Conservation introduction of top predator to an island triggers ecological cascades

Top predators play important roles in structuring ecosystems and protecting smaller biodiversity by suppressing herbivorous prey and smaller predators. This suppression is achieved through a combination of lethal (e.g. predation, intra-guild killing) and non-lethal (e.g. behavioural, competition) risk effects. Loss of top predators can trigger trophic cascades, which may result in detrimental effects on biodiversity. Reduced predation and competition by top predators can lead to increased abundance of prey and mesopredators, leading in turn to increased herbivory and predation on smaller prey species. Restoration of top predators to ecosystems is a potential tool for conserving biodiversity and controlling the impacts of invasive species. However, long-term manipulative experiments investigating the effects of top predators on ecosystems and their impacts on invasive species are rare. A natural experiment involving the removal of top predators is occurring on the island state of Tasmania, Australia, initiated with the extinction of the Thylacine (Thylacinus cynocephalus) in the 1930s. Tasmania's current top predator, the Tasmanian devil (Sarcophilus harrisii) has severely declined in the last 25 years due to a novel transmissible cancer, devil facial tumour disease (DFTD). Since it was first detected in 1996, DFTD has spread to most of the devil's distributional range, causing an overall population decline of 80%, with local declines up to 95%. Field studies suggest there has been an increase in abundance of two invasive species at different trophic levels, the feral cat (Felis catus) and black rat (Rattus rattus) in areas where DFTD has been present for long periods, and this may have caused declines in some small and medium-sized mammal species in long-diseased eastern Tasmania. There is debate, however, as to whether feral cats have responded to devil decline with a change in abundance or behaviour, and the mechanism responsible for these changes in cats and black rats is unclear. There is also evidence of an increase in ground-foraging behaviour in a native omnivore (the common brushtail possum, Trichosurus vulpecula), which is a major prey species for the devil, but changes in the distribution and abundance of this species in response to devil decline are largely unknown. Concern that the devil may be driven extinct by DFTD led to the assisted colonisation of devils to Maria Island, off the east coast of Tasmania to establish an insurance population of healthy devils separated from the disease. The island has resident populations feral cats, black rats, and common brushtail possums, all of which were previously introduced to the island. Also, the island supports several species of burrow-nesting seabirds that are potential seasonal prey for all of these introduced predators and omnivores. There are no other native or introduced terrestrial predators on Maria Island. In this thesis, I use the assisted colonisation of devils to Maria Island to test hypotheses about the ecological effects of increasing devil abundance on populations of feral cats, possums and black rats. A nearby site on the adjacent Tasmanian mainland served as a control site for the study, which maintained a very low and stable devil population. Further, a second predator-free island was also used as a control site for the work on seabirds. The hypotheses that I tested for this study were: First, that devils would alleviate predation pressure on a resident population of burrow-nesting seabirds, the short-tailed shearwater (Puffinus tenuirostris), by supressing cats and possums that prey on the species; Second, that an increase in devil abundance would reduce the abundance, distribution and individual survival of possums; Third, devils would either reduce the abundance of both cats and black rats as suggested in previous field studies. Or, devils would reduce the abundance of cats, and this would in turn result in the predatory release of black rats on Maria Island. Key results are as follows. Increasing devil abundance was associated with reduced cat and possum activity on shearwater colonies. Cat and possum activity declined on the shearwater colony and no possums were recorded at the colony when devil abundance was high. Devils preyed on possums while avoidance of devils at fine spatial scales seemingly decreased the hunting efficiency of cats. However, devils were more adept at killing shearwaters than either cats or possums, particularly because of the devils' ability to dig up burrows. Establishment of devils was followed by extirpation of the shearwater colony within four years, rejecting the first hypothesis that devils would alleviate overall predation pressure by suppressing cats and possums. Devils competed with cats for shearwaters, and at high abundance they may have restored the 'landscape of fear' for cats and especially possums. Devils reduced the abundance and individual survival of possums, resulting in abrupt shifts in habitat use to avoid potential conflicts with devils, confirming the second hypothesis. Devils killed and ate 50% of the GPS-collared possums that used open grassland habitats resulting in reduced possum density in open habitats. Surviving possums used structurally complex, dry forest habitats where they could avoid and escape predation by devils, thus increasing survival. Overall, possum abundance declined on Maria Island as the abundance of devils increased, in contrast to the control site where devil abundance was low and both possum and devil abundance remained stable throughout the study period. Predatory pressure from devils, through direct predation and non-lethal effects of risk of predation, determined the abundance and distribution of possums, highlighting the mechanisms and importance of predators in structuring ecological communities. Increasing devil abundance on Maria Island was associated with reduced abundance and recruitment of cats. Cat abundance almost halved on Maria Island as devil abundance increased, but remained stable at the control site where devil abundance was low and stable. Increasing devil abundance on Maria Island was also associated with reduced detection rate of black rats (possibly due to reduced abundance), in contrast to the control site where detection of black rats did not change. Hence, this result supported the first but not the second part of my hypothesis that suppression of the cat population by devils would result in the predatory release of black rats. Declines in abundance of both cats and black rats may result from exploitative and interference competition with devils for resources, such as seabirds and carrion. Devils may prey on kittens and young cats, and non-lethal effects could include changes in risk-sensitive behaviours in cats and black rats, such as spatial and temporal avoidance of devils. Overall, results of this large-scale and long-term manipulative study adds to the growing body of evidence that demonstrates the potential for top predator reintroductions as a management tool to limit overabundant herbivores and to mitigate the detrimental impacts of invasive predators on biodiversity and ecosystem function. The results of the study also highlight the conservation trade-offs that need to be considered when introducing a predator to an isolated ecosystem such as an island or where they have been absent for a long time.