Understanding the dietary ecology of animals provides information about their habitat requirements, facilitating informed conservation. We used last-bite diet and stable isotope analysis to assess the diet of juvenile and adult green turtles (Chelonia mydas) at two different habitats located 10 km apart within Port Curtis, Queensland, Australia. Last-bite diet analysis indicated that turtles had distinctly different diets in these two habitats: in one the diet was dominated by red macroalgae and in the other the diet was dominated by seagrass. Only juveniles (n = 12) were caught in the habitat where red macroalgae dominated the diet, while both juveniles (n = 9) and adults (n = 38) were captured in the habitat where seagrass dominated the diet. In the seagrass habitat there was no difference in diet between juveniles and adults, and no difference in diet between adult males (n = 17) and females (n = 21).

Because the red macroalgae and seagrass had distinctly different carbon stable isotope ratios, it was possible to detect a change in diet by comparing the carbon stable isotope ratio between serum and epidermal tissue sampled from the same turtle. In this region, a switch in diet would reflect a shift in foraging habitat. Such comparisons indicate that ∼50% of turtles switched diet, and therefore changed foraging habitat between the time when blood serum and epidermis were formed. This implies that switching foraging habitat by green turtles within this region is a common occurrence, which is somewhat surprising because previously it was thought that foraging green turtles had high site fidelity with relatively small home ranges.

Critical evaluations of bait attractiveness for camera trapping wildlife are scant even though use of the most attractive bait should improve detection of cryptic, threatened species. We aimed to determine the most attractive bait for camera trapping the northern hopping-mouse (Notomys aquilo) and sympatric mammals. We also tested the effectiveness of overhead camera trap orientation in identifying individual northern quolls (Dasyurus hallucatus) as this could be used to define a camera trap event for analysis purposes. Using white-flash camera traps, the attractiveness of four baits (peanut butter with oats, corn, sesame oil and sunflower kernels) and a control were compared for N. aquilo, D. hallucatus, the northern brown bandicoot (Isoodon macrourus) and the agile wallaby (Notamacropus agilis). Spot patterns of D. hallucatus were compared to determine the visitation rate of individuals. Peanut butter– and sesame oil–based baits were significantly more attractive to D. hallucatus, while I. macrourus strongly preferred the peanut butter bait. Bait type did not affect the mean number of events for N. aquilo or N. agilis. The consistently identifiable images of individual D. hallucatus were used to determine the optimal event delineator of 15 min. The improved techniques for camera trapping D. hallucatus should be valuable for future capture–recapture studies of this species. Camera trapping is a viable replacement for the ineffective method of indexing the abundance of N. aquilo using indirect signs.

Captive-bred brush-tailed rock-wallabies (BTRW) were reintroduced into the Grampians National Park, Australia, during 2008–12. Two release strategies (methods) were examined: ‘Small release with supplementation’ (Strategy 1) and ‘Larger release, no supplementation’ (Strategy 2).

Of the 39 animals released, 18% survived. Thirty-six percent of all mortality occurred within the first 100 days. Under Strategy 1, 22 animals were released in five groups. Twenty deaths occurred across 48 months, with predation estimated to account for 15% of mortalities. Under Strategy 2, 17 individuals were reintroduced across one month. Twelve deaths occurred in the five months following release, with predation estimated to account for 83.4% of mortalities. Of the independent variables tested for their relationship to survival time after release, release strategy was the only significant predictor of survival time after release with the risk of death 3.2 times greater in Strategy 2.

Independent variables tested for their relationship to predation risk indicated that release strategy was also the only significant predictor of predation risk, with the risk of death associated with predation 10.5 times greater in Strategy 2. Data suggested that fox predation was the main factor affecting BTRW establishment. Predation risk declined by 75% during the first six months after release. A significant positive relationship was also found between predation risk and colony supplementation events.

We conclude that predation risk at Moora Moora Creek is reduced in releases of fewer animals, that it declines across time and that disturbing BTRW colonies through the introduction of new animals can increase predation risk. We recommend that future reintroductions should employ diverse exotic predator control measures at the landscape scale, time releases to periods of lowest predator activity, and limit colony disturbance to maintain group cohesion and social structure. Furthermore, the preferred method of population establishment should be single, small releases over multiple sites without supplementation. Further testing of the reintroduction biology of this species is urgently required.

The genus Cyrtodactylus is the most diverse and widely distributed group of geckos in the world. Throughout their extensive range, species exploit a diverse range of habitats and are able to partition niches locally. Recent work has shown that Cyrtodactylus geckos in Queensland, Australia, have radiated in situ after colonisation by an arboreal Papuan ancestor and have undergone a habitat switch to rock dwelling during their evolutionary history. Using mitochondrial data we conducted a phylogeographic and molecular dating analysis to investigate the historical biogeography of Cyrtodactylus species in north Queensland. Our analyses show that after the arrival of a Papuan rainforest-dwelling ancestor, Cyrtodactylus diverged into two major lineages: one more restricted in northern Cape York and the other more widespread. Discordance in the timing of the speciation events and phylogeographic distribution within the two lineages likely reflect regional differences in the continuity of mesic rock habitats and climatic variability over the last 15 million years. Reconstructing the history of habitat use on a pre-existing global phylogeny reveals that switches between major habitat ecologies, rock and forest, have occurred multiple times in this genus. The ability to transition between different habitat types may have contributed to the global diversification of these geckos.

A detailed description of the methods used to house, maintain and assess the reproductive condition of captive Julia Creek dunnarts, Sminthopsis douglasi, that led to successful breeding of the species in captivity is provided. Basic features of the reproductive biology of this species of Sminthopsis have been established from observations made on captive animals. The females are polyoestrous, with a cycle length of ∼28 days. Young are born 13–16 days after mating and are dependent on the mother for ∼70 days. The age at which captive animals commence breeding ranged from 13 to 38 weeks (females) and 23 to 40 weeks (males). Both sexes are capable of breeding when two years old. Breeding in wild populations is seasonal and occurs over a six-month period from August to March, which encompasses the hottest and wettest time of the year. Both females and males are known to be capable of breeding in more than one season and females have the potential to rear two litters in a season. Recruitment of young to the population may be affected by heavy rainfall during the breeding season that can lead to closure of the cracks and holes in which the dunnarts shelter.

The spotted-tailed quoll (Dasyurus maculatus) is the largest marsupial carnivore on mainland Australia. It usually occurs at relatively low population densities and its cryptic nature makes it exceedingly difficult to observe in its natural habitat. On the mainland the species is also listed as nationally endangered and more information is needed to direct any meaningful conservation effort. In this study we aimed to elucidate quolls’ spatial requirements and activity patterns using GPS collars on 10 males and 4 females. Quolls were predominantly nocturnal but some individuals showed pronounced daytime activity. There was no apparent seasonal shift in the timing of activity. The movement of quolls appeared to be confined to home ranges that were relatively large for predators of their size. Furthermore, males used home ranges about three times as large as that of the smaller females. There appeared to be some spatial segregation between not only females, which have been considered territorial, but also males. Overall, it is likely that the larger areas used by males is partly caused by the sexual dimorphism in body mass that entails differences in prey requirements and spectrum, but probably is also a function of a promiscuous mating system. All of these could explain the observed more unidirectional movement and larger distances travelled per day by males.