Dissertations / Theses on the topic 'Rock-wallaby'

Thesis (Ph. D.)--University of Sydney, 2002.
Title from title screen (viewed Apr. 24, 2008). Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the School of Biological Sciences, Faculty of Science. Degree awarded 2002; thesis submitted 2001. Includes bibliography. Also available in print form.

Overabundant western grey kangaroo Macropus fuliginosus are known to impact agriculture, but how are they impacting threatened fauna sharing their habitat? In Paruna Wildlife Sanctuary, southwest Western Australia, kangaroos are suspected of competing with the sympatric and endangered black-flanked rock-wallaby Petrogale lateralis lateralis, however there is no research to support this. If kangaroos are negatively impacting rock-wallabies, kangaroo densities may need to be managed to ameliorate competitive pressures on rock-wallabies. We investigated the potential for dietary competition between M. fuliginosus and P. l. lateralis by measuring the overlap in their diets and foraging patches, as well as food resource availability. A combination of scat analysis, motion sensor camera trapping and vegetation surveys were employed. Petrogale lateralis lateralis diets were dominated by forbs and overlapped with those of M. fuliginosus which featured mostly browse and forbs (Schoener index: 0.56). Some of their shared preferred food resources were spatially and/or temporally limited. Their foraging patches also overlapped (33.9%), however these macropod species predominantly used different areas of the outcrop. Evidence over the duration of the study indicates potential for low levels of dietary competition, however the availability of shared food resources and resource partitioning suggest that P. l. lateralis were not being adversely impacted. In terms of the threatening processes limiting P. l. lateralis recovery, predation has been ranked higher than competition, a finding that is likely supported by the present study. This will likely remain true even if M. fuliginosus densities increase in the future. Conservation actions should therefore continue to prioritise the mitigation of predation threats to P. l. lateralis populations.

Based on the recommendations of both the 1993 Reintroduction biology of Australasian Fauna Conference and the 1994 Rock Wallaby Symposium, captive-bred Yellow footed rock wallabies were reintroduced into areas of their former ranges in both South Australia and Queensland

The black-flanked rock-wallaby (Petrogale lateralis lateralis) has suffered a significant decline in its distribution in Western Australia. This has been attributed to introduced predators (predominantly the red fox) and herbivores, fire, and habitat destruction due to clearing. Although since 2001 the Department of Environment and Conservation (DEC) had begun to reintroduce this species back into its former range, little was known of the behavioural ecology of this species. Fox control in the 1980s and 1990s resulted in population increases of rock-wallabies on several reserves in the central wheatbelt of WA. However, recently these populations have rapidly declined despite continuing fox control. All too often, management and conservation programs are based on little understanding on the natural history and ecology of animals, which can ultimately result in poor management. One of the major problems with our understanding of the ecology of P. l. lateralis was that there was little information on their behavioural, foraging and feeding patterns, and how these affect the population dynamics of the species. Although this species appears to be substrate bound, requiring complex rock structures that are believed to protect them from both predators and adverse climatic conditions, there was still little understanding of how these animals utilise this important resource, particularly in highly variable environments. There is a long-standing premise in behavioural ecology that highly variable environments can significantly affect the behaviour and demography of animal populations. Although this has been well documented in birds, and primates, behavioural elements are rarely incorporated into marsupial studies. For the success of any current and future reintroductions of rock- wallaby populations into new areas, we needed to understand the relationships between the landscape and climatic elements and the behavioural patterns and population dynamics of the species’. This study describes the behavioural ecology of the rock-wallaby subspecies P. l. lateralis in the central wheatbelt region of Western Australia. It focuses on environmental variables that have rarely been studied simultaneously, those of climate, fear of predation and conspecifics, and habitat structure. The Nangeen Hill Reserve population was chosen because it is one of seven small fragmented reserves, within a landscape, with strong seasonal climatic patterns. A methodological approach was developed that enabled identification of the key ecological determinants of the foraging and feeding behaviours of P. l. lateralis in this highly variable environment. The study included a comprehensive examination of all the behavioural characteristics and abilities of P. l. lateralis, compiling a detailed behavioural repertoire (ethogram), a first for this species, and included both nocturnal and diurnal observations. P. l. lateralis uses a complex arrangement of non-agonistic and agonistic behavioural acts that determines its social organisation, and indicates a linear dominance hierarchy. There is little vocal communication, and instead the rock-wallabies appear to rely on both ritualised