Metrics of progress in the understanding and management of threats to Australian birds

Although evidence-based approaches have become commonplace for determining the success of conservation measures for the management of threatened taxa, there are no standard metrics for assessing progress in research or management. We developed 5 metrics to meet this need for threatened taxa and to quantify the need for further action and effective alleviation of threats. These metrics (research need, research achievement, management need, management achievement, and percent threat reduction) can be aggregated to examine trends for an individual taxon or for threats across multiple taxa. We tested the utility of these metrics by applying them to Australian threatened birds, which appears to be the first time that progress in research and management of threats has been assessed for all threatened taxa in a faunal group at a continental scale. Some research has been conducted on nearly three-quarters of known threats to taxa, and there is a clear understanding of how to alleviate nearly half of the threats with the highest impact. Some management has been attempted on nearly half the threats. Management outcomes ranged from successful trials to complete mitigation of the threat, including for one-third of high-impact threats. Progress in both research and management tended to be greater for taxa that were monitored or occurred on oceanic islands. Predation by cats had the highest potential threat score. However, there has been some success reducing the impact of cat predation, so climate change (particularly drought), now poses the greatest threat to Australian threatened birds. Our results demonstrate the potential for the proposed metrics to encapsulate the major trends in research and management of both threats and threatened taxa and provide a basis for international comparisons of evidence-based conservation science.

How to improve threatened species management: An Australian perspective

Targeted threatened species management is a central component of efforts to prevent species extinction. Despite the development of a range of management frameworks to improve conservation outcomes over the past decade, threatened species management is still commonly characterised as ad hoc. Although there are notable successes, many management programs are ineffective, with relatively few species experiencing improvements in their conservation status. We identify underlying factors that commonly lead to ineffective and inefficient management. Drawing attention to some of the key challenges, and suggesting ways forward, may lead to improved management effectiveness and better conservation outcomes. We highlight six key areas where improvements are needed: 1) stakeholder engagement and communication; 2) fostering strong leadership and the development of achievable long-term goals; 3) knowledge of target species' biology and threats, particularly focusing on filling knowledge gaps that impede management, while noting that in many cases there will be a need for conservation management to proceed initially despite knowledge gaps; 4) setting objectives with measurable outcomes; 5) strategic monitoring to evaluate management effectiveness; and 6) greater accountability for species declines and failure to recover species to ensure timely action and guard against complacency. We demonstrate the importance of these six key areas by providing examples of innovative approaches leading to successful species management. We also discuss overarching factors outside the realm of management influence that can help or impede conservation success. Clear recognition of factors that make species' management more straightforward - or more challenging - is important for setting realistic management objectives, outlining strategic action, and prioritising resources. We also highlight the need to more clearly demonstrate the benefit of current investment, and communicate that the risk of under-investment is species extinctions. Together, improvements in conservation practice, along with increased resource allocation and re-evaluation of the prioritisation of competing interests that threaten species, will help enhance conservation outcomes for threatened species.

Bandicoots return to Booderee: initial survival, dispersal, home range and habitat preferences of reintroduced southern brown bandicoots (eastern sub species; Isoodon obesulus obesulus)

Context Reintroductions can be an effective means of re-establishing locally extinct or declining faunal populations. However, incomplete knowledge of variables influencing survival and establishment can limit successful outcomes. Aim We aimed to examine the factors (e.g. sex, body mass, release order) influencing the survival, dispersal, home range and habitat selection of reintroduced southern brown bandicoots (eastern subspecies; Isoodon obesulus obesulus) into an unfenced, predator-managed environment in south-eastern Australia (Booderee National Park). Methods Over 2 weeks in May 2016, six female and five male bandicoots were wild-caught in state forest and hard released into the park. Release locations were approximately evenly distributed between three primary vegetation types assessed as suitable habitat: heath, woodland and forest. Bandicoots were radio-tracked day and night for 4 weeks from the initial release date. Key results No mortality was detected. Males dispersed more than twice as far as females (male 704 m, female 332 m), but there was no significant sex bias in home range size. At the landscape scale, bandicoots preferentially selected home ranges that contained heath and avoided forest. Within home ranges, heath and woodland were both favoured over forest. Conclusions Post-release dispersal is sex-biased, but more data are required to determine the influence of other predictors such as body mass and release order. Within the release area, bandicoots favoured non-forest vegetation types. Implications Our study outlines factors influencing the establishment of reintroduced bandicoots. We recommend that future bandicoot reintroductions to Booderee National Park occur within areas of heath and woodland, and that subsequent releases consider the potentially larger spatial requirements and conspecific avoidance among male bandicoots. Our findings contribute new knowledge for improving translocation methods of a nationally endangered medium-sized mammal.

Effects of a large wildfire on vegetation structure in a variable fire mosaic

Management guidelines for many fire-prone ecosystems highlight the importance of maintaining a variable mosaic of fire histories for biodiversity conservation. Managers are encouraged to aim for fire mosaics that are temporally and spatially dynamic, include all successional states of vegetation, and also include variation in the underlying "invisible mosaic" of past fire frequencies, severities, and fire return intervals. However, establishing and maintaining variable mosaics in contemporary landscapes is subject to many challenges, one of which is deciding how the fire mosaic should be managed following the occurrence of large, unplanned wildfires. A key consideration for this decision is the extent to which the effects of previous fire history on vegetation and habitats persist after major wildfires, but this topic has rarely been investigated empirically. In this study, we tested to what extent a large wildfire interacted with previous fire history to affect the structure of forest, woodland, and heath vegetation in Booderee National Park in southeastern Australia. In 2003, a summer wildfire burned 49.5% of the park, increasing the extent of recently burned vegetation (<10 yr post-fire) to more than 72% of the park area. We tracked the recovery of vegetation structure for nine years following the wildfire and found that the strength and persistence of fire effects differed substantially between vegetation types. Vegetation structure was modified by wildfire in forest, woodland, and heath vegetation, but among-site variability in vegetation structure was reduced only by severe fire in woodland vegetation. There also were persistent legacy effects of the previous fire regime on some attributes of vegetation structure including forest ground and understorey cover, and woodland midstorey and overstorey cover. For example, woodland midstorey cover was greater on sites with higher fire frequency, irrespective of the severity of the 2003 wildfire. Our results show that even after a large, severe wildfire, underlying fire histories can contribute substantially to variation in vegetation structure. This highlights the importance of ensuring that efforts to reinstate variation in vegetation fire age after large wildfires do not inadvertently reduce variation in vegetation structure generated by the underlying invisible mosaic.

