The house cat, Felis catus, was introduced into Australia with European settlement of the mainland. Since its initial introduction, it now occupies all mainland habitats, Tasmania and many smaller offshore islands. Large numbers of cats were released intentionally into the environment in a misguided attempt to control the spread of other introduced mammalian pests, especially the European rabbit, Oryctolagus cuniculus. The feral cat is an invasive predator that has been implicated in the decline and extinction of many species of native small mammals across Australia, particularly in the arid regions and on offshore islands. Much of the research on feral cats in Australia has occurred in the continent’s arid and semi-arid regions. Consequently, little is known about the ecology of feral cats in tall forests. Additionally, the most generally effective population control technique, poison baiting with sodium monofluoroacetate (compound 1080), has wide ranging applicability in arid and semi arid areas but its use is restricted in the temperate and forested eastern states of Australia due to concerns about impacts on non-target species.
This thesis is divided into three parts. Firstly, I review the current knowledge of feral cats, particularly in relation to the actual and potential impact they have on native prey species. Secondly, I investigate the ecology of the feral cat in the temperate tall forests of Far East Gippsland, Victoria. The home range sizes, movement patterns and home range use of feral cats were determined. Thirdly, I examine a new technique for delivering poisons in a feral cat management program. The potential for all Australian non-target species to access the toxicant is examined using a desktop analysis, while field studies examine uptake by non-target species and the dynamics of prey species to determine acceptable times for baiting campaigns.
GPS and VHF collars were utilised to obtain fix data for feral cats in Far East Gippsland. Male cats had significantly larger home ranges (MCP100 455 ± 126 ha) than females (105 ± 28 ha), with male home ranges overlapping those of females. Some female home ranges overlapped extensively, with neighbouring females also having overlapping core areas within their ranges. These overlaps in female home ranges, in particular of the core areas, indicate that female cats in Far East Gippsland are tolerant of other females and do not actively exclude them.
Compared with the home ranges of feral cats in other regions of Australia and New Zealand, the cats in Far East Gippsland had smaller home ranges than those of cats occupying arid and alpine zones yet larger ranges than those of feral cats living in farmland or grassland. This variation probably reflects the availability of food resources, with cats in resource-poor areas requiring larger home ranges and cats with smaller home ranges generally inhabiting areas with greater, or more accessible, food resources.
The use of GPS collars to obtain accurate and high volumes of location data allowed the intra-home range movements of feral cats to be examined in ways not previously possible using conventional VHF radio telemetry. Location data were gathered at three different temporal intervals – 6 hourly, hourly and every 15 minutes. Feral cats followed a Lévy walk-style searching pattern as they moved through their home range. Employing a Lévy walk increases the likelihood of encountering prey items that are distributed sparsely in the environment, in turn maximising the potential hunting return for effort expended.
Each of the cats examined had large areas within their home range that they did not enter. To test the hypothesis that this resulted from a scarcity of prey in these areas, trapping grids were established to capture small prey-sized animals. There was no difference in the rate of capture of prey species in the areas of high and zero cat use, thus allowing the food hypothesis to be rejected. Modelling of abiotic environmental parameters was used to determine if these influence home range use. While the models explained much of the variation in the data, the global model was overdispersed, indicating that other unmeasured parameters were influencing home range use. The avoidance of these areas most likely arises from the presence of larger intraguild predators and subsequent employment of predator avoidance strategies by the cats.
Managing the abundance of feral cats using poison baiting requires that bait be distributed at times when cats are food-stressed. Generally this occurs in winter when prey species are in natural decline. To determine the most appropriate time for baiting feral cats, trapping grids were established to assess the population demographics of feral cat prey species. The 2 046 trap nights undertaken resulted in 176 captures of five prey-sized species. The breeding periods for the Antechinus spp. occur earlier in Far East Gippsland than would generally be expected based on the latitude and altitude of the trap sites, and have bearing on the optimal time for poison baiting. Based on these findings, the optimal time to manage feral cat populations in Far East Gippsland through poison baiting is between late August and mid November provided that the toxicant is enclosed within a hard shell delivery vehicle (HSDV) that maintains structural integrity or, alternatively, if the baits are suspended above the ground surface and out of reach of lactating female antechinus. Further research is proposed to supplement these findings.
Encapsulation of toxicants within an acid soluble HSDV which is then inserted into the bait media is being explored as a potential technique to minimise access of non-target species to the toxicant. A desktop analysis employing a decision tree process was used to examine the potential for non-target access to toxicant delivered in an HSDV. This analysis encompassed all non-aquatic vertebrate species in Australia. significantly fewer species would be susceptible to non-target poisoning if HSDVs were used when compared with directly injecting the toxicant into the bait media. Carnivorous mammals were the most likely to consume both the bait and the HSDV.
Using the systemic marker, Rhodamine B (Rb), in the HSDV, the ability of five species of small to mid-sized animals to access toxicants enclosed in the HSDV. This was compared with directly injecting it into the baits. Rhodamine B staining was apparent in the mystacial vibrissae of four of the five species at sites where Rb was injected into the baits. It was also present in three of the four species captured at the sites where the Rb was encapsulated within the HSDV. The longevity of the HSDV within the bait media was tested and found to decreased rapidly following insertion into the bait. This is most likely due to the bait media being slightly acidic. Since that experiment concluded, changes have been made to the pH of baits to extend the integrity of the HSDV and hence reduce leakage.
These key findings will allow managers to adopt a more targeted approach when undertaking cat control programs in these habitats. The use of GPS technology to obtain location data has allowed the analysis of intra-home range movements to an extent previously not possible with other techniques. This in turn will allow a more targeted approach to managing feral cats. The use of a decision tree approach to determining the susceptibility of non-target species during a baiting campaign can be applied to other poisoning campaigns regardless of the target species or the toxicant being used.