ABSTRACT The catastrophic 2009 wildfires in the Mountain Ash ( Eucalyptus regnans ) forests of the Central Highlands of Victoria provided an opportunity to gain new insights into the responses to fire by various elements of the biota. Ongoing long-term monitoring at a large number of permanent field sites for up to 25 years prior to the fire, together with 10 years of post-fire monitoring, has provided an unparalleled series of datasets on mammal, bird, and plant responses on burned and unburned sites. The empirical studies briefly summarized in this paper show patterns of steep declines in large old trees and declines in site occupancy by arboreal marsupials and birds. These changes contrast markedly with the responses of the two most common species of small mammals (the Agile Antechinus [ Antechinus agilis ] and Bush Rat [ Rattus fuscipes ]), which recovered within two generations after the fire. Declines in arboreal marsupials, birds and large old trees have also occurred on unburned sites, indicating an ecosystem-wide trend. In general, logging had a greater impact than fire on the majority of groups of birds and plants, particularly post-fire salvage logging that occurred in some areas following the 2009 wildfires. Beyond interactions between fire and post-fire (salvage) logging and their effects on forest biota, we have uncovered evidence of other kinds of interactions in Mountain Ash forests. These include interactions between: (1) the severity of fires and logging history, (2) post-fire bird population recovery and long-term climate and short-term weather conditions, and (3) impacts on forest soils. The structure and landscape composition of the Mountain Ash ecosystem has been radically altered over the last century. This has resulted from the combined impact of several large fires, including the 2009 fires as well as widespread clearfell logging that has been conducted within state forests over the last 50 years. The ecosystem now supports old growth cover that is 1/30 th to 1/60 th of what it was estimated to have been prior to European settlement. The ongoing decline of key components of the Mountain Ash ecosystem has led to it being classified as Critically Endangered and at high risk of ecosystem collapse. We argue that current forest policy and practices need to better mitigate the effects of fire on this already highly disturbed forest and enhance the possible persistence of species in this ecosystem. Several key strategies are required to do this. First, there is a need to significantly expand the extent of old growth within the Mountain Ash forest estate. This is because fire severity is diminished in such areas. Spatial contagion across old-growth dominated landscapes also may be suppressed relative to landscapes composed primarily of young forest. Allied management strategies include the protection of more mesic parts of Mountain Ash landscapes as these are less likely to burn or at least burn at high severity. Such enhanced protection should include an expanded network of buffers around drainage lines and waterways as these are where fire severity is likely to be lowest and also where old growth elements like large old hollow-bearing trees are more abundant. In addition, all existing living and dead hollow-bearing trees need to be protected by buffers of unlogged forest within wood production forests to promote their standing life and better conserve cavity-dependent fauna such as the Critically Endangered Leadbeater’s Possum ( Gymnobelideus leadbeateri ) and other declining taxa like the Greater Glider ( Petauroides volans ).

ABSTRACT The majority of studies into the response of birds to logging in Australian forests has been done in forest-dominated landscapes or relatively large forest blocks, where the surrounding landscape can ameliorate impacts. This is the first study to examine the response of birds to logging of a small, relatively isolated woodland remnant in a landscape dominated by agriculture, with a focus on declining woodland birds. Approximately two thirds of a 120 ha cypress-eucalypt woodland remnant was selectively logged. Eighty bird species in total were recorded 2–3½ years after logging, including 18 recognised as declining woodland birds. Sixty-four species were recorded in the unlogged area and 72 in the logged area. Of the 31 species recorded sufficiently frequently for analysis, 19 showed no statistical difference in occurrence between logged and unlogged areas at the power of the study, nine were more prevalent in unlogged areas and three more prevalent in logged areas. Declining woodland birds comprised 22½% of overall species and 33% of those more prevalent in unlogged areas. Mapping of records enabled the response of a subset of species to be examined in greater detail. Four species illustrate the range of responses by declining woodland birds: Eastern Yellow Robin Eopsaltria australis was confined to unlogged areas, White-browed Babbler Pomatostomus superciliosus favoured unlogged areas, Varied Sittella Daphoenositta chrysoptera favoured proximity to unlogged areas and Red-capped Robin Petroica goodenovii was widely distributed in both logged and unlogged areas. The likelihood of recovery of local woodland bird populations is discussed in the context of limited landscape connectivity and the recent colonisation of the remnant by the hyper-aggressive Noisy Miner Manorina melanocephala . White-browed Babbler and Eastern Yellow Robin are considered to be at high risk of local extinction in the remnant.