Interactive effects of fire and large herbivores on web-building spiders

Altered disturbance regimes are a major driver of biodiversity loss worldwide. Maintaining or re-creating natural disturbance regimes is therefore the focus of many conservation programmes. A key challenge, however, is to understand how co-occurring disturbances interact to affect biodiversity. We experimentally tested for the interactive effects of prescribed fire and large macropod herbivores on the web-building spider assemblage of a eucalypt forest understorey and investigated the role of vegetation in mediating these effects using path analysis. Fire had strong negative effects on the density of web-building spiders, which were partly mediated by effects on vegetation structure, while negative effects of large herbivores on web density were not related to changes in vegetation. Fire amplified the effects of large herbivores on spiders, both via vegetation-mediated pathways and by increasing herbivore activity. The importance of vegetation-mediated pathways and fire-herbivore interactions differed for web density and richness and also differed between web types. Our results demonstrate that for some groups of web-building spiders, the effects of co-occurring disturbance drivers may be mostly additive, whereas for other groups, interactions between drivers can amplify disturbance effects. In our study system, the use of prescribed fire in the presence of high densities of herbivores could lead to reduced densities and altered composition of web-building spiders, with potential cascading effects through the arthropod food web. Our study highlights the importance of considering both the independent and interactive effects of disturbances, as well as the mechanisms driving their effects, in the management of disturbance regimes.

An empirical assessment of the focal species hypothesis

Biodiversity surrogates and indicators are commonly used in conservation management. The focal species approach (FSA) is one method for identifying biodiversity surrogates, and it is underpinned by the hypothesis that management aimed at a particular focal species will confer protection on co-occurring species. This concept has been the subject of much debate, in part because the validity of the FSA has not been subject to detailed empirical assessment of the extent to which a given focal species actually co-occurs with other species in an assemblage. To address this knowledge gap, we used large-scale, long-term data sets of temperate woodland birds to select focal species associated with threatening processes such as habitat isolation and loss of key vegetation attributes. We quantified co-occurrence patterns among focal species, species in the wider bird assemblage, and species of conservation concern. Some, but not all, focal species were associated with high levels of species richness. One of our selected focal species was negatively associated with the occurrence of other species (i.e., it was an antisurrogate)-a previously undescribed property of nominated focal species. Furthermore, combinations of focal species were not associated with substantially elevated levels of bird species richness, relative to levels associated with individual species. Our results suggest that although there is some merit to the underpinning concept of the FSA, there is also a need to ensure that actions are sufficiently flexible because management tightly focused on a given focal species may not benefit some other species, including species of conservation concern, such of which might not occur in species-rich assemblages.

Landscape genetics of an early successional specialist in a disturbance-prone environment

Species that specialize in disturbed habitats may have considerably different dispersal strategies than those adapted to more stable environments. However, little is known of the dispersal patterns and population structure of such species. This information is important for conservation because many postfire specialists are at risk from anthropogenic changes to natural disturbance regimes. We used microsatellite markers to assess the effect of landscape variation and recent disturbance history on dispersal by a small mammal species that occupies the early seral stage of vegetation regeneration in burnt environments. We predicted that a postfire specialist would be able to disperse over multiple habitat types (generalist) and not exhibit sex-biased dispersal, as such strategies should enable effective colonization of spatially and temporally variable habitat. We found significant differentiation between sites that fitted an isolation-by-distance pattern and spatial autocorrelation of multilocus genotypes to a distance of 2-3 km. There was no consistent genetic evidence for sex-biased dispersal. We tested the influence of different habitat- and fire-specific landscape resistance scenarios on genetic distance between individuals and found a significant effect of fire. Our genetic data supported recently burned vegetation having greater conductance for gene flow than unburnt habitat, but variation in habitat quality between vegetation types and occupied patches had no effect on gene flow. Postfire specialists must evolve an effective dispersal ability to move over distances that would ensure access to early successional stage vegetation. Natural disturbance and natural heterogeneity may therefore not influence population genetic structure as negatively as expected.

Wildfires, fuel treatment and risk mitigation in Australian eucalypt forests: insights from landscape-scale simulation

Wildfires pose significant risks to people and human infrastructure worldwide. The treatment of fuel in landscapes may alter these risks but the magnitude of this effect on risk is poorly understood. Evidence from Australian Eucalyptus forests suggests that mitigation of risk using prescribed burning as a fuel treatment is partial because weather and fuel dynamics are conducive to regular high intensity fires. We further examine the response of risk to treatment in eucalypt forests using landscape simulation modelling. We model how five key measures of wildfire activity that govern risk to people and property may respond to variations in rate and spatial pattern of prescribed fire. We then model effects of predicted climate change (2050 scenarios) to determine how the response of risk to treatment is likely to be altered in the future. The results indicate that a halving of risk to people and property in these forests is likely to require treatment rates of 7-10% of the area of the landscape per annum. Projections of 2050 weather conditions under climate change further substantially diminished the effect of rate of treatment. A large increase in rates of treatment (i.e. circa. 50% over current levels) would be required to counteract these effects of climate change. Such levels of prescribed burning are unlikely to be financially feasible across eucalypt dominated vegetation in south eastern Australia. Despite policy imperatives to expand fuel treatment, a reduction rather than an elimination of risk will result. Multi-faceted strategies will therefore be required for the management of risk.

