Predictive distribution models and their application in wildlife conservation

Wildlife managers are often faced with the difficult task of determining the distribution of species, and their preferred habitats, at large spatial scales. This task is even more challenging when the species of concern is in low abundance and/or the terrain is largely inaccessible. Spatially explicit distribution models, derived from multivariate statistical analyses and implemented in a geographic information system (GIS), can be used to predict the distributions of species and their habitats, thus making them a useful conservation tool. We present two such models: one for a dasyurid, the Swamp Antechinus (Antechinus minimus), and the other for a ground-dwelling bird, the Rufous Bristlebird (Dasyornis broadbenti), both of which are rare species occurring in the coastal heathlands of south-western Victoria. Models were generated using generalized linear modelling (GLM) techniques with species presence or absence as the independent variable and a series of landscape variables derived from GIS layers and high-resolution imagery as the predictors. The most parsimonious model, based on the Akaike Information Criterion, for each species then was extrapolated spatially in a GIS. Probability of species presence was used as an index of habitat suitability. Because habitat fragmentation is thought to be one of the major threats to these species, an assessment of the spatial distribution of suitable habitat across the landscape is vital in prescribing management actions to prevent further habitat fragmentation.

Conservation ecology and breeding biology of the white-browed treecreeper, Climacteris affinis