ABSTRACT Long-term research is vital for understanding long-lived, dynamic systems like forests, yet for bats such research is notable for its absence. Recent advances in technology over the last 10–20 years means that long-term studies of bats are viable and can be cost-effective. I outline three case-studies of long-term research on forest bats using different techniques. These involve annual monitoring of populations at cave roosts, periodic monitoring of forest bat activity with ultrasonic detectors and advanced mark-recapture modelling of annual banding data to describe population dynamics. Annual counts over a 17 year period of a large maternity population of Eastern Horseshoe Bats Rhinolophus megaphyllus in a sandstone cave revealed a relatively stable population in the absence of major perturbations. There were only minor variations with climatic fluctuations, suggesting the cave and associated feeding grounds in a high rainfall gully could function as a climate refuge for the species. The second case study used acoustic detectors over a 14 year period to describe changes in bat use of forest regenerating (harvesting/fire) from the 1970s when specific environmental protection was not practised in New South Wales (NSW) forests. Forest regeneration was a dynamic process, especially in relation to changing vegetation density. Dense regrowth regenerating from harvesting rapidly dominated and persisted over a considerable period of time and provided unsuitable habitat for most bat species. Our third case study used mark-recapture of banded bats to estimate population dynamics in an experimental forest with environmental protection similar to that practised today in NSW forests (protection of riparian zones; habitat trees, rainforest). Analysis of 14 years of annual banding data demonstrated little effect (either no effect, or minor positive and negative effects) of harvesting history on survival and no effect on abundance or body condition. Environmental protection measures (and tracks) appeared to mitigate the impacts of dense regrowth and bats likely used these features at a local landscape scale. Long-term banding data also revealed negligible effects of weather extremes on survival and we suggest our high elevation study site represented a climate refuge that buffered bats from the effects of weather extremes. No single technique provides all the answers to bat ecology, rather each captures a different piece of the puzzle and, most importantly, each can be applied over long periods of time.

ABSTRACT The conservation of the Critically Endangered Leadbeater's Possum Gymnobelideus leadbeateri in Victoria's Mountain Ash Eucalyptus regnans forests is one of the most controversial native mammal conservation issues in Australia. Much of the controversy results from long-running conflicts between the demands of the native forest logging industry and associated impacts on Leadbeater's Possum and its Mountain Ash forest habitat. Here we argue that despite a legislative obligation to protect Leadbeater's Possum and some limited recent improvements in management, conservation efforts for the species have gone backwards over the past decade. The key problems we identify include that the Victorian Government has: (1) maintained levels of wood production that are too high given the amount of the forest estate that was burned in 2009, (2) failed to substitute clearfell logging practices with more ecologically-sensitive Variable Retention Harvesting Systems, (3) ignored the science (including by its own researchers) on the need for a large protected area for Leadbeater's Possum, (4) altered key definitions such as those for mature trees and old growth that have substantially weakened the ability to protect Leadbeater's Possum, and (5) overlooked the array of forest values beyond timber production (such as water and tourism) and which make a greater contribution to the economy. Our analyses suggest that populations of Leadbeater's Possum are undergoing a substantial decline, as are other hollow-dependent species such as the Greater Glider Petauroides volans . In light of this, it is clear that Victoria needs to substantially change native forest timber harvesting practices as well as improve its efforts to conserve Leadbeater's Possum and the Mountain Ash forests in which it occurs.