Use and characteristics of nocturnal habitats of the squirrel glider (Petaurus norfocensis) in Australian temperate woodlands

In Australian temperate woodlands, most squirrel glider (Petaurus norfolcensis) habitats exist outside formal conservation reserves, often in highly fragmented agricultural landscapes. To conserve squirrel glider populations in such woodlands it is essential to define important habitats and understand how they are used. This study examines the nocturnal habitat use of squirrel gliders across five sites within an agricultural landscape in south-eastern Australia. Over a five-month period we radio-tracked 32 gliders to 372 nocturnal locations. We quantify characteristics of key nocturnal habitats and describe their use. Gliders were more likely to use large eucalypt trees, particularly yellow box (Eucalyptus melliodora) and mugga ironbark (E. sideroxylon). Nocturnal activity mostly took place high in the canopy of eucalypts, accounting for 74% of fixes. Multiple regression models revealed that feeding was more likely to occur in large, healthy trees close to drainage lines, with a preference for E. melliodora, when eucalypts were not flowering. Flowering trees were preferentially sought and were strongly associated with being large healthy trees that occur on ridges and upper slopes. Showing that the squirrel glider utilises key feeding structures (large healthy Eucalyptus trees) in different parts of the landscape at different times has direct management implications in the conservation and restoration of squirrel glider habitat, particularly in fragmented temperate woodland.

Refuge site selection by the eastern chestnut mouse in recently burnt heath

Context Temporal reduction in shelter is an indirect primary impact of fire. Removal of animal refugia has implications for shelter site selection and fidelity – factors that have been largely overlooked in studies of Australian rodent fauna. This information is critical for guiding species conservation and appropriate land management including prescribed burning practices. Aims We sought to determine which features of burnt heath were selected as shelter sites by the eastern chestnut mouse, whether there was sex and/or seasonal variation in shelter site selection and whether we could identify primary refugia. Methods We completed a radio-telemetry study to identify diurnal refuge sites and compare habitat attributes with those of a matched set of control sites. We then used habitat features and fidelity parameters to classify refuge site use. Key results We found the eastern chestnut mouse selected shelters with the presence of specific structures and had taller, denser vegetation than randomly selected control sites. There were no differences in habitat selection between the sexes. Shelter sites in the non-breeding season had greater vegetation density compared with those used in the breeding season. In the breeding season, the eastern chestnut mouse showed no evidence of increased fidelity to particular refuges. Vegetation density in winter was the best predictor of a primary refuge compared with whether or not an animal returned to a shelter site or the amount of time spent in a shelter site. Mice were ephemeral and non-gregarious in their refuge use. There was some evidence for inheritance of refuge sites from a parent, as well as inter-season shelter site fidelity. Conclusions The eastern chestnut mouse selected refugia that had habitat attributes offering maximum protection. Seasonality in refuge site selection is likely to reflect the reproductive and environmental trade-offs in critical resources during different seasons. The maintenance of multiple, rarely occupied shelters by the eastern chestnut mouse is consistent with data for other mammals. Implications Fire management should ensure retention of vegetation structure on the ground layer, dense habitat patches in burned areas, and be carefully planned during the winter season to maintain shelter and refuge sites to assist population persistence.

The future of scattered trees in agricultural landscapes

Mature trees scattered throughout agricultural landscapes are critical habitat for some biota and provide a range of ecosystem services. These trees are declining in intensively managed agricultural landscapes globally. We developed a simulation model to predict the rates at which these trees are declining, identified the key variables that can be manipulated to mitigate this decline, and compared alternative management proposals. We used the initial numbers of trees in the stand, the predicted ages of these trees, their rate of growth, the number of recruits established, the frequency of recruitment, and the rate of tree mortality to simulate the dynamics of scattered trees in agricultural landscapes. We applied this simulation model to case studies from Spain, United States, Australia, and Costa Rica. We predicted that mature trees would be lost from these landscapes in 90-180 years under current management. Existing management recommendations for these landscapes--which focus on increasing recruitment--would not reverse this trend. The loss of scattered mature trees was most sensitive to tree mortality, stand age, number of recruits, and frequency of recruitment. We predicted that perpetuating mature trees in agricultural landscapes at or above existing densities requires a strategy that keeps mortality among established trees below around 0.5% per year, recruits new trees at a rate that is higher than the number of existing trees, and recruits new trees at a frequency in years equivalent to around 15% of the maximum life expectancy of trees. Numbers of mature trees in landscapes represented by the case studies will decline before they increase, even if strategies of this type are implemented immediately. This decline will be greater if a management response is delayed.

Contrasting mammal responses to vegetation type and fire

The response of terrestrial mammals and arboreal marsupials to past burning history as well as a year prior to, and then for 4 years after, a major wildfire in 2003 at Booderee National Park, Jervis Bay Territory was quantified. The present study encompassed extensive repeated surveys at a set of 109 replicated sites stratified by vegetation type and fire history. It was found that most species exhibited significant differences in presence and abundance between major vegetation types. Detections of long-nosed bandicoot (Perameles nasuta) increased significantly in all vegetation types surveyed, in both burnt and unburnt areas. Temporal patterns in captures of three species of small mammals (bush rat (Rattus fuscipes), swamp rat (Rattus lutreolus) and brown antechinus (Antechinus stuartii)) showed a trend for lower numbers of captures on burnt sites compared with unburnt sites. Three species of arboreal marsupials, common ringtail possum (Pseudocheirus peregrinus), greater glider (Petauroides volans) and common brushtail possum (Trichosurus vulpecula), were moderately common and all showed marked differences in abundance between vegetation types. Whereas P. peregrinus and P. volans exhibited a temporal decline between 2003 and 2006, T. vulpecula exhibited a general increase from 2003 levels. However, arboreal marsupial responses did not appear to be directly fire related.

