey and the opening to visitors were relevant factors in explaining the richness and abundance of birds. In general, a decrease in both variables was recorded as time passed after dawn and when there were visitors compared to when there were no visitors. At the mean value of time, the decrease in bird records associated with visitors was 49% for richness and 65% for abundance. Ground-perched birds showed no increased sensitivity to visitor exposure, because recording percentages remained consistent across both study years. Conclusions The results of this work suggest a benefit for wildlife in natural areas during the Anthropause of 2020. Implications Regarding management applications it is suggested to minimise the number of paths, so as to concentrate and limit negative effects, and to enhance the structural and floristic complexity of pathside vegetation.
Context The presence of the white shark (Carcharodon carcharias) off the coast of Brazil is rare and without records of its feeding behaviour. The Franciscana dolphin (Pontoporia blainvillei) is an endangered species in the same area, with declining populations. Aims Provide the first forensic record of C. carcharias predating on P. blainvillei off the coast of the south-western Atlantic, documenting a new white shark occurrence off the Brazilian coast. Methods The white shark wounds were identified according to the dimensions of the teeth marks, bite morphology, tooth allometry and scars orientation. The total length and weight of the white shark were estimated by using the measurement of the upper jaw bite circumference (UJBC), by using bite marks and TPSdig software for calculations. The age estimation of the Franciscana dolphin was established through the use of the von Bertalanffy model. The Franciscana dolphin was found stranded on the beach, with death estimated between 3 and 5 days earlier according to skin peeling and decomposition state. Key results The Franciscana dolphin (~1.72-m total length, TL) was an adult specimen (estimated to be ≥4 years), unsexed because of two sharp bites (Bite III) in the ventral area and one in the dorsal area. An upper jaw average interdental distance between teeth A1 and A2 on the right side measured 0.5 cm and the Bite I UJBC measured 60.3 cm. The white shark bite measurements indicated an estimated TL between 3.6 and 3.8 m, and weight between 400 and 490 kg. Bite patterns suggest a subsurface attack strategy, consistent with white shark predation on small cetaceans documented globally. Bite I was the most recent bite on the Franciscana dolphin. Bite II caused haemorrhage on the dorsal fin, indicating that the cetacean was still alive when it was predated. Conclusions The predation of C. carcharias on P. blainvillei confirmed that the species is an active inhabitant of the Brazilian southern coast, on the basis of forensic evidence obtained from the stranded specimen. Implications Given the rare records of white shark on the coast and the systematic decline of both species, it is essential to elucidate their roles in the ecosystem.
Context Feral pigs are one of the most destructive invasive vertebrate species in Australia. Aerial shooting is considered one of the most effective methods for causing a rapid reduction in feral pig density at a landscape scale. Aims We aimed to assess the effects of 7 years of annual aerial shooting operations in the northern agricultural region of Western Australia. We measured the catch per unit effort (CPUE) of aerial shooting operations and assessed demographic, reproductive and dietary characteristics of the feral pig population. Methods We used operational data, including number of animals culled, and hours flown, to measure CPUE, measured in pigs per hour across 7 years at four spatially independent shoot blocks. We performed necropsies on 1502 culled pigs, collecting a range of morphometric, demographic, reproductive and diet metrics. Key results Over the 7 years of control 13,554 feral pigs were culled, with an average CPUE of 48.10 pigs h−1. The CPUE increased over the course of control, with a mean annual population growth rate (r) of 0.16 (95% CI: 0.12–0.22). Males were larger than females (mean ± s.d., adult bodyweight of 56.1 ± 22.8 v. 46.4 ± 17.7 kg), and slightly older (12.8 ± 11.7 v. 11.8 ± 11.1 months). Probability of survival of feral pigs to 3 years old was 0.12. On average, 62.3% of sexually mature females were pregnant, the mean in utero litter size was 5.03 (±18.6 s.d.), and this increased 0.05 per year of age of the sows. Sows bred year-round with two peaks in births. Cereal crops, primarily wheat, were the most frequently consumed food items, along with green vegetation and lupins. Conclusions Aerial shooting of feral pigs in the region has not resulted in a decline in population density. This is likely to be due to the species’ high fecundity and strong juvenile survival rates. Further, there is no evidence that aerial shooting has caused any noticeable changes in demographic parameters over time. Implications These results highlight the need to assess additional or alternative approaches to feral pig control and to analyse trends in pig removals during management activities.