The White-browed Treecreeper Climacteris affinis is one of many woodland-dependent birds that are at risk from the encroachment of human-dominated land-uses into natural landscapes. The White-browed Treecreeper inhabits semi-arid woodlands in north-west Victoria, Australia, a vegetation community that has undergone extreme modification in the last century due to the expansion of agriculture in the region. Extant woodlands represent only 10% of the original woodland cover in the region, and are highly fragmented and disturbed in many districts. Thus, the survival of the White-browed Treecreeper may depend on active management. However, current knowledge of the ecology and biology of this species is virtually non-existent, and inadequate for informed and effective conservation actions. The aim of this thesis is to redress this situation and provide the ecological basis for sound conservation management of the species. The thesis consists of two parts: an investigation of habitat use at three spatial scales and a study of the social organization, nesting requirements, breeding behaviour and reproductive success of a population of White-browed Treecreepers. Fifty-six patches of remnant woodland in north-west Victoria were surveyed to determine the factors affecting the occurrence of the White-browed Treecreeper at the regional scale. It was detected in 16 patches, and was largely confined to two core districts - Yarrara and, Wyperfeld (Pine Plains). The floristic composition of the dominant tree species was an important determinant of patch occupancy, with the results providing quantitative support for the previously suspected affinity for Belah Casuarina pauper and Slender Cypress-pine Callitris gracilis — Buloke Allocasuarina luehmannii woodlands. However, the absence of the White-browed Treecreeper from several districts was due to factors other than a lack of appropriate habitat. Demographic isolation - the distance from the focal patch to the nearest population of the White-browed Treecreeper - was the most important variable in explaining variation in patch occupancy. Patches isolated from other treecreeper populations by more than 8.3 km in landscapes of non-preferred native vegetation, and 3 km in agricultural landscapes, were unlikely to support the White-browed Treecreeper. The impact of habitat loss and fragmentation on the capacity of individuals to move through the landscape (i.e. functional connectivity) is considered in relation to disruption to dispersal and migration, and the potential collapse of local metapopulations. Habitat use was then examined in a network of patches and linear strips of Belah woodland embedded in a predominantly cultivated landscape. A minimum area of 18.5 ha of Belah woodland was identified as the most important criterion for patch occupancy at the local scale. This landscape appeared to be permeable to movement by the White-browed Treecreeper, facilitated by the extensive network of linear habitat, and clusters of small to medium fragments. The third scale of habitat use investigated the frequency of use of 1-ha plots within tracts of occupied woodland. It is important to discriminate between habitat traits that operate at the population level, and those that act as proximate cues for habitat selection by individuals. Woodlands that have high tree density, extensive cover of low-stature shrubs, abundant lichen, a complex vertical structure, and relatively low cover of grass and herbs are likely to support larger populations of the White-browed Treecreeper. However, individuals appeared to be using tree dominance (positive) and tall shrub cover (negative) as proximate environmental stimuli for habitat selectivity. A relatively high cover of ground lichen, which probably reflects a ground layer with low disturbance and high structural complexity, was also a reliable indicator of habitat use. Predictive models were developed which could be used to plan vegetation management to enhance habitat for the White-browed Treecreeper. The results of the regional, landscape and patch-scale investigations emphasise that factors operating at multiple spatial scales influence the suitability of remnant vegetation as habitat for the White-browed Treecreeper. The White-browed Treecreeper is typical of many small Australian passerines in that it has high annual survival, small clutches, a long breeding season, multiple broods and relatively low reproductive rates. Reproductive effort is adjusted through the number of clutches laid rather than clutch size. They occupy relatively large, all-purpose territories throughout the year. However, unlike many group territorial birds, territory size was not related to the number of occupants. The White-browed Treecreeper nests in tree hollows. They select hollows with a southerly orientation where possible, and prefer hollows that were higher from the ground. At Yarrara, there was considerable spatial variation in hollow abundance that, in concert with territorial constraints, restricted the actual availability of hollows to less than the absolute abundance of hollows. Thus, the availability of suitable hollows may limit reproductive productivity in some territories, although the magnitude of this constraint on overall population growth is predicted to be small. However, lack of recruitment of hollow-bearing trees would increase the potential for hollow availability to limit population growth. This prospect is particularly relevant in grazed remnants and those outside the reserve system. Facultative cooperative breeding was confirmed, with groups formed through male philopatry. Consequently, natal dispersal is female-biased, although there was no skew in the sex ratio of the fledglings or the general adult population. Helpers were observed performing all activities associated with parenting except copulation and brooding. Cooperatively breeding groups enjoyed higher fledgling productivity than simple pairs, after statistically accounting for territory and parental quality. However, the difference reflected increased productivity in the 1999-breeding season only, when climatic conditions were more favourable than in 1998. Breeding commenced earlier in 1999, and all breeding units were more likely to attempt a second brood. However, only breeders with helpers were successful in fledging second brood young, and it was this difference that accounted for the overall discrepancy in productivity. The key mechanism for increased success in cooperative groups was a reduction hi the interval between first and second broods, facilitated by compensatory reductions in the level of care to the first brood. Thus, females with helpers probably achieved significant energetic savings during this period, which enabled them to re-lay sooner. Furthermore, they were able to recommence nesting when the fledglings from the first brood were younger because there were more adults to feed the dependent juveniles. The current utility, and possible evolutionary pathways, of cooperative breeding is examined from the perspective of both breeders and helpers. Breeders benefit through enhanced fledgling productivity in good breeding conditions and a reduction in the burden of parental care, which may impart significant energetic savings. Further, breeders may facilitate philopatry as a means for ensuring a minimum level of reproductive success. Helpers benefit through an increase in their inclusive fitness in the absence of opportunities for independent breeding (i.e. ecological constraints) and access to breeding vacancies in the natal or adjacent territories (i.e. benefits of philopatry). However, the majority of breeding unit-years comprised unassisted breeders, which suggests that pairs are selectively favoured under certain environmental or demographic conditions.