behaviour acts and chemical cues to exchange information about the physiological or behavioural state of the individual, thus evolving very diverse and complex social behaviours. Nangeen Hill rock-wallabies occupy a permanent central rock refuge, with strong signs of site fidelity. Their foraging patterns reflect those of a central place forager, but distances travelled are restricted in open habitats, with stronger preferences for areas of rock structural complexity. Their foraging behaviour is not strictly nocturnal nor can it be considered crepuscular, exhibiting significantly different seasonal patterns. Time allocation for foraging is strongly affected by fear of predators and to a lesser extent conspecifics, and certain climatic conditions. Although rock-wallabies use multiple behavioural strategies to reduce predation risk, energy costs, and intraspecific agonistic interactions, they can only can survive in their present environment if they have access to complex rock refuge. This rock resource not only enables them to avoid extreme ambient temperatures, but also gives them protection against predators particularly when environmental conditions are unfavourable. In addition, their cave refuge also gives them protection from precipitation, reducing heat loss, and provides a safe environment for both rearing young and for courtship without the cost of predation. Rock refuge is the most important resource that defines this species and is central to all aspects of its life history. Although this reserve has been under a fox-baiting program since 1982, and subsequently resulted in a rock-wallaby population increase and habitat expansion, within the rocky outcrop (Kinnear et al., 1998), it was clear from my research that the ecological situation was more complex. There is a strong predation influence on the behaviour indicating that the population is now predominantly fear-driven. It is the fear of predators and dominant conspecifics that restricts both the foraging range and time spent feeding. This results in animals being tightly restricted to their rock refuge, resulting in overgrazing and habitat degradation. These concerns led me to believe that if the current situation was not alleviated, then a population crash was imminent.A population crash subsequently happened in 2010, and the cause of this decline is a result of a complex ecological relationship, that includes direct and indirect predatory effects, weed invasion, and drought. The results of this research and the subsequent recent population decline, show the importance of including behaviour into an ecological study to have a better understanding. It provides a better understanding of a species as well as providing important insights into its evolutionary past, and how this has shaped their social and demographic patterns. This research also demonstrates how the use of a permanent central rock refuge both contributes to the animal’s continued survival, and restricts its future distribution, particularly in ecologically altered landscapes. Although this study was not designed primarily to resolve conservation and management problems, its findings are already being used to design an active management plan for the central wheatbelt rock-wallaby populations. The complex relationships between how an animal obtains and uses its resources, the availability of different vegetation patches, and population demographics creates significant problems for the management of a species. This emphasises the need for future research on all animal groups, to understand the relationships between the habitat and landscape elements, and that of behaviour and population dynamics. Understanding how animals perceive their environment and how they adjust to its changes will be paramount for the future management and survival of many species.

Bibliography: leaves 107-117. This thesis examines the rationale for managing biological diversity on a regional basis and develops recommendations for the use of two computational methods in biodiversity management planning by conducting a case study in the Flinders Ranges, centred on the Yellow-footed Rock Wallaby (abstract)

Petrogale lateralis lateralis (Black-flanked Rock-wallaby) has declined in its mainland distribution to a few isolated populations with extant populations known from six localities in the Wheatbelt, the Cape Range, the Calvert Range, and Barrow and Salisbury Islands (Pearson and Kinnear 1997). The conservation status of P. l. lateralis is classified as endangered. It is therefore extremely important that the threatening processes associated with this species are investigated to ensure appropriate management of the remaining populations. P. l. lateralis is a herbivore with a foraging range restricted to rocky outcrops; competition for resources from introduced herbivores has been implicated as a limiting factor to rock-wallaby populations. The aim of this study was to examine possible dietary overlap between P. l. lateralis and two co-occurring species Macropus robustus (euro) and Capra hircus (feral goat) at Cape Range National Park in order to determine if there is competition for food resources. Vegetation surveys combined with dietary analysis of plant epidermal fragments found in the faecal material of the three species and direct behavioural and feeding observations were carried out at two study sites Mandu Mandu Gorge and Pilgonaman Gorge in Cape Range National Park, Exmouth, Western Australia. A seasonal comparison of the diets of P. l. lateralis, M robustus and C. hircus at the study sites was conducted to highlight the time of year that competition for food