Context Increasing fire frequency and severity is a major conservation concern for forest-dependent species, particularly the nationally endangered southern greater glider (Petauroides volans), which has had populations severely affected by the 2019–20 Australian wildfires. Aims This study aimed to quantify the recovery of southern greater gliders 5 years after fire in Monga National Park (NP), New South Wales. Methods Double-observer spotlight surveys were conducted to estimate occupancy and density using multi-covariate distance sampling across 14 sites previously surveyed in 2020. Sites represented three fire severity classes: low (canopy unburnt), moderate (partial canopy burn) and high (complete canopy consumption). Estimates from 2020 (<1 year after fire) were compared with those from 2025 (>5 years after fire). Key results Fire severity class and elevation both influenced greater glider occupancy. The overall mean estimated density at Monga NP increased from 0.46 gliders ha−1 (95% confidence interval 0.26–0.65) in 2020 to 1.05 gliders ha−1 (95% confidence interval 0.55–2.01) in 2025. This increase was primarily driven by a rise in density of gliders at three low- and moderate-severity refuges, whereas the majority of occupied sites (5/8) showed slight declines in glider observations. No southern greater gliders were observed at high severity burnt sites in 2025, despite some records of gliders at one site 10 months post fire in 2020. Conclusions Although populations of southern greater gliders at Monga NP have become largely confined to a few low- and moderate-severity fire refuges, these sites have acted as strongholds for the species. Implications Areas that have experienced lower fire severity play a key role in sustaining southern greater glider populations and facilitating their recovery over the long term. Reducing wildfire severity and protecting less-affected habitats is essential, as 5 years post fire appears to be insufficient for recovery of gliders once found in high-severity sites.
Context The Testudine shell has contributed to the success of this clade for >200 million years by conferring protection against predators. Despite this, the slow life history of Testudine species increases their vulnerability in this rapidly changing world, and many are threatened with extinction. Ex situ conservation programs can rescue species from extinction, but maintaining optimal environmental conditions in captivity can be challenging. One species for which captive breeding is instrumental for recovery is the critically endangered western swamp tortoise (Pseudemydura umbrina). Anecdotal evidence suggests that the shells of captive tortoises are softer than those of individuals in the wild, which has the potential to have significant survival implications for zoo-bred individuals that are released into the wild. Aims The aims of our investigation were to (i) compare variation in environmental parameters, including ultraviolet B (UVB) levels that is important for Testudine shell growth, between a captive and wild population of the western swamp tortoise, and (ii) test empirically whether zoo-bred western swamp tortoises have softer shells than wild individuals. Methods We used remote sensing equipment to measure environmental variables (UVB light, temperature and relative humidity) within the captive and wild environments. Additionally, we used a tension-calibrated digital micrometer to evaluate differences in shell hardness among zoo-bred and wild individuals. Results We found that shell hardness in juvenile zoo-bred tortoises was significantly lower than that in wild juveniles, but did not detect differences among adults from the two populations. Our environmental measurements showed that UVB levels in the captive environment were significantly lower than levels recorded in the wild, and temperature and relative humidity were significantly higher. Conclusions Taken together, our results indicated that differences between wild and captive conditions are evident and may influence shell hardness of zoo-bred juveniles, an important insight for the captive breeding of a critically endangered species. Further research is needed to investigate the implications of softer shells for released tortoises, for example, regarding the likelihood of predator attack success. Implications Our research highlighted the need to consider the environmental conditions of the captive environment in ex situ conservation programs of endangered Testudines.