The characteristics of den trees used by the squirrel glider (Petaurus norfolcensis) in temperate Australian woodlands

Being able to recognise critical habitat features such as nesting and denning sites is essential for wildlife conservation. It is particularly true for the den trees of species threatened by habitat loss, such as the squirrel glider (Petaurus norfolcensis). Measurements of 146 den trees of the squirrel glider were made in fragmented woodlands of the South-west Slopes of New South Wales and data compared with random trees to quantify the key characteristics of den sites. The likelihood of use as a den tree increased with increasing numbers of visible tree hollows and tree size. Dieback was also a positive indicator of den use. However, when visible hollows within a tree are abundant, dieback-free trees were preferred. Measures of den tree context such as basal area, the number of neighbouring large trees and distance to the next nearest tree, were also found to be important determinants of the likelihood of usage. The above variables were combined into a multiple regression model. The squirrel glider favoured particular Eucalyptus species and some broader eucalypt groups. We believe such variations were most likely due to interspecific differences in hollow development and dieback among the various groups, rather than bark type, a factor previously cited as an important determinant of den tree usage. The ‘best’ model had high negative predictive power, suggesting it would be useful for identifying (1) trees that could be felled without a loss of this critical habitat resource (e.g. at development sites) and (2) areas unsuitable for potential squirrel glider relocation or habitat enhancement. Squirrel gliders show preference for a combination of tree and tree context features in selecting den trees. Understanding these features will help managers enhance and protect denning resources for this species.

Salvage logging in the montane ash eucalypt forests of the Central Highlands of Victoria and its potential impacts on biodiversity

The two major forms of disturbance in the montane ash eucalypt forests of the Central Highlands of Victoria (southeastern Australia) are clearfell logging and unplanned wildfires. Since the 1930s wildfire has been followed by intensive and extensive salvage-logging operations, which may proceed for many years after a wildfire has occurred. Although applied widely, the potential effects of salvage logging on native flora and fauna have been poorly studied. Our data indicate that the abundance of large trees with hollows is significantly reduced in forests subject to salvage harvesting. This has implications for thepersistence of an array of such cavity-using vertebrates as the endangered arboreal marsupial, Leadbeater's possum (Gymnobelidues leadbeateri). Salvage logging also reduces the prevalence of multiaged montane ash forests--places that typically support the highest diversity of arboreal marsupials and forest birds. Limited research has been conducted on the effects of salvage logging on plants; thus, we constructed hypotheses about potential impacts for further testing based on known responses to clearfell logging and key life history attributes. We predict many species, such as vegetatively resprouting tree ferns, will decline, as they do after clearfelling. We also suggest that seed regenerators, which typically regenerate well after fire or conventional clearfelling, will decline after salvage logging because the stimulation for germination (fire) takes place prior to mechanical disturbance (logging). Understoryplant communities in salvage-logged areas will be dominated by a smaller suite of species, and those that are wind dispersed, have viable soil-stored seed remaining after salvage logging, or have deep rhizomes are likely to be advantaged. We recommend the following improvements to salvage-logging policies that may better incorporate conservation needs in Victorian montane ash forests: (1) exemption of salvage logging from some areas (e.g., old-growth stands and places subject to only partial stand damage); (2) increased retention of biological legacies on burned areas through variations in the intensity of salvage logging; and (3) reduction in the levels of physical disturbance on salvage-logged areas, especially through limited seedbed preparation and mechanical disturbance.

Salvage logging, ecosystem processes, and biodiversity conservation

We summarize the documented and potential impacts of salvage logging--a form of logging that removes trees and other biological material from sites after natural disturbance. Such operations may reduce or eliminate biological legacies, modify rare postdisturbance habitats, influence populations, alter community composition, impair natural vegetation recovery, facilitate the colonization of invasive species, alter soil properties and nutrient levels, increase erosion, modify hydrological regimes and aquatic ecosystems, and alter patterns of landscape heterogeneity These impacts can be assigned to three broad and interrelated effects: (1) altered stand structural complexity; (2) altered ecosystem processes and functions; and (3) altered populations of species and community composition. Some impacts may be different from or additional to the effects of traditional logging that is not preceded by a large natural disturbance because the conditions before, during, and after salvage logging may differ from those that characterize traditional timber harvesting. The potential impacts of salvage logging often have been overlooked, partly because the processes of ecosystem recovery after natural disturbance are still poorly understood and partly because potential cumulative effects of natural and human disturbance have not been well documented. Ecologically informed policies regarding salvage logging are needed prior to major natural disturbances so that when they occur ad hoc and crisis-mode decision making can be avoided. These policies should lead to salvage-exemption zones and limits on the amounts of disturbance-derived biological legacies (e.g., burned trees, logs) that are removed where salvage logging takes place. Finally, we believe new terminology is needed. The word salvage implies that something is being saved or recovered, whereas from an ecological perspective this is rarely the case.