ABSTRACT Impacts on the forested bioregions of south–western Australia have, since first European settlement in 1826, been extensive and dramatic. Large–scale land clearing removed over two–thirds of the vegetation for agriculture and urbanisation. Other significant threats to the biota include: changed fire regimes; exotic predators, diseases and herbivores; and drought and climate change. Conservation reserves in the region were originally chosen on aesthetic appeal, often aligning poorly to modern CAR (comprehensive, adequate and representative) criteria aspiring to conservation of at least 10% of each bioregion. The entire period of European occupation of the region has been characterised by the need to balance different land uses within a changing political context, broadening in recent years to incorporate indigenous culture and heritage. Thus much may be gained by a historical perspective grounding contemporary policy in an understanding of the successes and failures of the past. Here, we consider the historical development of the unique perspectives from forestry, agriculture and conservation through the late 19th and early 20th centuries, before assessing their interplay with newer concerns. This experience illustrates that society can design management systems to conserve biodiversity and ensure long–term sustainable use of renewable resources, but that the political will is often lacking.

ABSTRACT There are 5.2 million hectares of public native forest within the coastal Integrated Forestry Operations Approval (IFOA) regions in New South Wales on and east of the Great Dividing Range. 4.3 million hectares or 83% of these forests are set aside in formal or informal conservation reserves. State Forests which comprise 1.55 million hectares or 30% of the public estate within the coastal IFOA regions in New South Wales are managed for a variety of reasons including timber production, recreation, and conservation. While commonly associated with timber harvesting, 43% (675,717 hectares) of the native State forest estate is set aside for conservation in informal reserves. State forest conservation areas include formal Flora Reserves, and a suite of informal reserves including riparian protections, ridge and headwater connection, old growth patches, rare and non– commercial forest types, rainforest, heath, rock outcrops, steep slopes, wildlife corridors, large forest owl protection areas and species specific exclusion zones. These informal reserves receive legal protection via the State Forest Management Zones (FMZ) across the landscape. These informal reserves facilitate the movement of forest fauna and provide important habitat for a unique subset of species. When harvesting occurs in adjoining forest land the reserved areas provide refugia and opportunities from which re–colonisation can occur. Additionally and importantly, the juxtaposition of the undisturbed forest in informal reserves and open areas created by harvest operations creates an ecotone and a variety of successional forest habitats. In addition to describing the current extent of these areas in State Forests in eastern NSW, we reviewed research that assessed their effectiveness as a complementary conservation measure to formal Reserves. Our review indicates that there is a considerable body of knowledge supporting their effectiveness at the local level. To date, however, there has been limited investment in extending and testing these findings more broadly across the landscape. The current forestry regulatory framework is a mix of broad landscape exclusions, general protective conditions implemented routinely regardless of the result of biodiversity surveys and also the option of implementing a condition in lieu of survey, but relies heavily on localised pre– harvesting surveys of threatened species. The costly nature of these surveys and the requirement for these surveys to comply with licence conditions, however, limits the ability to undertake post– harvest surveys or monitoring on a broader scale with available resources. A new model is currently being developed whereby minimum standards are set for informal reserves across the landscape, in conjunction with local scale protection of key habitat features. The new direction has arisen from recognition that effective biodiversity conservation demands a holistic approach. However, the effectiveness of this approach and proposed retention thresholds needs to be rigorously monitored. This would require transferring effort from pre–harvest surveys towards monitoring occupancy trends and status of biodiversity over time.

Nest boxes have been widely used across the world to provide shelter for animal species, often to restore areas following the loss of natural tree hollows. While the microclimates of installed nest boxes have been studied, limited attention has been paid to whether microclimate is influenced by nest box size, shape and entrance dimensions. In this study the temperature and humidity patterns were recorded within six nest box designs that were exposed to direct sunshine. All were constructed of 19 mm plywood but varied in length by a factor of 2X, in volume by 3.5X and in entrance areas by 5X. All next boxes behaved the same thermally, closely following ambient during the night, but during the day they heated to 5 °C warmer than ambient by mid-afternoon. Fluctuations in humidity varied, with small nest boxes with large entrances being closer to ambient humidity than those with small entrances or large volumes. Overall, nest box size and shape had no detectable influence on the internal temperature fluctuations, but did to a slight extent on humidity patterns. Construction material and nest box placement are the likely drivers of the temperature and humidity patterns within nest boxes, and need to be the focus of efforts to keep nest boxes habitable when deployed, and designs should be selected on the basis of the target animal's preferences for size and shape.