Context Spatial behaviour research involving GPS tracking is increasingly being used to support the management of wild (including feral) pigs (Sus scrofa) around the globe. Capturing and collaring pigs can have severe impacts on their health and welfare, but the implications of different capture and restraint methods practiced in large-scale studies remain unclear. Aims We compared the effectiveness and animal welfare outcomes of three methods for restraining wild pigs for GPS collar deployment: (1) remote darting with tiletamine-zolazepam and xylazine (TZX), (2) pole syringe injection with tiletamine-zolazepam (TZ), and (3) manual restraint without chemical immobilisation. Methods We retrospectively analysed data from 360 collaring events across three independent collaring programs in Australia. We quantified the durations of procedures, the frequency of adverse events (e.g. hyperthermia, mortality), and post-release movement patterns using a standardised animal welfare assessment framework. Key results Manually restrained pigs experienced the shortest exposure to capture and handling stressors, and showed no increase in adverse outcomes compared to chemically immobilised pigs. Pigs darted with TZX had more variable induction and recovery times than those administered TZ via pole syringe, but the use of reversal agents (yohimbine, atipamezole) reduced overall handling time. Pole syringe immobilisation with TZ produced more consistent inductions but longer recumbency durations. Post-release movement patterns were similar across all methods, with minor behavioural disruptions for about 4 days. Conclusions Manual restraint is a viable option for telemetry studies where experienced handlers are available and extended procedures are not required. When chemical immobilisation is desirable, animal welfare outcomes can be improved by using drug combinations with reversal agents and by applying careful monitoring and intervention. Implications These findings support evidence-based refinement of wild pig capture protocols and highlight the importance of publishing animal welfare outcomes to inform best practice in wildlife research and management.
Context Volant and non-volant mammals differ in their sensitivity to environmental and anthropogenic pressures, making species traits central to understanding biodiversity erosion. However, the extent to which forest structure predicts faunal composition remains unclear. Aims We tested how eco-behavioural traits and rainforest structure jointly influence detection and occupancy in mammal assemblages, and whether habitat characteristics reliably explain variation in community composition. Methods Over 12 months, we combined ground and arboreal camera trapping, bat flight-performance indices, and vegetation measurements in mature lowland forest in the Yangambi Biosphere Reserve, Democratic Republic of the Congo. Single-season multi-species occupancy models were used to estimate variation in detection and occupancy as a function of species traits and habitat characteristics while accounting for imperfect detection. Key results Detection probability declined sharply with increasing body mass and sociality (Pi = 0.06, β = −39.08), indicating non-random depletion of large-bodied, group-living mammals consistent with trait-mediated filtering under human pressure. Stem density positively affected occupancy (ψ = 0.30, β = 9.06), but this effect was insufficient to offset evidence of defaunation, showing a mismatch between forest condition and assemblage composition. By contrast, bats showed high occupancy (ψ = 0.85) across structural gradients (β = −1.92), whereas detection probability (Pi = 0.162) was unrelated to flight traits, suggesting greater ecological flexibility. Conclusions Non-volant mammal assemblages were shaped primarily by trait-mediated filtering rather than forest structure alone. By contrast, bats maintained high occupancy across heterogeneous habitats, supporting functional connectivity and showing contrasting responses to environmental change in tropical forests. Implications Structural metrics alone are insufficient proxies for ecosystem intactness. Integrating species traits with occupancy-based approaches can provide early-warning indicators of biodiversity change and support resilience-based management and restoration of tropical forests.