The effects of habitat fragmentation via forestry plantation establishment on spatial genotypic structure in the small marsupial carnivore, Antechinus agilis

Dispersal is an important influence on species' distributions, patch colonization and population persistence in fragmented habitat. We studied the impacts of habitat fragmentation resulting from establishment of an exotic pine plantation on dispersal of the marsupial carnivore, Antechinus agilis. We applied spatial analyses of individual multilocus microsatellite genotypes and mitochondrial haplotypes to study patterns of gene flow in fragmented habitat and natural habitat 'control' areas, and how this is affected by the spatial dispersion of habitat patches, the presence of corridors and a 'mainland' source of migrants. Spatial analysis of molecular variance and partial Mantel tests confirmed the absence of cryptic barriers to gene flow in continuous habitat, which if present would confound the comparison of genetic structures in fragmented vs. unfragmented habitats. Spatial genotypic structure suggested that although dispersal was male-biased in both habitat types, fragmentation restricted dispersal of males more than that of females and the degree of restriction of male dispersal was dependent on the geographical isolation of the patch. The scale of positive genotypic structure in fragmented habitat was restricted to the two closest patches for females and the three closest patches for males. Our results provide evidence for significantly increased gene flow through habitat corridors relative to that across the matrix and for significantly lower gene flow between 'mainland' unfragmented habitat and habitat patches relative to that within either habitat type, suggesting a behavioural barrier to crossing habitat interfaces.

The effects of habitat fragmentation on the social kin structure and mating system of the agile antechinus, Antechinus agilis

Habitat fragmentation is one of the major contributors to the loss of biodiversity worldwide. However, relatively little is known about its more immediate impacts on within-patch population processes such as social structure and mating systems, whose alteration may play an important role in extinction risk. We investigated the impacts of habitat fragmentation due to the establishment of an exotic softwood plantation on the social kin structure and breeding system of the Australian marsupial carnivore, Antechinus agilis. Restricted dispersal by males in fragmented habitat resulted in elevated relatedness among potential mates in populations in fragments, potentially increasing the risk of inbreeding. Antechinus agilis nests communally in tree hollows; these nests are important points for social contact between males and females in the mating season. In response to elevated relatedness among potential mates in fragmented habitat, A. agilis significantly avoided sharing nests with opposite-sex relatives in large fragment sites (but not in small ones, possibly due to limited nest locations and small population sizes). Because opposite-sex individuals shared nests randomly with respect to relatedness in unfragmented habitat, we interpreted the phenomenon in fragmented habitat as a precursor to inbreeding avoidance via mate choice. Despite evidence that female A. agilis at high inbreeding risk selected relatively unrelated mates, there was no overall increased avoidance of related mates by females in fragmented habitats compared to unfragmented habitats. Simulations indicated that only dispersal, and not nonrandom mating, contributed to inbreeding avoidance in either habitat context. However, habitat fragmentation did influence the mating system in that the degree of multiple paternity was reduced due to the reduction in population sizes and population connectivity. This, in turn, reduced the number of males available to females in the breeding season. This suggests that in addition to the obvious impacts of reduced recruitment, patch recolonization and increased genetic drift, the isolation of populations in habitat patches may cause changes in breeding behaviour that contribute to the negative impacts of habitat fragmentation.

The life history of Pseudocheirus occidentalis (Pseudocheiridae) in the jarrah forest of south-western Australia

Life-history attributes are described for the threatened ngwayir or western ringtail possum (Pseudocheirus occidentalis) in inland jarrah (Eucalyptus marginata) forest east of Manjimup, south-western Australia. Data on 81 individuals were collected over 18 months. There was no sexual dimorphism and body size was similar to that found in other P. occidentalis populations, but larger than the closely related P. peregrinus in eastern Australia. Breeding at Chariup was more strongly seasonal than that of coastal populations, with 77% of births in May–June and the remainder in October–November. All neonates were singletons except for one instance of non-viable twins. No females bred twice in the same year. The growth rate of the head length of pouch young (<5 months of age) was 0.245–0.362 mm day–1 and curvilinear toward an asymptote thereafter. Temporal variations in body condition, coat condition and ectoparasites were significant. Mortality was highly seasonal (84% of deaths were April–September) and predominantly caused by predation, mainly by fox (Vulpes vulpes) and cat (Felis catus). More effective and strategic control of introduced predators prior to and during autumn/winter, could therefore improve the viability of jarrah forest populations. Nutrition appears to influence many of the life-history traits of P. occidentalis. Nutrition also may partly explain the differences in size, life history and conservation status between P. occidentalis and P. peregrinus.

A comparison of survey methods for arboreal possums in jarrah forest, Western Australia

Comparative trials of different survey methods were conducted in the southern jarrah (Eucalyptus marginata) forest to determine the most efficient means of detecting koomal (common brushtail possum, Trichosurus vulpecula hypoleucus) and ngwayir (western ringtail possum, Pseudocheirus occidentalis). In particular, we examined different trapping and spotlighting methods and compared these with scat surveys. Six different trapping methods (derived by combining three bait types and two trap positions) were compared at six sites. Significantly fewer koomal were caught on ‘universal’ bait (i.e. peanut butter, rolled oats and sardines) than on flour-based baits using rose oil or Eucalyptus oil as lures. Significantly more individuals of both possum species were caught in arboreal traps than in ground traps (P < 0.001 in both cases). Recapture rates of koomal were high, whereas ngwayir were rarely retrapped. There were no detection differences between six different spotlighting methods (derived by combining three spotlight intensities with two filter colours) for koomal. Significantly more ngwayir were detected using 50-W or 100-W lights than 20-W lights (P = 0.01). There were no significant differences in the detection rates for ngwayir using red or white light. There were, however, significant observer differences in the number of possums of both species detected (koomal, P = 0.025; ngwayir, P = 0.004). Spotlighting detected, on average, only 4.9% of the koomal ‘known to be alive’ by trapping. However, spotlighting with a 50-W or 100-W spotlight detected more ngwayir than did trapping. Koomal abundance measures derived from scat surveys were not related to trapping or spotlight abundance estimates. For ngwayir, however, scat counts were strongly related to spotlight counts and there were no significant observer differences for the former. We conclude that koomal are more effectively surveyed using arboreal trapping with rose or Eucalyptus lures. Ngwayir are best surveyed using scat surveys or 50-W spotlights.