resources is most prevalent. Vegetation profiles and the vegetation surveys conducted within the Mandu Mandu Gorge and Pilgonaman Gorge show that plant species vary in abundance and occur in different areas within the gorges. Vegetation surveys showed that the vegetation structure between Mandu Mandu Gorge and Pilgonaman Gorge differed but contained an overlap of plant species. Mandu Mandu Gorge consisted of mainly Triodia grasslands with Acacia sp. The vegetation in Pilgonaman Gorge becomes increasingly dense the farther east into the gorge with the dominant species Ipomoea costata and Ficus brachypoda with a variety of herb and grass species. Hence, the plant species that herbivores forage on in Mandu Mandu Gorge and Pilgonaman Gorge are limited spatially due to changes in vegetation structure within the gorges, with dense patches occurring on the gorge floor and scattered plants occurring around P. l. lateralis refuge sites and on the rocks of the gorge walls. The faecal analysis found that approximately 60% of P. l. lateralis diet consisted of dicotyledon species in summer with consistently higher proportions, (approximately 70% dicotyledons) in winter. The species occurring in the highest proportions in the P. l. lateralis diet at Pilgonaman Gorge in summer were Ficus platypoda (12.5 %), Ptilotus obovatus (11.5%) and Ipomoea costata (7.5%) F. platypoda comprised of (17.6%), I. costata (10.6%), P. obovatus (11.5%) and Plumbago zeylanica comprised (10.6%) of the diet in winter. The percentages of identified dicotyledon species in the summer diet at Mandu Mandu Gorge are as follows; F. brachypoda comprising of (15.7%), I. costata (11.8%), P. obovatus (11.8%), and Solanum sp. (2%) The main plant present in the P. l. lateralis diet in winter at Mandu Mandu Gorge was F. brachypoda (14.5%), followed by I. costata (13.3%), P. obovatus (7.2%), then Solanum sp. (3.6%), P. zeylanica (2.4%) and Achyranthes aspera (2.4%) The proportion of monocotyledon and dicotyledon species for M robustus remained constant between both summer and winter in both gorges. There was little variation in the proportion of monocot species consumed between summer and winter in Mandu Mandu Gorge and Pilgonaman Gorge. Approximately 25 % of the diet consisted of dicotyledon species. M robustus consumed Myrtaceae sp., Sida sp. Ptilotus obovatus and Ipomoea costata in Pilgonaman Gorge in summer and Myrtaceae sp., Sida sp., P. obovatus and Ficus brachypoda in winter. In Mandu Mandu Gorge, M robustus consumed Myrtaceae sp., Sida sp. and P. obovatus in summer and Sida sp and P. obovatus in winter. A high proportion of the C. hircus diet in this study consisted of dicot species. There was little variation in the percentage of dicot species consumed in summer and winter for C. hircus in both Mandu Mandu Gorge (63% in summer and 65% in winter) and Pilgonaman Gorge (67% in summer and 69% in winter). The identified dicot species in the C. hircus diet in the winter months at Pilgonaman Gorge comprised of I. costata (12.4%), P. zeylanica (9%), P. obovatus (9%), Sida sp. (5.6%), F. brachypoda 4.5%, Solanum sp. 3.3%, A. aspera (2.3%) and Myrtaceae sp. (1 %). The dicot species consumed by C. hircus at Mandu Mandu Gorge include Sida sp. (3.5 %), P. obovatus (17.5%), I. costata (8.8%), F. brachypoda (7%) and Solanum sp. (5.3%) of the summer diet. The winter diet for C. hircus at Mandu Mandu Gorge comprised of P. obovatus (15.1%), I. costata (8.1%) F. brachypoda (10.5%) and Solanum sp. (4.7%) Direct observations of P. l. lateralis eating support the identification of plant species found within their faeces during the scat analysis of this study. Furthermore, physical evidence of grazing on leaves within Pilgonaman Gorge suggests that herbivores are actively grazing upon F. brachypoda, Malvaceae sp., Solanum sp., I. costata, Sida sp., P. zeylanica and Cenchrus ciliaris. P. l. lateralis, M robustus and C. hircus were found to consume a selection of these species in various proportions in the scat analysis. A significant dietary overlap of plant species was found to occur between C. hircus, M robustus and P. l. lateralis in Pilgonaman Gorge and Mandu Mandu Gorge. P. l. lateralis, M robustus and C. hircus had a dietary overlap consuming the same plant species; Myrtaceae sp., Sida sp. C. ciliaris, I. costata and F. brachypoda and P. obovatus at differing percentages. C. hircus was found to forage on the following species in common with P. l. lateralis; Myrtaceae sp. Sida sp. C. ciliaris, I. costata and F. brachypoda, P. obovatus, Solanum sp., P. zeylanica and A. aspera. M robustus consumed Myrtaceae sp., Sida sp. C. ciliaris, I. costata and F. brachypoda, P. obovatus and E. caevuless in common with P. l. lateralis. The dietary overlap between C. hircus, M robustus and P. l. lateralis in Pilgonaman Gorge in summer was found to be significant with a probability of N= 22 W= 43.46 and P < 0.01. The dietary overlap between the species was consistent in winter N=22 and W= 41.02 and had the same probability value of P < 0.01. A significant dietary overlap occurred in summer with a probability of N=ll, W= 20.8 and P < 0.05 and winter N= 12, W= 24.25 and P < 0.02 in Mandu Mandu Gorge. Finally, behavioural observations on interactions between feral goats and the rockwallabies indicate that when feral goats are at close range they are disruptive to the normal activities of rock-wallabies. This suggests that the goats are not only competing for food resources but that they may also elicit a form of interference competition. Implications of competition are discussed and recommendations for further research and management are considered.