Reptile surveys were conducted in the early 1990s in the Dorrigo Forest Management Area, an area formerly comprising timber-production forests managed by the Forestry Commission of New South Wales. The surveys consisted of searches for reptiles along 54 transects together with targeted surveys. Twenty-two 500 m transects were located in previously logged forest and 32 in unlogged forest. Each transect was surveyed twice at an average seven days apart in the spring-summer of 1993 by two persons simultaneously for a duration of 90 min each. A total of 26 reptile species were detected during systematic surveys and 22 reptile species during targeted diurnal searches and opportunistic sightings. The highest species richness and density of reptiles were recorded in the logged New England Hardwood and Gorge Red Gum forest types. The Moist Hardwood type, with a closed mid-canopy had the lowest reptile species richness and density. Neither species richness nor number of individuals varied significantly with logging history. Analysis of abundances of two species of skink that accounted for 88% of detections during transect searches found that the Grass Skink Lampropholis delicata was more abundant in logged sites and during the second survey, whereas the Scute-snouted Calyptotis Calyptotis scutirostrum had higher densities in unlogged sites and in the first count. However, species richness did not vary significantly with logging history, forest type or count. Stephens' Banded Snake Hoplocephalus stephensii , listed under the TSC Act (1995), was observed in the survey area by State Forests' staff. Another 15 species of reptiles recorded during surveys are considered as conservation priority species (Scotts 1996). The Major Skink Bellatorias frerei had not previously been recorded in the survey area and represented a southern range extension. The survey was unable to assess the impacts of logging on threatened species because of low numbers detected. Approaches to improve this situation in future surveys are discussed.

At dusk on the 20 March 2012 bird watcher David Koffel observed a large gliding possum glide between trees at the southern end of Lady Carrington Drive in Royal National Park. David reported his observation to the Royal National Park Visitors Centre and a spotlighting excursion was organised to investigate the sighting. On the 30 March 2012 David Koffel, Debbie Andrew and Glen Harvey located the glider and confirmed its identity as a greater glider Petauroides volans . This was the first confirmed observation of a greater glider in Royal National Park since the wildfire of 1994, 18 years prior, which burnt more than 90% of Royal National Park and 50% of neighbouring Garawarra State Conservation Area. An account of the sightings and background to the observations are provided.

A pilot study, using remotely deployed ultrasonic bat detectors, was undertaken in the Conondale Ranges as part of a research programme to assess the impacts of wet sclerophyll logging on native wildlife. The remote system was of the voice activated type and tested because of its costs relative to other systems. The remote detection technique was inefficient as a means for identifying ail potential species that occur in the area, because of slow response by the equipment to switch on in response to bat calls, problems with high noise to signal ratio and the behaviour of some species relative to the remote equipment. Only eight species were recorded out of a potential eighteen species that occur in the Conondale Ranges. The highest diversity and activity measures were recorded in the intermediate site, last logged in 1961, and the lowest in the regrowth site. These preliminary results are not conclusive because of small sample size and difficulties with the methodology.

Eighty-two species of vertebrates were recorded from an area of cypress pine and ironbark forest within Goonoo State Forest, located 42 Idiometres north-east of Dubbo. New South Wales. Grass-seed-eating birds were uncommon due to the lack of gasses within the forest. Small birds such as fairy-wrens, thornbills and warblers were only found where patches of dense shrub were found, usually where the canopy has been disturbed. Flowering ironbarks were found to attract a variety of bird species utilizing both the pollen and nectar as well as the insects that were attracted. Four threatened species (Glossy Black Cockatoo, Koala. Yellow-bellied Sheathtall Bat. Little Pied Bat) were recorded as well as a new range extension for the Eastern Pygmy Possum. Retention of ironbarks over 47.0 cm diameter at breast height is needed to preserve the presence of large tree hollows as well as eucalypt blossom sources within the forest.