Context The management of feral cats (Felis catus) is crucial to effective species conservation. One management approach is to aerially apply toxic baits (containing sodium monofluoroacetate) to achieve landscape scale control. However, limited bait types are available to target cats. Aims In the absence of specific baits (i.e. Eradicat® or Curiosity®) for the control of feral cats in New South Wales (NSW), we aimed to investigate the effectiveness of red meat baits, intended for canid control, to reduce local feral cat density. Methods We deployed a grid of 65 camera traps, spaced at approximately 1-km intervals, in the Guy Fawkes River National Park, NSW. The feral cat population was surveyed 7 weeks prior to and post-baiting, separated by a 4-week baiting period that was unmonitored. Cats detected by camera were individually identified and a spatial-mark-resight model was used to estimate the change in feral cat density. Key results In total, 34 individual feral cats were identified. The feral cats present after baiting were mostly different to those individuals detected prior to baiting. However, feral cat density was not substantially different between surveys (pre-baiting: 0.56 cats km−2, 95% CI: 0.39–0.81 cats km−2; post-baiting: 0.58 cats km−2, 95% CI: 0.40–0.84 cats km−2). Conclusions Aerially deployed red meat baits, intended for canid control, had a negligible effect on feral cat density. The density estimate reported was two times higher than the average cat density previously modelled for natural environments across Australia. Implications With a higher-than-expected density, and no clear population reduction from red meat baiting, it is possible the local feral cat population may be having a greater impact on regional biodiversity than previously expected. However, given the lack of replication, we caution against pre-emptively generalising the outcomes of this study across different seasons and ecoregions.
The eradication and control of feral cats (Felis catus) on offshore islands is a conservation priority in Australia to protect threatened species. However, this task is challenging and resource-intensive, particularly in remote and inaccessible locations. This study aimed to evaluate the effectiveness of lethal aerial baiting and Felixer grooming traps in reducing feral cat activity on Three Hummock Island, Tasmania, to support the establishment of a hooded plover (Thinornis rubricollis) stronghold. Two rounds of aerial baiting using Curiosity poisoned baits were undertaken in May 2021 and September 2023. Additionally, five Felixer units were deployed with a 6-month non-lethal period before being switched to lethal mode in November 2021. Felixers use image recognition to identify cats and apply a sodium fluoroacetate (1080) poison gel, which is ingested during grooming. A network of 20–60 camera traps, operating over 37,175 trap days during 2019–2024, was used to monitor changes in feral cat activity and site usage. Feral cat relative activity (proportion of days with cat events) steadily declined following control efforts, with site usage halving after Felixer deployment in 2021 (dynamic-occupancy model slope for extinction without Felixers = −2.22, s.e. = 0.929, P = 0.0121). By contrast, aerial baiting did not significantly affect site usage (first round: estimate = 0.921, s.e. = 1.04, P = 0.378; second round: estimate = −1.021, s.e. = 0.687, P = 0.137) or activity. Despite targeting only nine cats, the Felixers contributed to a substantial decline in cat presence, suggesting a small, possibly inbred population prior to control. Camera trap images revealed that most cats were ginger, and many had kinked tails, further supporting the likelihood of inbreeding depression. Felixers were effective in reducing feral cat activity with minimal intervention, whereas aerial baiting alone had limited impact. These findings highlight the potential of automated, low-effort control methods for managing feral cat populations on remote islands. This study demonstrates that Felixers may be an effective alternative to traditional high-effort control methods, such as shooting, trapping, and baiting. However, ongoing monitoring and additional control efforts are required to ensure progress toward complete eradication.
