Trish Fleming

The use of road underpasses by feral cats (Felis catus) and native wildlife. Context: Underpasses are incorporated into roads to encourage animal movement across fragmented landscapes. However, observations of introduced predators – such as feral cats – has raised speculation that they target underpasses for hunting native species and utilising shelter. Aims: The study examined (1) if feral cats and native wildlife used underpasses, (2) which underpass characteristics were correlated with more animal use, and (3) if underpasses could act as ‘lures’ for targeted cat control. Methods: Located at Alcoa’s two bauxite mines in Western Australia, 64 motion camera traps were deployed across 16 road underpasses over a period of four months. Four cameras were set up at each underpass; one on each end of the underpass, each paired with another camera 30m away in the adjacent bush. Key results: Underpasses were used by native species, including echidna, chuditch, rakali, and mardo to cross beneath roads; quokkas and bandicoots were detected at the edge of underpasses, but rarely entered. Observations of feral pigs and foxes using underpasses to pass beneath roads were also recorded. Additionally, 14 individual cats were identified using underpasses as a thoroughfare; one cat was observed regularly at underpasses approximately 5km apart, and another cat took refuge inside an underpass out of the rain for 49 minutes. Conclusions: While the study provided no direct evidence of predation by feral cats at underpasses to confirm the ‘prey-trap hypothesis’, it also could not be dismissed due to frequent feral cat detections at underpass sites, the regular return of individuals, and extended periods of time spent inside underpasses. The study successfully identified underpasses as potential locations for targeted cat control for more effective management of feral cats for the conservation of Australian biodiversity.

Behavioural responses of dingoes (Canis familiaris) to camera traps in Western Australia. Context: In Western Australia, dingoes are an ongoing threat to grazers and farming communities due to direct (livestock predation) and indirect impacts (e.g., mis-mothering, keeping livestock from key resources). Therefore, landholders seek to suppress dingo populations through management practices. To quantify the success of control activities, monitoring population responses through camera trapping is increasingly important. Camera traps are generally assumed to be a non-invasive monitoring method; however, if dingoes alter their behaviour around camera traps, becoming more wary of the devices, they could subsequently avoid detection, which would influence density estimates and potential management of the species. Aims: This study investigated the behaviour of dingoes towards camera traps on dingo managed properties across Western Australia, to determine if their behaviour was influenced by reproductive seasons, time of day and positioning of the camera trap. For a subset of the data, images were identified to individual dingo, allowing assessment of individual behavioural responses. We hypothesised that there would be greater detectability of dingoes exhibiting trap-wary than trap-happy behaviour during the breeding seasons, at nighttime, and with the camera position angled down tracks. Methods: Dingo interactions were scored for 180 camera traps deployed on roads across properties in the Gascoyne, Murchison, Northern Agricultural and Nullarbor regions between 2020 and 2023. For each independent trigger event, dingo responses were categorised and scored as curious, cautious and neutral behaviour. Key results: Of the 2446 detection events, dingoes exhibited more neutral behaviour than either curious or cautious behaviour. A first model testing all dingo detections found a significant negative effect on the behavioural score during the Pre-Breeding season (p=0.015). A second model testing a data subset including individual identification of dingoes found a significant positive effect on the behavioural score during the Whelping season (p=0.029). Neither time of day (model one, p=0.187; model two, p=0.802) nor camera orientation to the road (model one, p=0.358; model two, p=0.957) had any significant impact on dingo interactions with camera traps. There was significant variability amongst individual dingoes in their behavioural responses (ICC = 0.319).

Conclusions: We surmise that camera traps do not significantly influence overall dingo behaviour across controlled regions of Western Australia. Variability of behavioural score suggests individual dingo responses differ based on previous exposure to camera traps, human activity and experiences with lethal management. Implications: Results indicate behavioural responses influence the detectability of dingoes particularly across different periods of the year. Variation in individual responses should be considered in future monitoring designs and baiting events for lethal control management.

Can fire fighting water points be used to monitor feral pig presence? Feral pigs (Sus scrofa) are one of the most significant mammalian pests in Australia. Feral pigs disrupt ecosystems and can transmit diseases, threatening environmental, agricultural, social and cultural assets. The Warren Bioregion of Western Australia hosts a range of endemic species and communities threatened by feral pigs. Throughout this region, a series of fire-fighting water points (i.e. natural and artificial reservoirs used for firefighting) provide a perennial source of water in an otherwise water-limited landscape, making them potentially valuable resources for fauna, including water-dependent feral pigs. To investigate the occurrence of feral pigs at fire water points, we deployed camera traps at 57 fire water points from May to September 2024. We assessed the association between camera trap rates of feral pigs and native vertebrates with site characteristics: canopy cover, understory density, vegetation structure, leaf litter, percentage of native vegetation for a 2km-radius buffer, and rainfall. Over 6,580 trap nights (average 92.1 ± 45.0 nights), feral pigs where only detected at 13 of the 57 sites (30 independent detections). No significant associations were identified between the trap rate of feral pigs and site characteristics. Rainfall was negatively correlated with western grey kangaroo (Macropus fuliginosus) and emu (Dromaius novaehollandiae) trap rates. The very low trap rate of feral pigs during this study resulted in an inability to conclusively assess factors driving their presence or activity. The measured site-based environmental factors did not explain the variation in feral pig occurrence across the study area, suggesting that other environmental or anthropogenic variables may determine their habitat use. Further research should prioritise larger time scale data collection to observe seasonality of water point use. The current study highlights the importance of local assessments of pig presence, rather than predictive approaches.