Jill Shephard

Can the flight dynamics of the Carnaby’s cockatoo be predicted from its morphology? In pursuit of the 2050 zero-emissions target, renewable energy, notably windpower, has emerged as a key player in the quest to reduce carbon emissions. However, the rapid expansion of wind energy has raised pressing concerns, particularly regarding its impact due to turbine strikes on flighted species, especially those classified as threatened/endangered or with long lifespans. Collision risk modelling can be used to predict turbine strike risks, but this approach heavily relies on species-specific flight data, which is not always readily available. This underscores the pivotal role that theoretical flight data can play when GPS derived flight data is unavailable. In this thesis, I compared the wing morphology and flight capabilities of Carnaby's cockatoos (Zanda latirostris) in Western Australia, compared theoretical flight data generated by the 'afpt' R package and equivalent data derived from GPS tracking and discussed the results in the context of collision risk modelling. I describe the distinct wing morphology of the Carnaby's cockatoo from wing traces of 48 individuals, which was characterised by relatively short and wide wings with pronounced slotting. Notably, juveniles featured thinner wings with shorter wing widths and higher aspect ratios compared to their adult counterparts. These significant morphological differences translated into differences for the model-generated data, with juveniles exhibiting significantly lower speed and lower powered flight performance. There was substantial variation between the model-generated flight data and the observed flight data for 11 individuals where both sets of data could be calculated, particularly in terms of flight speed and energy expenditure; greater sample size could add to these analyses. This research deepens the understanding of Carnaby's cockatoo flight behaviour and highlights the potential of theoretical flight data in enhancing collision risk modelling.

The distribution, behaviour and environmental impact of feral pigs in the Kimberley. Feral pigs (Sus scrofa) were first documented in the Kimberley Region almost a century ago, and were reported as occurring in small, isolated patches of habitat, at the mouths of the Fitzroy and Isdell rivers. In the time since, they have slowly increased in population size and geographical extent. Today, feral pigs are reported across important conservation areas in the central Kimberley, and their continued expansion triggered the current investigation. This thesis includes a literature review and two research chapters:

• Study 1 made use of 1,437 known locations for pigs over a 10-year period from 2009 to 2019 to model the potential distribution of feral pigs under current climatic conditions in the Kimberley Region. Modelling used MaxEnt to create a set of models which explicitly included or excluded measures of survey bias to examine the effect on distribution predictions. In this case compensating for survey bias increased Maxent’s ability to discriminate suitable from unsuitable habitat and showed that pigs may potentially increase their distribution range by up to 27%.

• Study 2 analysed satellite telemetry data collected for 24 feral pigs across five sites in central and north Kimberley to investigate sex, body size and environmental influences on range area for feral pigs in the Kimberley. Male pigs had larger home ranges than females. Maximum temperature was a significant driver of male home range size, with smallest home ranges for the hottest months (around August), when feral pigs were most constrained to permanent water ways.

Seasonal variation in home range size can inform effective management of this species, identifying the time of year when feral pigs are most constrained to water and therefore have restricted movements. Findings of this study contribute to the existing understanding of feral pig distribution and behaviour in the Kimberley and should prove a useful tool for land managers.