Reintroductions are challenging and many fail. The red-tailed phascogale (Phascogale calura) is a small semi-arboreal Australian dasyurid marsupial that has been extirpated from 99% of its pre-colonisation range. Nine reintroductions have been attempted since 2005 across five States and Territories, six into feral predator-free fenced areas (safe havens) from which feral cats and foxes were extirpated, and three into unfenced areas where cat and/or fox control occurs. We summarise the methods and results of these reintroduction attempts to inform future management. Of the nine reintroduction attempts, three were considered successful and support populations extant for more than five years (2 safe havens and 1 unfenced), three did not lead to population establishment (2 safe havens, 1 unfenced), two safe haven populations remain extant two years after initial reintroduction, and one unfenced reintroduction is still underway. All three successful reintroductions released wild-sourced adult animals, sometimes in combination with captive-bred animals, and were within 280 km of extant wild populations. In comparison all releases to date that did not result in established populations primarily released captive-bred animals. The main challenges encountered during red-tailed phascogale reintroductions included predation by introduced and native predators, hyperdispersal, toxoplasmosis, and severe weight loss post-release. All but one reintroduction (Western Australia, safe haven, wild-sourced founders) reported low recapture rates of founders or high mortality of radio-collared individuals, suggesting low initial survival of translocated animals. Red-tailed phascogales may be more challenging to reintroduce and monitor than some other marsupials due to their small size, making them 1) more vulnerable to predation from both native and invasive (i.e. feral cats and foxes) predators, 2) more susceptible to starvation after release, and their 3) more likely to hyperdisperse through safe haven fences. Captive breeding may affect the probability of post-release survival through inappropriate anti-predator behaviours, unfamiliarity with food sources at release sites and underdeveloped hunting behaviour, and possibly an increased susceptibility to toxoplasmosis. We recommend that practitioners consider sourcing wild animals for translocations where possible or, if using captive-bred animals, employing rigorous toxoplasmosis assessment and hygiene alongside a staged soft-release strategy involving gradually weaning each release group off supplementary food in situ.
Context Zoo-based conservation breeding and release programs help mitigate biodiversity loss from human-driven change, using adaptive management to evaluate actions, guide research priorities, and inform policy. Zoo-based conservation breeding has played a significant role in the recovery program of the critically endangered regent honeyeater. Over 400 zoo-bred regent honeyeaters have been released into New South Wales (NSW) and Victoria (VIC), Australia between 2000 and 2022 to bolster wild populations. Aims The aim of the study was to develop a set of key actions to shape the adaptive management strategy for the zoo-based regent honeyeater breeding and release program. Methods We analysed the post-release apparent survival and reproductive success of 414 zoo-bred regent honeyeaters released to the wild in NSW and VIC. The data from the Victorian release were examined in a previous publication; however, they are included here with updated information from the most recent post-release sightings and a revised analytical approach. To generate the management recommendations, we used a combination of evidence-based data generated from the first and second review, literature review and stakeholder engagement. Key results The first two actions are specifically highlighted as important by the results of the current study: (i) release the majority of birds at the optimum age of 1–3 years, and (ii) consider parents’ age when pairing birds and where possible breed from older mothers but younger fathers. The other four actions are generated from the synthesis of results from both reviews or from evaluation of the literature and expert consultation: (iii) undertake song tutoring for zoo bred birds and assigning song quality score, (iv) birds to spend a minimum of 8 weeks, prior to release, in a multispecies free flight aviary, (v) manage clutches per pair per season and (vi) release birds with breeding experience (where appropriate). Conclusions This review highlights the importance of the adaptive management framework to drive the enhancement of the program, and these actions form a flexible framework that may be further adapted as new information emerges. Implications We recommend regular evaluation of the regent honeyeater breeding and release program so that new actions can be implemented iteratively in the adaptive management strategy.