In this thesis a number of statistical methods have been developed and applied to habitat suitability modelling for rare and threatened species. Data available on these species are typically limited. Therefore, developing these models from these data can be problematic and may produce prediction biases. To address these problems there are three aims of this thesis. The _rst aim is to develop and implement frequentist and Bayesian statistical modelling approaches for these types of data. The second aim is develop and implement expert elicitation methods. The third aim is to apply these novel approaches to Australian rare and threatened species case studies with the intention of habitat suitability modelling. The _rst aim is ful_lled by investigating two innovative approaches for habitat suitability modelling and sensitivity analysis of the second approach to priors. The _rst approach is a new multilevel framework developed to model the species distribution at multiple scales and identify excess zeros (absences outside the species range). Applying a statistical modelling approach to the identi_cation of excess zeros has not previously been conducted. The second approach is an extension and application of Bayesian classi_cation trees to modelling the habitat suitability of a threatened species. This is the _rst `real' application of this approach in ecology. Lastly, sensitivity analysis of the priors in Bayesian classi_cation trees are examined for a real case study. Previously, sensitivity analysis of this approach to priors has not been examined. To address the second aim, expert elicitation methods are developed, extended and compared in this thesis. In particular, one elicitation approach is extended from previous research, there is a comparison of three elicitation methods, and one new elicitation approach is proposed. These approaches are illustrated for habitat suitability modelling of a rare species and the opinions of one or two experts are elicited. The _rst approach utilises a simple questionnaire, in which expert opinion is elicited on whether increasing values of a covariate either increases, decreases or does not substantively impact on a response. This approach is extended to express this information as a mixture of three normally distributed prior distributions, which are then combined with available presence/absence data in a logistic regression. This is one of the _rst elicitation approaches within the habitat suitability modelling literature that is appropriate for experts with limited statistical knowledge and can be used to elicit information from single or multiple experts. Three relatively new approaches to eliciting expert knowledge in a form suitable for Bayesian logistic regression are compared, one of which is the questionnaire approach. Included in this comparison of three elicitation methods are a summary of the advantages and disadvantages of these three methods, the results from elicitations and comparison of the prior and posterior distributions. An expert elicitation approach is developed for classi_cation trees, in which the size and structure of the tree is elicited. There have been numerous elicitation approaches proposed for logistic regression, however no approaches have been suggested for classi_cation trees. The last aim of this thesis is addressed in all chapters, since the statistical approaches proposed and extended in this thesis have been applied to real case studies. Two case studies have been examined in this thesis. The _rst is the rare native Australian thistle (Stemmacantha australis), in which the dataset contains a large number of absences distributed over the majority of Queensland, and a small number of presence sites that are only within South-East Queensland. This case study motivated the multilevel modelling framework. The second case study is the threatened Australian brush-tailed rock-wallaby (Petrogale penicillata). The application and sensitivity analysis of Bayesian classi_cation trees, and all expert elicitation approaches investigated in this thesis are applied to this case study. This work has several implications for conservation and management of rare and threatened species. Novel statistical approaches addressing the _rst aim provide extensions to currently existing methods, or propose a new approach, for identi _cation of current and potential habitat. We demonstrate that better model predictions can be achieved using each method, compared to standard techniques. Elicitation approaches addressing the second aim ensure expert knowledge in various forms can be harnessed for habitat modelling, a particular bene_t for rare and threatened species which typically have limited data. Throughout, innovations in statistical methodology are both motivated and illustrated via habitat modelling for two rare and threatened species: the native thistle Stemmacantha australis and the brush-tailed rock wallaby Petrogale penicillata.

In this thesis a number of statistical methods have been developed and applied to habitat suitability modelling for rare and threatened species. Data available on these species are typically limited. Therefore, developing these models from these data can be problematic and may produce prediction biases. To address these problems there are three aims of this thesis. The _rst aim is to develop and implement frequentist and Bayesian statistical modelling approaches for these types of data. The second aim is develop and implement expert elicitation methods. The third aim is to apply these novel approaches to Australian rare and threatened species case studies with the intention of habitat suitability modelling. The _rst aim is ful_lled by investigating two innovative approaches for habitat suitability modelling and sensitivity analysis of the second approach to priors. The _rst approach is a new multilevel framework developed to model the species distribution at multiple scales and identify excess zeros (absences outside the species range). Applying a statistical modelling approach to the identi_cation of excess zeros has not previously been conducted. The second approach is an extension and application of Bayesian classi_cation trees to modelling the habitat suitability of a threatened species. This is the _rst `real' application of this approach in ecology. Lastly, sensitivity analysis of the priors in Bayesian classi_cation trees are examined for a real case study. Previously, sensitivity analysis of this approach to priors has not been examined. To