Establishment, behaviour and ecology of the SA mainland tammar wallaby (Macropus eugenii eugenii) following an experimental reintroduction.

The South Australian mainland tammar wallaby (Macropus eugenii eugenii) was presumed extinct in the wild from the early 1930's, until a feral population was re-discovered in New Zealand. Eighty-five animals were returned to Australia as part of a repatriation program into their former range. The establishment phase after a reintroduction is a critical time as animals may fail to survive if they cannot find resources and avoid predators in an unfamiliar habitat. To maximise reintroduction and establishment success, reintroductions need to be planned with a good understanding of the animals’ ecology and anti-predator strategies. To improve this understanding requires experimental reintroductions and detailed monitoring. This thesis investigates the experimental reintroduction of 46 wallabies into Innes National Park in South Australia and examines the influence of release group familiarity on establishment. Part of this was an investigation of home range, habitat requirements and social behaviours during establishment and seasonally post-establishment. The thesis includes three data chapters which focus on (1) home range and core area, home range stability and degree of overlap with conspecifics; (2) habitat selection at the landscape scale and for day and night use within home range; and (3) factors influencing fine scale habitat use and social grouping behaviours in light of predation risk. In this study, the animals’ perceived risk of predation is assumed to be reflected by surrogate measures of risk, such as distance to cover, likelihood of using cover, group size, and distance to nearest neighbour. Release group familiarity was established by housing animals together in captivity for at least one month prior to release (“familiar” groups), whereas “unfamiliar” release groups comprised animals housed separately but released together. After each reintroduction the first month was considered to be a time of “establishment” in the environment. During the establishment month, home ranges were not randomly located within the landscape, as indicated by the biased occupancy of particular habitat types. Habitat types selected at the landscape scale were similar for familiar and unfamiliar release groups. However, animals released in unfamiliar groups showed a stronger preference for denser high cover habitat during their nocturnal activities within their home range. As tammars use cover to conceal themselves from predators, this result suggests that animals released in unfamiliar groups were more cautious than those released in familiar groups. Indeed at the fine scale, it was also found that animals released in unfamiliar groups were more likely to be found in high cover habitat, and forage closer to cover at night than did those released in pre-established familiar groups. Using habitat with more caution and capitalising on communal vigilance in an unfamiliar habitat may ultimately improve the likelihood of survivorship and overall reintroduction success. Comparing habitat decisions and social behaviours during the establishment period to similar times of year post-establishment suggested that animals' naivety about their new environment influenced some decisions they made. While habitat selection at the landscape scale was similar during establishment and at an equivalent time of year post establishment, analyses showed that they preferred to use melaleuca (a high cover habitat) during their nocturnal activities during establishment much more strongly than they did once they had established. Home range and core areas were also significantly smaller during the establishment month than at an equivalent time of year post establishment. This result supports the idea that animals will restrict their movements when they are unfamiliar with the habitat and predation risks. It also suggests that some habitat choices improved once they were familiar with their new environment and presumably predation risks. Their habitat choices reflected better anti-predator behaviour than those made during the establishment period: they were more likely to use high cover habitat, they remained significantly closer to cover while foraging, and group sizes were larger than during establishment. Seasonal habitat selection at the landscape and home range scales suggested that the five habitat types within the study area provided different fundamental resources for the animals, as they were preferred at different times of year. Some differences in habitat selection were observed between the sexes, and the females were more selective in their diurnal and nocturnal activities. These differences most likely reflected, in part, differences in predator avoidance and reproductive strategies of the sexes, where females' preference at the landscape scale shifted towards high cover habitat during spring, the time of year when pouch young vacate the pouch and start to become independent. While no such selection at the landscape scale was observed for the males, it was also observed that within their daily activities both sexes were more likely to be found in high cover habitat during spring than any other time of the year, perhaps suggesting that as males were following the females. Overall, females generally selected Eucalyptus diversifolia, Acacia anceps and grassland at the landscape scale, and used E.diversifolia for refuge during the day and the other two habitats for foraging at night, whereas males generally preferred Melaleuca halmaturorum instead of E.diversifolia for diurnal refuge. Eucalyptus rugosa was mostly avoided by both sexes. From month to month, both sexes expanded rather than shifted their home ranges to incorporate new areas, and these new areas were explored with conspecifics (when the amount of new area increased the amount of sharing also increased), highlighting their reliance on communal vigilance when in unfamiliar habitat. More new areas were incorporated into home ranges from July to December than from January to May. Time since release was not influential, which also supported the conclusion that perhaps home ranges moved to follow resources or overlap conspecifics more. Indeed, in one circumstance when neighbouring animals had died, a male wallaby was observed to move four kilometres through unfamiliar habitat and completely shift his home range in search of other residence. Compositional analysis of habitat use versus availability indicated that monthly home ranges were selectively positioned in the landscape and were always larger than 4ha. Males' home ranges were larger than females', and males shared more of their home ranges than females did, supporting the usual sex bias observed for polygamous species. Core areas were proportional to the size of the home range, with similar sizes held by males and females and throughout the year. Core size was not influenced by the degree of overlap with conspecifics, with similar amount of core area shared by both sexes year round. The time of year influenced home range size, the smallest were held in winter when food resources were likely to be most abundant, but also when inclement weather was likely to restrict movements, as the animals' ability to detect predators may be hindered due to wet and windy conditions. Living with conspecifics is known to assist predator detection by group vigilance. Indeed, this study found the amount of home range overlap and the time two individuals spend together was positively correlated, and the size of home ranges decreased when more of it was shared with conspecifics, which suggested that sharing of home range was important. The degree which home ranges were shared was observed to be a fairly stable requirement for both sexes and did not change with season or time since release. Despite previous isolation from predators, the wallabies displayed anti-predator behaviours which incorporated interrelated benefits obtained from group vigilance and using protective cover. Additionally, these behaviours were adjusted according to their familiarity with the habitat. Post-establishment, animals were observed to go further from cover when they were a greater distance from their nearest neighbour but surrounded by larger numbers of conspecifics. Whereas during the establishment period, animals ventured further from cover when they were closer to a nearest neighbour, but group size was not influential. It is known that larger groups of animals have more false alarms to predators, and false alarms result in the animals‟ wasting energy in fleeing. If false alarms are more prevalent while occupying unfamiliar habitat with unfamiliar risks, then relying on large numbers of conspecifics while establishing may have been more of a liability than a benefit during the establishment period. However, some anti-predator strategies were commonly used, regardless of familiarity with their habitat. During the establishment month and post-establishment, animals were always more likely to be found in high cover habitat when they were further from their nearest neighbour, or were surrounded by fewer conspecifics. Some strategies and habitat decisions may have reflected differences in reproductive needs. While females, with and without pouch young did not differ in how far they would forage from protective cover, females with pouch young remained closer to their nearest neighbour than those without. This finding perhaps reflected the importance of relying on communal vigilance when their flight time from a predator may be hindered due to increased weight and bulk of a pouch young. These findings supported the theory that group vigilance anti-predator strategies are somewhat innate in the tammars, as having previously been completely isolated from predators their responses could have been lost, and once released there was no opportunity to socially learn the appropriate responses off an established population. Some habitat and social grouping behaviours were occasionally