Context. With over 1500 species, bats provide essential ecosystem services but face growing threats worldwide. As keystone species in cave ecosystems, they serve as bioindicators of ecosystem health. Their sensitivity to climate change and human disturbances highlights the need for long-term phenological studies. Although being well documented in Europe and North America, bat phenology remains understudied in many regions and species. Aims. This study examines the annual reproductive patterns, body mass variation, and activity levels of three threatened bat species inhabiting Gora'an Cave, Safi, Morocco. Methods. Over a 2-year survey, we assessed the reproductive status of mist-netted individuals by using sex-specific morphological criteria. Additionally, acoustic surveys monitored seasonal changes in flight activity at the cave entrance. Key results. Two years of acoustic monitoring recorded 17,570 bat passes from eight species, with Myotis punicus (6149), Rhinolophus mehelyi (5642), and Rhinolophus blasii (5311) being the most detected. Activity patterns of M. punicus, and R. mehelyi were similar, lower in winter and higher in spring and summer. Activity increased with temperature in both species and with relative humidity in R. mehelyi. In contrast, R. blasii exhibited higher activity in winter, moderate activity in spring and autumn, and lower summer activity, which was negatively correlated with temperature. A total of 355 individuals of these three focal species were captured. Breeding was synchronized from March to May. Lactating M. punicus females appeared in April, while R. mehelyi and R. blasii started in May, with weaning in June. Males displayed signs of reproductive readiness from July to November. Body mass varied among species, sexes and months, increasing during gestation in April and May. Conclusions. This study has highlighted the importance of subterranean caves as critical habitats for threatened bat species during key reproductive phases. It also demonstrated species-specific response to microclimatic variation in shaping seasonal flight dynamics, and underscore the value of combining morphological and acoustic monitoring to enhance understanding of bat phenology, and activity patterns.
Context Wildlife management in the Anthropocene is a complex and challenging task, that involves balancing the needs of different species and ecosystems. There is now broad consensus that a network of large, feral predator-free fenced areas must form part of any strategy to prevent further extinctions of Australian mammals. However, fencing to protect threatened species from introduced predators can have drawbacks such as restricting natural migration and dispersal, and creating barriers to gene flow or ecological processes. Therefore, developing innovative solutions to optimise the use of fencing and minimise negative effects is an important goal for Australian conservation practitioners. Aims We aimed to develop a method for allowing target species to move from one side of a conservation fence to another, while maintaining the fence as a barrier for all other species. Methods We designed and built a prototype artificial intelligence (AI)-operated Smart Gate system to allow target species to move through a fence while preventing non-target species from doing so. We conducted two rounds of field-testing at Karakamia Wildlife Sanctuary to train the AI species classifier to recognise brush-tailed bettongs (woylies, Bettongia penicillata) and to test the AI-operated Smart Gate system. Key results The field-trials provided a convincing proof of concept, with 42 successful and safe woylie transits and zero non-target species transits through the Smart Gate system from 73 events when an animal entered the system. Conclusions The AI-operated Smart Gate has the potential to be used as a management tool to allow transit of target species across conservation fences. Implications There are many potential future applications for this technology, including allowing target species to enter or exit a fenced area, improving connectivity between a fenced area and the surrounding landscape by allowing some species to pass through the fence in either direction, or acting as a trap for a highly targeted monitoring or management purpose. Although further refinement is needed before a real-world application, devices such as this Smart Gate are likely to become part of the toolkit for managing conservation fences in Australia and beyond.
Context Commercial forestry plantations can provide shelter, food and connectivity among native forest remnants in landscapes with extensive habitat clearance. In their southern range, koalas (Phascolarctos cinereus) have responded favourably to the establishment of blue gum (Eucalyptus globulus) plantations, but their use of other plantation species remains undocumented. Understanding this is critical for the conservation of a genetically distinct koala population in the heavily modified Strzelecki Ranges region of Victoria. Aim We aimed to determine the relative density of koalas in four plantation species compared with remnant native forest in the Strzelecki Ranges, Victoria. Methods Sites (N = 59) were stratified across six forest types, including four plantation species (blue gum, Eucalyptus globulus subsp. globulus; shining gum, E. nitens; mountain ash, E. regnans; and radiata pine, Pinus radiata) and two native forest classifications (high or low suitability for koalas). Sites were each surveyed for koalas once, using a drone mounted with a thermal camera and spotlight (DJI Matrice 300 RTK) that was flown in manual mode and systematically searched each site. Koalas were detected in real time by the pilot. Survey effort was determined from the area of canopy and subcanopy surveyed, which ranged from 2.2 to 14.6 ha per site. A generalised linear model with negative binomial distribution was used to determine the influence of forest type on relative koala density. Key results Relative koala density was highest in blue gum plantations, with 0.28 koalas ha−1 (95% CI: 0.13–0.61) and high-suitability native forest, with 0.38 koalas ha−1 (95% CI: 0.18–0.82). Other plantation types had relative densities of fewer than 0.1 koalas ha−1 on average. Large confidence intervals for all forest types suggested high variability or uncertainty. Conclusions Blue gum was the only plantation species favoured by koalas; however, it is grown in only a small proportion (<5%) of plantations in the region. Implications Most of the plantation estate is not habitat preferred by koalas. Conserving koalas in this region therefore relies on maintaining and increasing the extent, connectivity and quality of native forest. Further study is needed to improve our understanding of koala tree preferences and the influence of landscape characteristics on use of different forest types.