address the second aim, expert elicitation methods are developed, extended and compared in this thesis. In particular, one elicitation approach is extended from previous research, there is a comparison of three elicitation methods, and one new elicitation approach is proposed. These approaches are illustrated for habitat suitability modelling of a rare species and the opinions of one or two experts are elicited. The _rst approach utilises a simple questionnaire, in which expert opinion is elicited on whether increasing values of a covariate either increases, decreases or does not substantively impact on a response. This approach is extended to express this information as a mixture of three normally distributed prior distributions, which are then combined with available presence/absence data in a logistic regression. This is one of the _rst elicitation approaches within the habitat suitability modelling literature that is appropriate for experts with limited statistical knowledge and can be used to elicit information from single or multiple experts. Three relatively new approaches to eliciting expert knowledge in a form suitable for Bayesian logistic regression are compared, one of which is the questionnaire approach. Included in this comparison of three elicitation methods are a summary of the advantages and disadvantages of these three methods, the results from elicitations and comparison of the prior and posterior distributions. An expert elicitation approach is developed for classi_cation trees, in which the size and structure of the tree is elicited. There have been numerous elicitation approaches proposed for logistic regression, however no approaches have been suggested for classi_cation trees. The last aim of this thesis is addressed in all chapters, since the statistical approaches proposed and extended in this thesis have been applied to real case studies. Two case studies have been examined in this thesis. The _rst is the rare native Australian thistle (Stemmacantha australis), in which the dataset contains a large number of absences distributed over the majority of Queensland, and a small number of presence sites that are only within South-East Queensland. This case study motivated the multilevel modelling framework. The second case study is the threatened Australian brush-tailed rock-wallaby (Petrogale penicillata). The application and sensitivity analysis of Bayesian classi_cation trees, and all expert elicitation approaches investigated in this thesis are applied to this case study. This work has several implications for conservation and management of rare and threatened species. Novel statistical approaches addressing the _rst aim provide extensions to currently existing methods, or propose a new approach, for identi _cation of current and potential habitat. We demonstrate that better model predictions can be achieved using each method, compared to standard techniques. Elicitation approaches addressing the second aim ensure expert knowledge in various forms can be harnessed for habitat modelling, a particular bene_t for rare and threatened species which typically have limited data. Throughout, innovations in statistical methodology are both motivated and illustrated via habitat modelling for two rare and threatened species: the native thistle Stemmacantha australis and the brush-tailed rock wallaby Petrogale penicillata.

The allied rock-wallaby, Petrogale assimilis, is a medium-sized, monomorphic, continuously-breeding macropodid marsupial that lives in rocky habitats in the climatically unpredictable wet-dry tropics of north-eastern Australia. Long-term capture-recapture records of individually marked P. assimilis from an isolated population inhabiting Black Rock, a sandstone escarpment in western Queensland, were used to investigate the population demography of the species over 12 years. In natural populations, both extrinsic environmental variation and intrinsic density-dependence contribute to variability in demographic fitness components. Changes in these fitness components, and trade-offs between them, determine the dynamics of populations. Almost no information is available on the temporal variability of the demographic fitness components and their relationship with rates of population change for kangaroos and wallabies. This research provides the first long-term study of a macropod species that is based on individually-marked animals, which are required to estimate such fitness components and their temporal variation. A major aim of this research was to determine the impact of variability in the fitness components on the population dynamics of this macropod species, and to compare these results with those available for other long-lived herbivores. The demographic components of the life history of P. assimilis at Black Rock were investigated, along with the interrelationships between long-term temporal variability in these components and intrinsic and extrinsic factors, and individual fitness. The long-term fitness of the population and the influence of temporal covariation between the demographic components on variability in population fitness were also examined. The growth relationship between the size and age of P. assimilis at Black Rock was determined using various nonlinear models within a mixed-effects framework. The age and individual variation in growth of animals of unknown age were estimated from measurements of body size. Parametric nonlinear models did not fit the growth pattern, but semi-parametric spline models adequately described the growth pattern and accounted for individual variation. Differences in growth between the sexes were small, suggesting no strong sexual size dimorphism in P. assimilis. Growth rates were highest during pouch development, were lower during the weaning period, and decreased dramatically after weaning. The age structure of the population of P. assimilis at Black Rock varied substantially over the study period. The log-transformed relationships between body mass and various body size measures were nonlinear and the head length was the best predictor of body mass of P. assimilis at Black Rock. Indices of body condition were calculated from the residuals of mixed-effects models that estimated the form of the size-mass relationship using splines. Substantial variation in body condition was explained by annual and seasonal variability and lagged rainfall, as well as variation between individual animals. Variation in body condition was not associated with sex or age-class, and