Ten State Forests on the south-west slops 01 New South Wales were surveyed for bats at 39 forest sites and 24 water-bodies using ultra-sonic bat detectors, harp-traps, mist-nets and trip-lines. Thirteen species of bats were recorded for the region. Many species were widespread. with six of the 13 species recorded at more than 50% of sites; however four occurred at less than 20% of sites. Each State Forest sampled supported at least eight species. The species of bats present were representative of the biogeographical region of the study area, lying in the transitional zone between moist climates of the east and arid environments to the west. Within the range of elevations sampled (390-1 419 m). Chalinolobus gouldii, Vespadelus vulturnus, Mormopterus planiceps lp and Scotorepens balstoni were either absent or maintained lower activity levels above 1000 m. Bat activity and species richness was also low in forests typical of the most extreme climate (Snow Gum/Black Sallee), however no differences were found between forests with other floristic associations. Structural complexity appeared more important. as it was related to greater activity levels and higher species richness. Targetting water-bodies increased the efficiency of our survey as the number of bat passes recorded over water-bodies was three times greater than at forest sites. The use of Delay Switches revealed temporal differences in activity for a number of species. it is suggested that bat surveys should attempt to sample the entire night and only Compare activity recorded in the same time period.

This study investigated relationships between durations of inundation in River Red Gum wetlands and waterbird breeding, in particular the durations of flooding under nest trees that waterbirds need to breed successfully; and developed guidelines for managing water regimes in River Red Gum wetlands for waterbird breeding. The main study area was the middle section of the Murrumbidgee River in New South Wales. Average lag times between flooding under nest trees and first nests of Darter, cormorants, herons, egrets, Australian White Ibis and spoonbills ranged from two to four months following spring floods, and from four to seven months following autumn floods. The average lag times from flooding under the nest trees to first broods of ducks were one to two months, regardless of season of flooding. Average lag times to peak breeding of Darter, cormorants, herons, egrets, Australian White Ibis and spoonbills were three to five months following a winter/spring flood and six to nine months following an autumn flood. The equivalent figures for ducks were four months following a winter1 spring flood and five months following an autumn flood. Taking egg laying, incubation of eggs and fledging of young into account, waterbirds need 5 to 10 months inundation under nest trees to complete breeding. Most Darters, cormorants, herons, egrets, Australian White Ibis and spoonbills nested at wetlands which retained water permanently in their deeper, open areas. In contrast, most ducks bred at wetlands which dried out fully between floods. Water levels in River Red Gum wetlands should be managed to provide breeding habitat for waterbirds by allowing the ground under the nest trees to flood for at least 5 to 10 months, and then dry back before the next flood. River Red Gums should not be flooded for longer than 18 months, but water can be retained permanently in deeper, open areas of wetlands.

During March 1997 we surveyed for Myotis macropus at 18 water bodies along the Murray River in River Red Gum Eucalyptus camaldulensis forests in the Millewa and Perricoota/Koondrook State Forests, using Anabat II detectors and harp-traps. Although no M. macropus were caught, they were recorded by Anabat detectors at two sites, one at Moira Lakes Flora Reserve and one on the Murray River at Barmah Bridge. These records were within 7 km of each other and near a previous record on the Murray River (Hoye 1995). Despite the rarity of M. macropus in these forests, high levels of microchiropteran activity from 12 other taxa were recorded. On average, one bat pass was recorded for every minute of sampling in the first hour after dark. Eight taxa ( Vespadeius vulturnus , V. regulus , V. darlingtoni , Chalinolobus gouidii . C. morio , Mormopterus sp. and Tadarida australis ) were recorded at more than 70% of sites, while four species ( Scotorapens balstoni , Nyctophiius gouldi , Nyctophilus geoffroyi and Myotis macropus ) were recorded from less than 40% of sites. River Red Gum forests are particularly important to bats as a mesic habitat in a predominantly semi-arid landscape. Further studies on M. macropus are required to determine the reason for its rarity in an area with abundant surface water.