Context Human activities increasingly reshape wildlife habitats by modifying resource availability, risk levels, and interspecific interactions. Understanding how sympatric species adjust their trophic niches under anthropogenic pressure is essential for predicting community responses to environmental change.Aims We investigated trophic niche differentiation and dietary overlap between the African golden wolf (Canis anthus) and the red fox (Vulpes vulpes) during the dry season across three habitat types in southern Tunisia (natural, agricultural, and urban areas), representing a gradient of human disturbance.Methods Over a 6-month period (April to September), 833 fecal samples were collected and analyzed using standard scat analysis methods. Sampling across all sites occurred within the same seasonal window to minimize potential seasonal variation in prey availability. Both species exhibited marked dietary plasticity across habitats.Key results The African golden wolf consumed primarily plant material in all habitats, with rodents dominating in agricultural areas and livestock remains occurring mainly in urban environments. In contrast, the red fox relied largely on invertebrates, rodents, birds, and plant material, with relative contributions varying among habitats. Trophic niches differed between species and among habitats, indicating clear dietary differentiation. Dietary overlap varied along the anthropogenic gradient, with higher overlap in agricultural and urban habitats than in natural areas, indicating increased resource sharing in human-modified landscapes during this seasonal snapshot.Conclusions Anthropogenic disturbance strongly influences trophic niche structure and dietary overlap between sympatric canids in southern Tunisia.Implications These findings highlight the influence of human activities on both prey availability and trophic interactions, and demonstrate how dietary plasticity facilitates the coexistence of sympatric canids in increasingly fragmented ecosystems, at least during the dry season.
Context. Little ravens (Corvus mellori) have emerged as a prominent predator on the clutches of burrow-nesting little penguins (Eudyptula minor) in recent decades on Phillip Island, Australia, raising concerns for conservation and eco-tourism. It has been suggested that consumption of penguin clutches may vary between individuals and years, with some individual diets likely dominated by penguin in some years. Aims. We investigated raven energy assimilation in the world's largest little penguin colony using stable isotopes. Methods. 813C and 815N were extracted from feathers of marked individuals and potential food sources (marine and terrestrial items) to explore the energy assimilated by individual ravens. Key results. Ravens had large variation in food resource use. We found dietary differences with age and year, with the proportional contribution of dietary sources varying more in juveniles and fledglings compared with adults and differing between sampling years. Penguin, mouse and other mammals contributed substantially to the energy assimilated by ravens. There was no evidence of variation in penguin consumption with age or year. This was true for both 'culprits' and 'non-culprits' (i.e. those that had not been observationally detected consuming penguin), which showed little difference in their dietary composition and little dietary specialisation between individuals. The invasive house mouse Mus musculus formed a prominent component of raven diet in some but not all years. Ravens prey on penguin eggs and chicks, and persist within penguin colonies where they consume a variety of food types. Conclusions. The energy assimilation we describe infers a generalist diet, which in the years and birds we sampled included conservation-sensitive prey in the form of penguin. Implications. We failed to detect a subset of individuals which specialised on consuming penguin. Selective removal of problematic individuals therefore appears unfeasible.