did not depend on the lactation status of females. The estimated index of body condition appeared to represent the nutritional status of individual P. assimilis at Black Rock and was used as a predictor of variation in demographic rates associated with individual fitness. Directional goodness-of-fit tests for Cormack-Jolly-Seber capture-recapture models showed that individual P. assimilis at Black Rock had a trap happy response to capture. The likelihood of recapture varied over time and depended on whether animals were captured on the previous sampling occasion. Recapture probability was generally very high, and was lower for adult females not marked as young than for other females marked as young and for all males, regardless of age at marking. Similar effects on recapture probability were identified using generalised linear mixed models, though the time-dependent effect could be simplified to a year by season interaction plus the effect of trapping effort. There was strong agreement between estimates of population size based on different methods. Population size fluctuated substantially over the study with periods of consistent increase and decline, and showed evidence of rapid population recovery from relatively low numbers under positive environmental conditions. Support for the Trivers-Willard hypothesis (TWH), which states that if the costs of reproduction differ between the sexes then the offspring sex ratio will vary depending on the parent’s ability to allocate resources, was evaluated. The sex ratio at birth was equal and not correlated with mother’s age, body mass or body condition. Sex ratio varied seasonally; being female-biased in the mid- to late-dry season and male-biased in the late dry and early wet seasons. Survival from birth to pouch emergence was correlated with environmental conditions, depending on sex, in the direction consistent with TWH; male survival was higher than female survival under good conditions and lower under poor conditions. Also, survival was higher for pouch young of heavier mothers or mothers in good condition. Survival to pouch emergence was also density-dependent. In support of TWH, the sex ratio at pouch emergence was male-biased under good environmental conditions and was male-biased for mothers in good condition or heavier mothers. Mothers in good condition produced offspring in good condition, satisfying an assumption of TWH, and higher mass at pouch emergence resulted in improved survivorship to weaning. Survival from pouch emergence to weaning was higher for females than males, and was positively correlated with environmental conditions. Females had higher survival to weaning than males in the hotter seasons, but there was no sex difference in survival in the cooler seasons. Survival was also higher for females than for males born to lighter mothers, but there was no difference between the sexes born to heavier mothers. Male-biased sex ratios at weaning were observed under good environmental conditions and for mothers with higher body mass, providing support for TWH at this life history stage. The body mass of offspring at weaning was correlated with maternal body mass, satisfying an assumption of TWH, and offspring mass was also higher under high rainfall conditions. Patterns of variability in the sex ratio and pre-weaning survival of P. assimilis at Black Rock were influenced by a combination of: (1) adaptive allocation of resources between the sexes depending on maternal condition; and (2) non-adaptive extrinsic modification associated with environmental stochasticity, allowing mothers to respond quickly in a variable environment to maximise their current reproduction and future survival. Male P. assimilis matured later and were heavier at maturity than females. Early maturity was favoured for females but not for males under high rainfall conditions, independent of size at maturity. Age at maturity was density-dependent for males only, with delayed maturity at lower population sizes. Maturation was delayed for both sexes when body mass at weaning was low relative to individuals with high weaning mass. Subadult male survival was lower than for females, and survival of both sexes increased under high rainfall conditions. Adult survival was higher than subadult survival, independent of sex, and survival of older adults was lower than that of prime-aged adults, indicating senescence. Survival increased under positive climatic conditions and was density-dependent, with lower survival at high population sizes. Males had lower survival to sexual maturity than females, and increased body mass improved survival for both sexes. Population growth rate was most sensitive to changes in prime-aged adult survival, as is typical for longer-lived vertebrates. The temporal variance in the juvenile survival stages was much higher than in adult survival. Therefore, consistent with other longer-lived, iteroparous vertebrates, the impact of relative changes to adult survival on population growth were much greater than changes of the same magnitude to either survival in the juvenile stages, or to the fecundity rate. There was a negative relationship between temporal variation in the demographic rates and their sensitivities, indicating that the more variable vital rates had low effects on long-term population growth. The most variable demographic rates covaried suggesting that common factors contributed to temporal variability. Covariation in survival between the juvenile and pre-maturity life history stages accounted for most of the variability in the population growth rate, indicating survivorship in the juvenile stages was more important than adult survival in determining changes in the population growth rate of P. assimilis at Black Rock. The survival patterns of different life history stages were variable for P. assimilis at Black Rock, and responded differently to environmental variability and population density. Density-dependent and density-independent limiting factors primarily acted on the juvenile survival components of the life history, and covariation among these fitness components influenced the dynamics of the population. Survival of prime-aged adult P. assimilis at Black Rock was much less sensitive to these limiting factors, showed the highest elasticity, and appeared to be buffered against temporal variability. These results add to growing evidence from natural populations that natural selection may favour traits that are highly buffered against environmental variability.