Contribution of multi-tenure reserve networks to biodiversity conservation

A consistent and integrated approach to nature conservation across the landscape and regardless of tenure is widely recognised as essential in ensuring the effective conservation of biodiversity. 'Multi-tenure reserve networks', which incorporate public and private lands managed for conservation, are considered a means of achieving landscape scale conservation. Biosphere Reserves (BR) and Conservation Management Networks (CMN) are characteristic models in Australia. This thesis aims to evaluate the role of such networks in protecting biodiversity, specifically by: (1) analysing the spatial configuration (size, shape, connectivity) of networks and their individual components; (2) evaluating the contribution of networks (in real terms and in reporting procedures) to biodiversity conservation objectives; (3) analysing the influence of the attitudes and perceptions of land managers on the functionality of networks; and (4) evaluating the influence of coordinating bodies on network functionality. In order to account for deficiencies in existing classifications of conservation lands, a new classification system was developed for this thesis - the Conservation Lands Classification. This classification incorporates conservation mechanisms on public and private lands and forms the basis for comparing network components in three Australian case studies - the Bookmark BR located in the Murray Mallee of South Australia, the Gippsland Plains CMN on the eastern Gippsland Plains of Victoria and the Grassy Box Woodlands CMN across the inland slopes of New South Wales. The spatial configuration of individual components within networks was measured using spatial analysis techniques within a geographic information system (GIS). GIS was also used to measure the contribution that networks made to a comprehensive, adequate and representative reserve system through the ecosystems they protected. The attitudes and perceptions of landowners and managers within the networks were obtained using questionnaires. Questionnaires were also sent to network coordinators. Statistical and descriptive analysis was conducted on the results. The sizes of individual components varied markedly between the three networks, however within each network public reserves were on average larger than private conservation lands. Although levels of physical connectivity varied between networks, Bookmark BR and Gippsland Plains CMN showed greater similarity to each other than to the Grassy Box Woodlands CMN. The findings raise important questions about the real and perceived differences in the BR and CMN models. All networks, and particularly those components outside the public protected area estate, contributed to enhancing the protection of ecosystems unrepresented or under-represented in the reserve system, although the extent of this contribution varied between networks. Trade-offs between reserve design efficiency and a contribution to a comprehensive, adequate and representative reserve system were evident between networks. Bookmark BR was characterised by high connectivity, strong reserve design integrity but a lower contribution to protecting under-reserved ecosystems, whereas the opposite was evident in the Grassy Box Woodlands CMN. Over 88% of managers considered their involvement in multi-tenure reserve networks to be a positive or very positive experience. A lack of resources and time for management were considered major limitations of these networks. The majority (80%) of private land managers within networks were willing to be included in a national reserve system of conservation lands. This has important implications for the Australian National Reserve System, which currently incorporates mostly public land. The changing nature of the network coordination arrangements suggests an organic fluid evolution of network structures is likely, contrasting with the desire for legalistic and administrative rigidity promoted by government agencies. The thesis concludes that all the networks studied contribute in varying degrees to biodiversity conservation. The key factors influencing the current and potential contribution that such networks make are: (1) the aims, directions and restrictions set by or imposed upon the coordinating body; and (2) the biophysical nature of the surrounding bioregion and resultant historical land use and tenure pattern. Although the successful operation of such 'multi-tenure' networks ultimately relies on the willing participation of private landholders, ongoing institutional support is likely to be required for maintaining networks in the longer term. Considering networks are increasingly formed outside of the influence of government institutions, this presents a significant challenge for effective coordinated conservation.