Effects of trap position, trap history, microhabitat and season on capture probabilities of small mammals in a wet eucalypt forest

In this study, we use data drawn from a series of trapping events on four 0.5-ha trapping grids surveyed in the wet eucalypt forests of central Victoria, south-eastern Australia, to identify relationships between capture probabilities and several factors of interest for three species of small mammals that are common throughout the forests of this region: the agile antechinus (Antechinus agilis), the dusky antechinus (Antechinus swainsonii) and the bush rat (Rattus fuscipes). The design of our study – four regular trapping grids – generated spatio-temporal data with binary responses and many covariates. We used powerful and relatively new statistical methodology to deal with the spatio-temporal dependence patterns in the data – analytical problems that are common in trapping data such as these modelled here. Although A. agilis, A. swainsonii and R. fuscipes are among the best studied mammals in Australia, our data analysis produced new perspectives on their probability of being captured. In particular, we quantified how capture probability is affected by trap position within a trapping grid, day of capture in a sequence of trapping days, history of trap occupancy over time by different species and sexes of those species, time of the year or season, and microhabitat attributes. Our insights are discussed in terms of their consequences for trapping protocols that might be applied in the field.

The life history of Trichosurus vulpecula hypoleucus (Phalangeridae) in the jarrah forest of south-western Australia

The common brushtail possum (Trichosurus vulpecula) is noted for its morphological, biological and ecological variability across its range. Despite having suffered substantial population declines since European settlement, relatively little has been published on the south-western Australian subspecies, the koomal (T. v. hypoleucus). This study reports morphological, reproductive and general life-history data from an 18-month study of a population in the southern jarrah (Eucalyptus marginata) forest at Chariup (part of Perup), near Manjimup, in south-western Australia. As one of the smallest subspecies, adult males of T. v. hypoleucus averaged 1616 g and females averaged 1470 g. Sexual dimorphism also occurred with head length and pes length, but not tail length. A single autumn breeding season occurred in both 2002 and 2003, in which all adult females bred and produced a single young between February and May. The onset of autumn births was associated with the end of the summer drought. Unlike many other Trichosurus populations, no spring breeding pulse or ‘double-breeding’ events were observed. At least 83% of pouch young survived to pouch emergence. The growth rate of offspring was initially linear, but became curvilinear and approached an asymptote after ~5 months. Most females bred for the first time when they were 1 year old. On the basis of testis size, males also matured at 1 year old. The body condition of adult males, but not adult females, changed significantly over time and followed an apparently seasonal pattern in which their condition was poorest in winter and best in summer. While many of the life-history traits of the Chariup population were similar to those of other south-western Australian populations of T. v. hypoleucus, the most striking variations included age at maturity, extent of spring breeding pulse and female fecundity. Further comparisons with conspecifics elsewhere in Australia and New Zealand also highlight the variability exhibited by T. vulpecula across its range. Some aspects of the biology of T. v. hypoleucus were particularly similar to those observed for T. v. arnhemensis in northern Australia.

Patch use by the greater glider (Petauroides volans) in a fragmented forest ecosystem. II. Characteristics of den trees and preliminary data on den-use patterns

This paper presents the results of a study of the use of den trees by the greater glider (Petauroides volans) in five patches of remnant eucalypt forest embedded within an extensive radiata pine (Pinus radiata) plantation near Tumut in south-eastern Australia. Radio-tracking was used to identify 171 den trees occupied by 40 animals over 948 animal-tracking days between September 1997 and September 1998. All radio-tracked P. volans used multiple den trees. Males used significantly more den trees than females and a greater proportion of these were used for single visits. Males also used fewer new den trees over the study period in the smaller patches, although they still used more than females. In the larger patches, males and females used similar numbers of den trees. Commonly used den trees tended to be situated in (or close to) core areas of an individual’s home range. Den tree sharing, either concurrently or independently, was predominantly between adult males and females, or between adults and their young. Trees most likely to be used by more than one individual had the same characteristics as trees that had the highest probability of use per se – that is, they were of large ‘average size’.

Sound recording of bird vocalisations in forests. I. Relationships between bird vocalisations and point interval counts of bird numbers - a case study in statistical modeling

Morning vocal activity data for birds were collected using automatic sound recorders at 165 sites in fragmented forests at Tumut in south-eastern Australia. A model was developed to describe the statistical properties of the vocal activity data and study the relationship between the number of 'elements' (or notes) broadcast by birds and the number of birds recorded by human observers. We discuss the practical issues of finding a model where variance heterogeneity and skewness is a feature of the data, while the context of the problem required that multiplicative relationships be preserved. A satisfactory fit to our vocal activity data was obtained by using a Gamma distribution.For most groups of birds, the observed relationships between vocal activity and the number of birds recorded by human point counts were weak but statistically significant.Our analysis suggests that these data provide limited useful information about vocal activity per bird. Automatic sound-recording data may be informative for drawing inferences about temporal patterns in vocal activity but do not seem useful as a method for estimating the abundance of birds. A significant relationship was identified between the number of species vocalising, as measured by sound recorders, and the number of species observed by point count method. However, the relationship was far from isometric and there were many more species detected by the sound recorders than by human observers.Results from analyses of vocal activity data gathered at Tumut were confirmed by additional data obtained in a subsequent, small calibration study in the Namadgi National Park in the Australian Capital Territory.