The rock wallaby, Petrogale assimilis, is a small macropodid living in isolated colonies in rocky outcrops in the wet-dry tropics of northern Australia, a region where the onset, magnitude and duration of the wet season are unpredictable. Eighty individual P. assimilis were caught in traps between one and 42 times during more 4000 trap days over 44 monthly field trips to Black Rock (19° 05'S., 144° 07'E.) between June 1986 and June 1990, as part of a longitudinal study of this species. Petrogale assimilis was a typical macropodid for its size, with regard to its life history and reproductive ecology. Both sexes were capable of reproducing continuously; gestation was about the same length as the oestrous cycle (approximately one month); a single young was born and, a post-partum oestrus and embryonic diapause probably occurred. Pouch young remained permanently attached to the teat until 110 - 143 days (n=11). Permanent exit from the pouch occurred at 180 - 231 days (mean=201 days, n=25), and weaning occurred between 267 - 387 days (n=5). Sexual maturity occurred at a minimum age of 17.5 months in females and 23 months in males. The growth of pouch young of P. assimilis at Black Rock was compared with that of captive animals maintained in ad libitum conditions at Macquarie University, Sydney. No significant differences in pouch young growth were found between these two groups of young despite the potentially different nutritional conditions of their mothers. Individual pouch young of P. assimilis varied in their rate of growth, although the data were too limited to examine the potential causes. Exponential and Gompertz growth curves were described for the head and pes length respectively of pouch young and were used to predict the age of pouch young of P. assimilis. Body measurements were of limited use for predicting the age of older rock wallabies and cannot be used accurately for sexually mature animals. Petrogale assimilis exhibited only limited sexual dimorphism in its external morphology and growth during adult life with males slightly larger than females in the average length of the head (4%), pes (4%), tail (7%) and ear (3%). Sexual dimorphism was most pronounced in upper limbs (and presumably in their associated musculature) (11%) and in weight (15%). Radiographs were taken of the teeth of rock wallabies older than pouch young to measure the eruption stage and position of the teeth. Regression curves which were fitted to data from known age animals were used to estimate the age of rock wallabies older than pouch young. Available data, albeit limited, suggested that the rates of molar eruption and movement are similar in both wild and captive P. assimilis. Molar progression was mostly limited to movement induced by wear. This movement occurred too slowly and with too much variability to be useful in ageing animals in which the head had ceased growing or which had full molar eruption. The use of molar eruption for determining the age of macropodids was limited to animals with less than full molar eruption (estimated mean age = 7 years). As P. assimilis appears to have a maximum life span of approximately 12 years, molar eruption was a useful tool for ageing animals during a major proportion of their adult life. The influence of the seasons on changes in the body condition of adult rock wallabies probably resulted from the availability of water and quality of forage. Although below average annual rain fell in both years of my study (1987 and 1988) which could be fully examined, the pattern of rainfall was quite different during each of these years and this pattern of rainfall influenced the body condition of adult rock wallabies. In 1987 rock wallabies lost weight (mean loss = 10%) and the packed cell volume of their blood decreased as the dry season progressed. In 1988, when a higher proportion of the year's rain fell during the dry seasons and green forage was available throughout the year, rock wallabies maintained weight and the packed cell volume of their blood increased. The seasons were also crucial to the survivorship of pouch young. Pouch young born in the late dry season were twice as likely to survive total lactational dependence than those born during the wet or early dry seasons. However, the length and harshness of the late dry season is dependent on when the unpredictable wet season arrives. This may explain some of the loss of 46% of pouch young during total lactational dependence. If the wet season does not arrive, the mother may lose her pouch young early in pouch life when her investment is low, in the expectation that the next pouch young is more likely to survive to take advantage of the wet season flush. The cause of the mortality of young pouch young is unknown, although anecdotal evidence suggests that components of the mother's diet may be crucial. There was considerable variation between mothers in the survivorship of pouch young during both total lactational dependence and late pouch life. However, I was able to identify only one maternal correlate of pouch young survivorship: that smaller-framed females were more successful in rearing young in late pouch life than their larger conspecifics. I was unable to examine the influence of other factors (such as maternal age and position in the dominance hierarchy) which have been shown to influence the survivorship of young in other mammals. There was a negative correlation between the mother's success at rearing a pouch young through total lactational dependence and her success in rearing pouch young through late pouch life, although data are limited. Despite the individual variation in reproductive success observed, the ultimate survivorship of pouch young born during my study was very low (15%). Almost no young born during this period survived to sexual maturity. This result and the recent improvement in offspring survivorship, particularly young at foot, with better seasons suggests that infant survivorship occurs in pulses with a higher proportion of young from most females surviving during more favourable years. The age structure of the population provides supporting evidence for the hypothesis that infant survivorship occurs in pulses. Previous research at Black Rock during the above average rainfall years of the early 1970's suggested that there were a high proportion of pre-reproductive animals present in the population. During my study, the age structure of the population was skewed towards very young animals (pouch young) and very old animals (full molar eruption). Almost no sub-adults or young adults were present at Black Rock. While the older animals could not be accurately aged, I speculate that these animals may be part of a cohort which was born during the above average rainfall years in the late 1970's and early 1980's.