Movement ecology of Australia’s avian apex predator: the Waalitj / Warlawurru (Wedge-tailed Eagle Aquila audax). Almost half of the world’s 75 species of eagle are threatened, and in nearly all cases the causes are anthropogenic. The Wedge-tailed Eagle Aquila audax, Australia’s largest eagle, is widespread despite hundreds of thousands being culled during a century of European persecution. This project investigated the ecology of two breeding ‘populations’ occupying different climatic zones (‘arid’ and ‘mesic’), by monitoring 11–98 breeding pairs annually between 2009–2019, and employing satellite telemetry to better understand movements and behaviour. The two populations supported 44 and 54 evenly spaced territories with an average of 93% occupied by breeding pairs. Mean annual productivity of 0.77 fledglings/pair/year in the mesic zone was significantly greater than in the arid zone (0.13; the lowest value recorded for the species). A comparison of methods for estimating home range size involving analysis of data from GPS/satellite tracking and nest distribution revealed telemetry-derived (auto-correlated kernel density) estimates averaged 15.8 and 15.9 km² for arid and mesic eagles, respectively. These were not correlated with estimates derived from nest distribution data, and were smaller than previously determined by spatial methods. Home range size was not related to estimated prey abundance or climate. Twenty-two juvenile eagles were tracked with GPS/satellite transmitters. The duration of the nestling period was 77–114 days and males fledged significantly earlier than females. The post-fledging period was 102–221 days and there was no detectable difference between the sexes. Eagles left their natal areas between 16th March and 3rd June, with most immediately dispersing long distances, covering 192–2,212 km/month, with cumulative distances of 11,230–27,093 km in the first year of independence. Using banding records and data from tracked individuals, survival was 96% as nestlings, 83% during the first year, and 97% and 80% in the pre-adult and adult years, respectively. These findings provide up-to-date information on the status of the Wedge-tailed Eagle, and show this important apex predator persists in varied environments and roams widely over large expanses of country, which allows it to avoid or overcome some of the anthropogenic threats other species face in Australia’s ever-changing landscape.

Navigating the urban matrix: activity patterns, home range dynamics, and habitat preferences of the Western Australian quenda (Isoodon fusciventer). Background. As urbanisation increases globally, it becomes critical to preserve and maintain wildlife in cities, for human health benefits and to maintain essential ecosystem services. The increasing urbanisation of Perth, Western Australia, has significant implications for a local urban adapter, the quenda (Isoodon fusciventer), a critical-weight-range, ground-dwelling marsupial. Aims. My research, initiated and supported by the City of Mandurah, explores how urban gardens, habitat preferences, and interactions with domestic pets influence quenda behaviour. Methods. I used a combination of camera trapping and GPS tracking to identify resource use by quenda for residential gardens and urban bushland reserves. Key findings. Residential gardens provide supplementary food sources, which alter quenda activity patterns and potentially increase their injury risk due to vehicle strike or predator encounters. Despite the predation risks posed by domestic pets, particularly cats, camera trapping and GPS tracking indicates that quenda do not appear to modify their behaviour to avoid these predators, highlighting the importance of pet management strategies to reduce injury risks. Habitat selection from GPS location and movement data show that quenda favour areas with native vegetation and dense undergrowth, underscoring the need for urban planners to prioritise the conservation and integration of such habitats within urban landscapes. Additionally, the selective use of bushland corridors and road verges by quenda suggests that these areas play a crucial role in maintaining connectivity between habitat patches. Conclusions. The implications for urban landscape management are clear: strategies must incorporate the preservation and enhancement of suitable habitats, the regulation of pet interactions with wildlife, and the careful design of urban spaces to support the ecological needs of species like the quenda.

The south-west of Western Australia (SWWA) is home to three species of threatened endemic black cockatoos. Anthropogenic activities including urbanisation and agriculture have led to the loss of at least 70% of pre-colonial native vegetation and much of what remains exists in a highly fragmented state within a matrix of human modified landscapes. The black cockatoos of SWWA interact in varying degrees with these modified landscapes, with consequences to fitness that are largely unknown.

In this project, I used modern tracking technologies to study the ecology of Carnaby’s cockatoo (CC) and the forest red-tailed black cockatoo (FRBC) in modified habitats. Using GPS units with integrated accelerometer devices, I gained insight into the fine-scale habitat use of these birds, as well as informing the interaction between habitat, behaviour, and energy expenditure. In addition, I analysed foraging resources for energetic content to further inform patterns of activity revealed by tracking data.

FRBC were tracked in both their natural forest habitat and the urban environment into which they have recently expanded. Urban birds were found to expend 25% more energy per day, with little evidence that this was adequately offset by foraging on exotic foods of higher calorific value. CC were also tracked in the urban area where birds spending more 􀆟me utilising the smallest fragments of remnant vegetation were found to travel significantly further each day. Daily distance travelled was in turn associated with greater energetic expenditure. Breeding CCs were tracked in the wheatbelt where the recent introduction of canola to the diet has been associated with improved reproductive outcomes. The existence of an upper limit to canola foraging was discovered, even when high temperatures reduced time available for foraging on native vegetation.