Patch use by the greater glider (Petauroides volans) in a fragmented forest ecosystem. III. Night-time use of trees

Night-time use of feed trees by 40 radio-collared individuals of the greater glider (Petauroides volans) was recorded within five remnant patches of eucalypt forest near Tumut in south-eastern Australia. Radio-collared animals were observed making 663 night-time visits to 433 trees. For these observations, we recorded the number of visits by an animal to each tree, the number of different animals using each tree, the characteristics of trees that animals used, and the category of activity or behaviour displayed by animals (classified as feeding, perching, and moving). We found no evidence of a significant difference in patterns of behaviour between male and female P. volans in their night-time use of trees within remnant patches, regardless of patch size or population density. There were few records of animals (4%) from the radiata pine (Pinus radiata) plantation that surrounded the eucalypt remnants, although 20 observations were made of P. volans feeding on the young male cones and buds of this introduced tree species. There was evidence of preference for feeding in ribbon gum (Eucalyptus viminalis), mountain gum (E. dalrympleana) and narrow-leaved peppermint (E. radiata), with 72% of feeding observations of P. volans coming from these three tree species. Animals were generally solitary and spent most of the night feeding and perching in the upper canopy. Most trees were used by a single individual, with 96% of observations being of a single animal in a tree. The limited number of cases of sharing and co-use of trees were generally between an adult male and adult female (assumed mates), and females and their young. The probability that a tree was used increased with the average size of a tree (a composite measure of diameter, height and crown features) until approaching an asymptote of 1.0, i.e. all large trees were used. The number of visits a tree received from P. volans also was positively related to the measure of its size.

Sound recording of bird vocalisations in forests. II. Longitudinal profiles in vocal activity

As early morning bird vocalisation is a major feature of many bird communities, longitudinal profiles of vocal activity data, collected using sound recorders, were compared for a range of habitat types in the Tumut area of south-eastern Australia. There was a significant, and roughly linear, decline in vocal activity across the morning after an initial early peak of activity. Vocal activity persisted longer at sites located within large areas of continuous eucalypt forest than in the strip- and patch-shaped eucalypt remnants surrounded by extensive stands of radiata pine or at sites dominated by stands of radiata pine. There was evidence that the pattern of persistence of vocal activity differed among the different bird groups.

Patch use by the greater glider (Petauroides volans) in a fragmented forest ecosystem. I. Home range size and movements

This paper examines home-range attributes of 40 greater gliders (Petauroides volans) in five patches of remnant eucalypt forest surrounded by stands of radiata pine (Pinus radiata) near Tumut in south-eastern Australia. Fixed-kernel smoothing methods were used to estimate home-range size for P. volans. For males, home-range size varied from 1.38–4.10 ha (mean = 2.6 ± 0.8 ha, n = 12) and was significantly larger (P < 0.05) than for females (1.26–2.97 ha, mean = 2.0 ± 0.6 ha, n = 11). Home-range size increased significantly with increasing patch size and reduced patch population density. Thus, small patches had more animals per unit area with smaller home ranges and greater home-range overlap. Our findings illustrate flexibility in the use of space by P. volans. Such results have not previously been reported for P. volans or any other species of arboreal marsupial. Considerable home-range overlap (at 95th percentile isopleth level) was observed between male and female P. volans. Pairs of females also exhibited home-range overlap. Males tended to maintain home ranges exclusive of other males, although some shared common areas. Contrary to the large variations observed in home-range area, core areas (50th isopleth) remained relatively constant, regardless of patch size, population density or sex. This may indicate that core areas are an essential requirement for individuals and resources they contain cannot be shared with congeners.

The use of nest boxes by arboreal marsupials in the forests of the Central Highlands of Victoria

The results are reported of a nest-box study conducted in two locations in the mountain ash (Eucalyptus regnans) forests of the Central Highlands of Victoria (south-eastern Australia) to compare usage of different nest-box designs located at different heights in trees. A total of 96 nest boxes was established using a rigorous experimental design – two regions (Powelltown and Toolangi State Forests), two forest age classes (20-year post-logging regrowth and 60-year fire- and salvage-logging regrowth), two nest-box designs (large boxes with large entrance holes and small boxes with small entrance holes), and two heights at which nest boxes were attached to trees (3 m and 8 m above the ground). The study entailed setting out four nest boxes at each of 24 sites to meet the design criteria. Evidence of occupancy by vertebrates was recorded in a total of 19 of 96 boxes on 11 of 24 sites site during regular inspections over more than three years. Thirteen boxes were used by Leadbeater's possum (Gymnobelideus leadbeateri), six by the mountain brushtail possum (Trichosurus cunninghami) and seven by the common ringtail possum (Pseudocheirus peregrinus). The common ringtail possum and mountain brushtail possum were seen only in high–large boxes but Leadbeater's possum used all but the low–large boxes. There was evidence of spatial dependence in usage patterns, with all four boxes at a given site showing signs of eventually being occupied. Only two nest boxes located in mountain ash forest regenerating after the 1939 wildfires were occupied. Relatively limited use of nest boxes supports concerns about the use of a nest box over large scales and long timeframes as an effective recovery tool for species threatened by the loss and subsequent shortage in the numbers of naturally occurring hollows.