Context. Mammals inhabiting arid and semi-arid regions are often exposed to extreme conditions and may rely on refuges for survival. The greater stick-nest rat (Leporillus conditor) is a threatened, nest-building rodent that once occupied large parts of semi-arid Australia. The species became restricted to two offshore islands, but has now been translocated several times, with translocation failures attributed to predation, unsuitable habitat and drought. Aims. We studied a translocated population of greater stick-nest rats at Australian Wildlife Conservancy's predator-proof fenced area at Mt Gibson Wildlife Sanctuary in Western Australia. We documented the population trajectory over time, characterised and modelled the thermal properties of above-ground stick-refuges and recorded data on habitat use. Methods. We used live trapping data to estimate population abundance over time. We documented the structure and thermal and humidity profiles of stick-refuges and conducted vegetation surveys comparing areas inhabited by greater stick-nest rats to adjacent uninhabited areas. Results. We found that: (i) greater stick-nest rat abundance declined to a barely detectable level during a severe hot drought; (ii) internal refuge temperatures were typically cooler than external temperatures, but the buffering effect was ineffective on extremely hot days; and (iii) stick-nest rats inhabited sites with a relatively high abundance of salt-tolerant shrubs, which are highly susceptible to drought and of which there is limited availability within the fenced area. Conclusions. Extreme heat wave conditions and drought were related to a sharp population decline. Stick-refuges occupied by greater stick-nest rats at our study site provided limited thermal buffering, and the drought likely reduced the availability of food resources. As the extent of salt-tolerant shrubland is limited within the fenced area at Mt Gibson, a lack of suitable habitat may have contributed to the observed population decline during the hot, dry period. Implications. Our findings suggest that regions with more moderate climates and larger areas of suitable vegetation may be better suited as translocation sites for this species in the future. More broadly, our study highlights the need for further research on how Australia's rodents will respond to elevated temperatures and increased drought as climate change progresses.
Context The ability of animals to exhibit homing behaviour, returning to familiar locations by using spatial memory, is central to understanding movement ecology, conservation planning, and the success of reintroduction programs. Despite this, little is known about the homing behaviour of mature captive-reared crocodiles once released into the wild.Aims This study investigated post-release movements of captive-reared, wild-caught, and translocated saltwater crocodiles (Crocodylus porosus) in Bangladesh's Sundarbans mangrove ecosystem, with the aim of assessing homing tendencies and how movement patterns differ among release types.Methods Five mature crocodiles were monitored, including three captive-reared individuals, one wild-caught and released, and one wild-caught translocated animal. Satellite telemetry and continuous-time state-space modeling were applied to estimate daily movements and home-range sizes across tracking periods of 53-127 days per individual.Key results Movement patterns varied substantially among individuals. The translocated crocodile exhibited the greatest daily movements (up to 34.3 km/day) and an expansive home range (2515 km2), consistent with previously documented strong homing and exploratory responses to translocation. In contrast, captive-reared and wild-caught released crocodiles moved shorter daily distances and maintained smaller, more defined home ranges (14.2-125 km2). Importantly, captive-reared crocodiles showed no indication of homing towards prior husbandry sites, despite long-term captivity.Conclusions Mature captive-reared crocodiles were able to establish ranges in the wild without showing fidelity to former captive or husbandry sites. In contrast, the pronounced exploratory behaviour of the translocated crocodile underscores the challenges associated with adult translocations.Implications These findings highlight the potential of releasing subadult and adult captive-reared crocodiles as a viable conservation strategy to bolster wild populations while avoiding the high predation risks of juvenile releases. At the same time, they emphasise the importance of carefully considering release sites and incorporating measures such as soft-release protocols to mitigate homing behaviour in translocated animals. The results provide an evidence base for refining crocodile reintroduction and restocking efforts across their range.