The research undertaken for this thesis has added to our understanding of the ecology of both species and the fitness consequences imposed on each by interaction with modified landscapes. The information herein will contribute to the conservation of Carnaby’s cockatoos and forest red-tailed black cockatoos by directly informing their respective Recovery Plans.

Historically it has been difficult to gain information on the movement ecology of psittacine species in Australia. Using a novel double-tagging telemetry method, this research, aimed to: investigate regional differences in movement of the three black cockatoo species endemic to Western Australia; identify key roost and foraging sites for these species across regions; and estimate home range sizes for flocks in resident areas, using a combination of GPS and satellite PTT tags.

Tagged birds served as markers of flock movement once integrated into a wild flock of conspecifics, which was confirmed through means of behavioural change point analysis and field observations. Linear mixed models were used to determine differences in movement across regions, revisitation analysis was used to identify key habitat sites, and an auto-corrected Kernel density estimator was used to estimate the home ranges.

Results showed that key roosts sites for the three species predominantly occurred on public green space and private property. These were closely associated with foraging habitat which mainly occurred as remnant vegetation in the landscape or as nature reserves. Riparian zones and roadside vegetation were shown to play a crucial role as foraging habitat and in providing connective landscape structures. Daily movement distances differed both between and within regions depending on habitat matrix, resulting in varying home range sizes. These results suggest that movement for the three black cockatoo species is region specific, driven by food resources in the landscape. In addition, between species, movement varied as each species uses the landscape in different ways, depending on seasonal movements and ecological requirements.

This research has provided critical baseline data required to address knowledge gaps listed in Recovery Plans for these species of black cockatoo. Further research is now required to include these data in resource and habitat selection models to identify how the landscape matrix affects movement, which will facilitate adaptive habitat management and conservation plans for black cockatoos in Western Australia.

The three black cockatoo species endemic to south-west Western Australia – Carnaby’s cockatoo (Calyptorhynchus latirostris), Baudin’s cockatoo (C. baudinii) and forest red-tailed black cockatoo (C. banksii naso) are threatened and have Recovery Plans guiding conservation efforts. Threats include habitat loss due to land clearing for urban, agricultural and industrial development; competition with other species for nest hollows; poaching; disease; vehicle-strike and illegal shooting.

This research built on previous black cockatoo research with an overall aim to develop and validate reliable methods to track all three species, to gain insight into their movement, distribution, habitat use, activity and behaviour.

In an initial proof of concept trial, we attached tail-mounted tags to two Baudin’s cockatoos. Both birds were successfully tracked for several months after release, demonstrating satellite telemetry can be used to locate and track forest species.

We then developed a double-tag mounting protocol to attach a tail-mounted ARGOS PTT satellite tag and back-mounted solar-powered UvA-BiTS GPS tag to captive black cockatoos. The combination of UvA-BiTS back mount and ventral tail mounted Telonics tags was the best tolerated and provided excellent GPS and ARGOS satellite location data with no interference between the two types of tag.

The focus then moved to the development of an automated classifier tool that used accelerometer data from UvA-BiTS GPS tags to remotely identify behaviours and calculate activity budgets. Using accelerometer data from 15 birds post-release, we determined black cockatoos spend most of their time at rest, interspersed with foraging activity through the day and some movement between roost sites and feeding habitat.

To maximise the retention time of tail-mounted tags, the tail feather life span and time of moulting was studied using moulted tail feathers from captive cockatoos and tagged bird post-release. Captive cockatoos had a mean feather lifespan of 410 days, suggesting tail feathers do not always moult annually. Peak tail feather moulting occurs from December to March, the non-breeding period. The optimal time to attach tail mounted tags is from May to September.

The development and optimisation of tracking methodologies for use on black cockatoos has facilitated the tracking of all three species in the wild. This research has provided data which have enabled identification of key roosting, foraging and breeding habitat and determination of flock movement patterns and habitat use at a landscape scale across the species’ distribution ranges. This information is being used to guide black cockatoo conservation management in relation to habitat protection and restoration.