Conservation biology is primarily concerned with the amelioration of species decline. The Yellow-footed Rock-wallaby (Petrogale xanthopus xanthopus) is a medium-sized Macropod that inhabits the semiarid rangelands of South Australia and New South Wales. Its conservation status is Vulnerable C2a(i). In this study, population modelling, spatially explicit habitat modelling and Population Viability Analysis (PVA) have been used to better understand the factors that affect the abundance and distribution of the P. x. xanthopus in South Australia. The processes that drive the population dynamics of a species operate at different scales. As such this research involves a collection of several inter-related and scale-specific empirical studies that provide insights about the population dynamics of P. x. xanthopus. Each of these studies captures environmental, demographic and behavioural process acting on the population at different scales. These include the analysis of relative abundance data derived from an aerial census, mark recapture sampling of demographic parameters in relation to rainfall patterns and a collection of habitat models derived at different scales using presence-absence data. Spatially explicit PVAs enable the population dynamics of a species to be modelled in space and time. Using these data, a PVA is conducted to explore and rank the importance of the factors that threaten this species and help guide their future monitoring and management. Movement is also a key issue when considering problems such as isolation and inbreeding. Given that little is known about the dispersal behaviour of this species, a range of different dispersal behaviours are also simulated in the PVA using random and non-random mating algorithms, to estimate the potential for inbreeding.
thesis (PhD)--University of South Australia, 2004.

Australia has one of the world's worst rates of mammal extinctions, accounting for about one third of the mammal species that have become extinct in the world since 1600 (Burbidge and Manly 2002). Most of these documented declines and extinctions have been concentrated towards mammals in the arid-zone of central Australia. In South Australia (SA), one of the arid-zone species that has undergone substantial declines in range and abundance is the black-footed rock-wallaby, Petrogale lateralis (MacDonnell Ranges race). Animals are now restricted to the Anangu Pitjantjatjara Yankunytjatjara (APY) Lands in the state's far north-west, where there remain two known metapopulations. The decline of this race, which is known as 'warru' by Western Desert Indigenous people, is presumed to have been precipitated by introduced predators. However, in order to exclude other potential causes and establish if recovery is feasible, the current project investigated the ecology of warru in the SA APY Lands. In order to make inferences about the habitat and environmental conditions that best facilitate warru persistence, a modelling approach at multiple spatial scales was used. At a landscape scale, warru presence was correlated to geology, slope, soil type and elevation. At a finer scale, fieldwork established that a number of parameters, including aspect, vegetation, rock complexity and refuges' physical characteristics were important determinants of warru presence and use. The work presents the first analysis of P. lateralis habitat requirements. To establish animals' use of this preferred habitat, the thesis provides preliminary data on warru home range and movement patterns. Unfortunately, catastrophic failure of Global Positioning System collars meant that data were limited to those from one adult female in one month. Nevertheless, during this time, the animal had a larger than expected home range and undertook a number of long-range movements, including a sojourn off the hill. These results have implications for management, including predator and fire management strategies. An investigation of the genetic structure of animals in the three largest-known warru colonies indicated that all colonies are genetically diverse, with levels of heterozygosity approximating expected values. All three also exhibited substantial population structuring, with genetic groups correlating to geographic colonies. However, there was also evidence for sub-structuring being present within one of the colonies and for migration occurring into two colonies, with approximately 5% of sampled individuals having mixed ancestry. An analysis of the parentage of offspring indicated that warru have a polygamous mating structure, although one instance of intra- and inter-year monogamy was also established. Population modelling was used to provide an estimate of the sizes of each of the three studied colonies. Combined with recent aerial survey data, these data suggested that there are likely to be fewer than 200 warru remaining in SA. However, all three colonies showed signs conducive to potential recovery, including high average reproductive rates, even sex ratios and high adult survivorship (>75%). Juvenile survival, however, was significantly lower (51%) and positively correlated with winter rainfall, indicating that access to water is important during the drier winter months. The potential for warru recovery was supported by an analysis of warru blood chemistry in both the wild and captive colonies. The latter was established as part of the recovery program and aimed to produce animals for in-situ supplementation and reintroduction. In 2009, when blood samples were taken, warru colonies did not manifest with results that are indicative of population-wide disease. However, the biochemical parameters of animals in one of the in-situ colonies, 'New Well', suggested some level of nutritional and water stress. This suggests that managers could consider providing warru at New Well with supplementary food and water during drought, and/or using fire to promote vegetational diversity. The study presents the first detailed haematological reference values for P. lateralis and potentially, a methodology for other threatened species recovery programs to follow in order to establish the health of their populations. The current results have implications for management of both the in- and ex-situ warru populations. For example, results pertaining to habitat selection can be used to guide selection of appropriate reintroduction sites. Overall, results indicate that although the extant colonies have small population sizes, the animals therein have high reproductive rates, are not suffering the effects of disease and are genetically diverse. The rock-wallabies' polygamous mating strategy is also likely to continue to facilitate this genetic diversity. Furthermore, although warru have specific habitat preferences, they have the capacity to disperse to surrounding available habitat. Overall, this suggests that, given appropriate management, warru recovery is feasible. Some of these data may also be pertinent for management of P. lateralis in the Northern Territory and Western Australia.
Thesis (Ph.D.) -- University of Adelaide, School of Earth and Environmental Sciences, 2011