A regional survey of 253 forest sites on the southwestern slopes and adjacent highlands of New South Wales recorded a total of 530 animals from five species of nocturnal forest birds and nine species of arboreal marsupials. One additional species, the Squirrel Glider Petaurus norfoicensis , was tapped during supplementary searches. Elevation (climate) and best type were the major factors accounting for the distribution of this fauna. Minor environmental gradients included geology, the density of hollow-bearing trees, topography, logging intensity and fire. The patterns of landuse in the region contributed to interactions between elevation, forest type and management history (logging, fire and grazing) which restricted inferences about the habitat requirements of animals. Forest types occurring at either end of the elevational (wet-dry) gradient (Alpine Ash and Box-Stringybark) were poorer in species than forests occurring at intermediate elevations (Alpine Gums, Wet Peppermint and Dry Peppermint). The Wet Peppermint forest type, which was the richest in species of nocturnal animals, has been deliberately selected in the past for conversion to pine plantations. The fauna of the region was generally more sparse and patchy compared to four other regions (northeastern and southeastern New South Wales, and northeastern and Central Highlands of Victoria) where similar studies have been undertaken. Numbers of large forest owls and the Greater Glider Petauroides volans were recorded less commonly than expected and may have declined in the region. The Common Brushtail Possum Trichosurus vulpecula was particularly abundant. This species and the Common Ringtaii Possum Pseudocheirus peregrinus appeared to maintain their numbers in small (<2 000 ha), isolated forest fragments.

We present a detailed case study of conservation and restoration of the Australian arboreal marsupial Leadbeater's Possum (Gymnobelideus leadbeateri) and its Mountain Ash forest habitat to illustrate the important intersection between forest restoration principles and the general principles for forest biodiversity conservation. Mountain Ash forests have been extensively modified through a century of intensive logging, recurrent wildfires and post-fire salvage logging. These disturbances have led to a reduction in old growth forest to 1/30th-1/60th of the extent of historical levels, a rapid collapse followed by a prolonged (≯30-year) shortage of populations of hollow-bearing trees throughout the Mountain Ash forests (which are critical habitat elements for many species of cavity-dependent vertebrates), and an increased risk of re-burning of landscapes dominated by young, regrowth forest. The consequences of the severe decline and consequent ‘temporary extinction’ of large old trees will be the potential global extinction of Leadbeater's Possum whose distribution is significantly associated with the number of large old trees. We outline the conservation and forest restoration principles and practices that are needed to address these problems. We discuss how general principles for forest restoration must be multi-faceted and multi-scaled by encompassing strategies ranging from retaining existing key residual elements of original natural forest cover (e.g. remaining populations of target species, key structures, habitats, and patches) through to restoring patterns of forest cover and key ecosystem processes. We also outline how forest restoration principles intersect strongly with similarly multi-faceted and multiscaled general principles for forest biodiversity conservation - in particular, those corresponding to conserving populations of particular species and their habitats, maintaining stand structural complexity, maintaining patterns of landscape heterogeneity, and perpetuating key ecosystem processes. Finally, we outline the potential for positive cumulative benefits of multiple restoration and conservation strategies by outlining how actions at one scale can create benefits at other (smaller or larger) scales.

Riparian zones are an important habitat for a range of bat species and, as a consequence, understanding whether land use practices such as timber harvesting influence their use is important for conservation and management. This small-scale study used bat activity as a measure of the use of riparian and up-slope zones along headwater streams by bats, and to determine whether past timber harvesting influenced the use of these areas by bats by comparing regrowth with no retained riparian buffers and mature forest. This study found no significant differences in bat activity between treatments, but did find a trend of higher bat activity in riparian zones compared to up-slope zones, particularly in regrowth forest. However, many sites had little to no bat activity indicating activity was influenced by factors not measured in this study. Despite limited data, this study highlights the potential value of retaining riparian habitat in harvested forests for bats, as is practised elsewhere in Australia, and the need for future research into the effectiveness of forest retention measures for wildlife.