Public and stakeholder values and knowledge of wildlife in Victoria, Australia

Prompted by a lack of human dimensions research in Australia, this study investigated the values and knowledge relating to wildlife held by members of the public within distinct demographic subsets of the Victorian population and members of wildlife management stakeholder groups; and compared these characteristics with how Victorian wildlife managers perceive these groups. A combination of semi-structured interviews and postal questionnaires were used. Fifteen in-depth interviews were conducted to explore how wildlife managers perceive the values and knowledge of wildlife held by members of various subsets of the Victorian population. A total of 1,431 questionnaires were completed by members of 13 public and stakeholder groups throughout Victoria, and these were analysed to explore values and knowledge relating to wildlife in Victoria. The findings of this study suggest that Victorian people have a strong emotional attachment to individual animals (the humanistic value), and an interest in learning about wildlife (the curiosity/learning/interacting value). The dominionistic/wildlife-consumption, utilitarian-habitat, aesthetic and negativistic values were not expressed by the majority of respondents from the public samples. The data also suggest that Victorian people have relatively low levels of factual knowledge about Australian wildlife. Thus, wildlife managers should expect support for wildlife management objectives that reflect the strong humanistic orientation of Victorians and tailor management and education programs to appeal to this value and Victorians' interest in learning about wildlife. Members of the Field Naturalists Club of Victoria (FNCV), Bird Observers Club of Australia (BOCA), Australian Conservation Foundation (ACF) stakeholder groups and management agency Parks Victoria expressed a strong interest in learning about wildlife. Members of the Royal Society for the Prevention of Cruelty to Animals (RSPCA) obtained high humanistic value scores; and members of the Victorian Field and Game Association (VFGA) obtained high domimomstic/wildlife-consumption value scores. Importantly, the humanistic and curiosity/learning/interacting values were the most strongly expressed values in all six groups and these values could be the key to more effective communication and collaboration between groups. Relationships between demographic factors, and values and knowledge relating to wildlife were found. For example, rural Victorians held a stronger dominionistic/ wildlife-consumption value than urban Victorians; females held stronger humanistic, curiosity/learning/interacting and negativistic values than males; young Victorians (18-34 years) held a lower curiosity/learning/interacting value and lower factual knowledge of wildlife than older Victorians; and more highly educated Victorians were more knowledgeable about wildlife than people with less formal education. No statistically significant differences were found between the values and knowledge of wildlife held by different income classes. While relationships between demographic factors, and values and knowledge relating to wildlife were found, they were generally much smaller than expected based on wildlife managers' perceptions and previous research. For example, the results suggest that Victorian females have a slightly stronger humanistic value of wildlife than males do. However, the important message emerging from the data is that males and females both express a strong emotional attachment to individual animals. Importantly, the results indicate that the effects of demographic factors on values and knowledge relating to wildlife are not always consistent across different geographic locations and stakeholder groups. For example, the slightly stronger interest in learning about wildlife among females when compared with males was observed in the rural and urban-fringe samples but not in the urban samples. This suggests that caution must be used when generalising the findings from human dimensions studies from one type of community or stakeholder group to another. Management programs should be tailored to the specific characteristics of the target audience. The findings also indicate that Victorian wildlife managers have diverse perceptions about the values and knowledge of wildlife held by members of different publics and stakeholder groups, and that the perceptions held by wildlife managers are not always consistent with the actual values and knowledge of wildlife held by members of different publics and stakeholders. For example, counter to the perceptions expressed by the interviewed wildlife managers, the interest in and factual knowledge of wildlife held by members of voluntary conservation groups equalled or surpassed that of wildlife managers; young Victorian adults (18-34 years) held a slightly lower curiosity/learning/interacting value and slightly lower level of factual knowledge of wildlife than older Victorians; and rural and urban communities in Victoria held low dominionistic and utilitarian values. Such discrepancies highlight the importance of investigating the actual values and knowledge held by members of such groups, so that appropriate and effective wildlife management programs can be implemented. Inaccurate perceptions and assumptions may contribute to ineffective communication between managers, stakeholders and publics; and adversely effect the success of wildlife management programs.

Nest-tree selection by the threatened brush-tailed phascogale (phascogale tapoatafa) (marsupialia: dasyuridae) in a highly fragmented agricultural landscape

The conservation of roosting and nesting resources is of critical concern for many hollow-dependent species around the world. We investigated the nest-tree requirements of the threatened brush-tailed phascogale (Phascogale tapoatafa) in a highly cleared agricultural landscape in south-eastern Australia. We documented the physical characteristics of selected nest trees and describe the spatial and temporal patterns of nest-tree use as revealed by radio-tracking. Nine phascogales (seven females, two males) were radio-tracked between March and July 1999 in an area where most woodland habitat is confined to linear strips along roads and streams or small patches and scattered trees in cleared farmland. Female phascogales were monitored for 13–35 days over periods of 5–15 weeks and two males were monitored for 2 and 9 days respectively. A total of 185 nest-tree fixes was collected and all nests occupied by phascogales were in standing trees. Eighty-three nest trees were identified, ranging in diameter at breast height (dbh) from 25 to 171 cm, with a mean dbh for the trees used by each individual phascogale of >80 cm. Phascogales did not discriminate between canopy tree species in selecting nest trees, but showed highly significant selection for trees in the largest size class. All individuals used multiple nest trees, with the seven females occupying an average of 11.4 nest trees from a mean of 25 diurnal locations. The number of nest trees continued to increase throughout the study, suggesting that more would be identified during a longer or more intensive study. Occupied nest trees were located throughout each individual’s home range, highlighting the importance of a continuous spatial distribution of suitable nest trees across the landscape. Nest trees were also located in adjacent farmland up to 225 m from roadside vegetation, demonstrating the value that scattered clumps and even single trees in farmland can have for wildlife conservation.