Landscape surrogates of forest fragmentation: Synthesis of Australian Montreal Process case studies

Habitat loss and fragmentation are key biodiversity indicators of the Montreal Protocol for monitoring progress towards ecologically sustainable forest management. Over the last 15 years, an array of landscape metrics have been developed as spatial measures of habitat loss and fragmentation. However, most metrics require rigorous empirical testing if they are to provide scientifically credible information to managers and policy makers. We present a synthesis of three Australian case studies for developing Montreal Indicator 1.1e, fragmentation of forest type, each representing different levels of landscape modification: St Mary State Forest, south-east Queensland; Tumut, southern New South Wales; and the Central Highlands, Victoria. Collectively, the studies found that no single landscape metric captured the response of the target species and fauna assemblages, or served as a reliable ecological surrogate for the conservation of a large set of species. Rather, species demonstrated a diversity of responses to habitat loss and fragmentation. Fragmentation effects were more important for the Tumut study, but not important for the Central Highlands study. Stand-scale habitat variables and area of suitable habitat were dominant explanatory variables for the St Mary study. Differences in observed response are partly explained by: (i) differences in landscape structure, particularly the proportion of preferred forest habitat remaining; (ii) differences in the ecology of target species; and (iii) the insensitivity of the landscape measures. Based on the outcomes of the three case studies, we propose principles for developing landscape surrogates for conserving biodiversity in Australia's eucalypt forest landscapes.

Dispersal patterns and population structure in a small marsupial, Antechinus agilis, from two forests analysed using microsatellite markers

Juvenile male agile antechinus (Antechinus agilis) are thought to disperse immediately after they leave the nest, whereas females are philopatric. However, no genetic data are available to support a hypothesis of male-biased dispersal, and the possible effects of habitat characteristics on dispersal are not known. We used five microsatellite loci to describe the dispersal patterns and population structure of A. agilis at two different forested sites (Mt Donna Buang and Mt Disappointment in central Victoria, south-eastern Australia). Within each site, DNA samples were collected from individuals from different trapping grids (four grids in Mt Donna Buang, collected over two years, and seven grids at Mt Disappointment, collected over one year). To assess levels of genetic structuring among the populations, Fst values (the proportion of genetic variance among populations) were calculated for each site and sex separately, and a test for isolation by distance was performed. Overall, Fst values were low, and did not increase with distance, and this may be the result of high levels of gene flow for both sexes at both sites, at least on the scale of less than 10 km. To investigate microsatellite patterns at a finer scale, we calculated pair-wise relatedness values separately for sites, years and sexes. By comparing these values for animals within and between grids, evidence was found for male-biased dispersal. A genetic assignment test provided further evidence for this: males were more likely to be assigned to grids other than their capture site. Most females had high assignment values for their capture site, although some of the females may have been immigrants. This result could be due to inaccuracies in the assignment test, or due to a higher frequency of dispersal of females than previously thought. The sex-biased dispersal was less pronounced at Mt Disappointment than at Mt Donna Buang. This may be related to the level of disturbance: logging patches, roads and fire-breaks (grass strips) disrupt the forest at Mt Disappointment more than at Mt Donna Buang.

Geographic dimorphism in the mountain brushtail possum (Trichosurus caninus): the case for a new species

The morphological and genetic characteristics of the mountain brushtail possum (Trichosurus caninus) are described for animals from a range of locations throughout its known geographic distribution in eastern Australia. Although there is considerable variation among populations, unequivocal morphological and genetic differences exist between northern and southern populations of the species. Specimens from southern populations (from Victoria) have a significantly (P < 0.001) larger ear conch, a significantly (P < 0.001) longer pes, and a significantly (P < 0.001) shorter tail than do specimens from northern populations (from New South Wales and Queensland). Animals can be clearly distinguished using a simple index based on these three morphological measures, which are gathered from live animals. North–south dimorphism is strongly supported by patterns in genetic data that show genetic distances of 2.7–3.0% between the southern and northern populations. The combined outcomes of morphological and genetic analyses suggest the existence of two distinct species. We recommend that the northern form, distributed from central New South Wales north to central Queensland, retain the name Trichosurus caninus; the southern form from Victoria is described here as Trichosurus cunninghami, sp. nov. The common names of these new species should be the 'short-eared possum' and the 'mountain brushtail possum', respectively.

The ecological roles of logs in Australian forests and the potential impacts of harvesting intensification on log-using biota

A review is presented of the ecological values of logs in Australian eucalypt forests. Logs are a key component of stand structural complexity and have critical functional roles for forest biodiversity including:- (1) providing nesting and sheltering sites for biota, (2) providing foraging substrates for predators like snakes and predatory invertebrates such as velvet worms, (3) providing basking and hibernation sites for reptiles, (4) facilitating animal movement, (5) providing places for key social behaviours, (6) acting as plant germination sites, (7) providing substrates to promote the growth of fungi, (8) providing mesic refugia for organisms during drought and/or fire, and (9) contributing to heterogeneity in the litter layer and patterns of ground cover. Logs also play significant roles in nutrient cycling in forests. The role of logs is often ignored in forestry operations, including those where harvesting intensification will occur through the removal of dead and/or "defective" standing trees and logs under the guise of removing so-called waste or logging "residues". Recently proposed intensive large-scale forestry operations in the Australian native forest estate (e.g., biomass burning power plants and charcoal plants) have the potential to reduce stand structural complexity, alter forest ecosystem function and negatively impact upon log-dependent species in those part of the landscape where harvesting takes place. The risks of such impacts have not been adequately measured in Australia, but they need to be addressed urgently. Prescriptions for the retention and future recruitment of logs must be developed to avert possible losses of biodiversity.

Hematological and plasma biochemical values of the greater glider in Australia

Reference hematological and plasma biochemical values are presented for the greater glider (Petauroides volans) at Tumut (southeastern New South Wales, Australia). Nineteen animals were sampled during a capture period of 1 wk in August 1999. Values for red cell counts were significantly higher in male animals (mean +/- SE; males: 5.6 +/- 0.1; females: 5.2 +/- 0.1). Young animals had higher white cell counts than older ones (mean +/- SE; young: 4.9 +/- 0.4; older: 2.8 +/- 0.4). Lymphocytes were the predominant white blood cell type in this species.