Succession of small mammal communities after fire and reintroduction of the swamp antechinus 'Antechinus minimus'

This thesis is involved with changes that have occurred to small mammal populations following a major disturbance in the Anglesea region as a result of the 1983 Ash Wednesday fires. Fire, with its effects on spatial and temporal heterogeneity, was found to be an important factor in the maintenance of vegetation and small mammal community structure and diversity in the region. Successional changes in vegetation and small mammal communities were described by multivariate analyses, using data collected annually from 22 study sites. The use of factor analysis techniques, in reducing the annual capture data content, enabled long-term changes in the structure of mammal communities to be interpreted. The small mammal communities in the coastal heath and forest vegetation in the Anglesea region show evidence of a general resilience, (the degree and speed of recovery), to disturbance. Two phases of successional response to fire by mammal species have been proposed; a ‘re-establishment’ phase which occurs in the initial 5-6 years post-fire and is accompanied by rapid increase in species’ abundance, and a subsequent ‘maintenance’ phase accompanied by relatively minor changes in abundance. Habitat Suitability Indices were produced relating to these phases. Vertical density measures of understorey shrubs and herb layers showed significant relationships with small mammal species abundance at the study sites. Long term studies following major disturbances are needed to distinguish between short term recovery of plant and animal species and long term changes in these species. Studies extending over a number of years enable a better directional view of changes in small mammal communities than can be determined from . observations made over a short period. As a part of the investigation into temporal change, it was proposed to undertake trial reintroductions of the Swamp antechinus, Ant echinus minimus, a marsupial dasyurid species which was trapped in the area prior to the 1983 fire, but rarely subsequently. Other more commonly observed native small mammal species (e.g. Rattus fuscipes,R. lutreolus, Antechinus stuartii, Sminthopsis leucopus) had re-invaded the proposed reintroduction site after this fire. Failure of A. minimus to re-establish may have been due to spatial separation of the pre-fire populations coupled with the extensive area burnt in 1983, A source population of the species was located about 100km to the west and habitat utilization and interspecific and niche relationships between the species making the small mammal community explored. Discriminant analysis revealed some spatial separation of species within a habitat based on structural vegetation factors rather than floristic factors. Temporal separation of species was observed, asA. minimus were more active than Rattus species during daylight periods. There was evidence of micro-habitat selection by species, and structural vegetation factors were most commonly identified in statistical analyses as contributing towards selection by small mammal species. Following a theoretical modelling study three reintroduction trials were carried out near Anglesea during 1992-94. Individuals were subsequently radio tracked, and habitat relationships between the species in the small mammal community investigated. Although successful breeding of A, minimus occurred during the latter two trials, the subsequent fate of offspring was not determined. Invasive techniques required to adequately monitor young animals were considered potentially too damaging. Telemetry studies indicated a preference of A. minimus for short, wet heath vegetation. Structural vegetation factors were identified as being significant in discriminating between capture locations of species. Small scale and inexpensive trial reintroductions have yielded valuable additional data on this species and may be viewed as a useful tool in the conservation of other small native mammals.

Population ecology and captive breeding of Pseudomys novaehollandiae at Anglesea

The New Holland Mouse is critically endangered within Victoria. A decline in the species abundance resulted from below average rainfall and structural changes to habitat vegetation. A captive breeding program was established as part of the species recovery plan. Induced breeding throughout the year, under controlled conditions, was limited by the species aggressive behaviour.

Effects of a fox removal program on small and medium-sized mammal population dynamics within the Anglesea Heathlands, Victoria

This thesis assessed the diet of foxes within the ALCOA lease area of Anglesea to determine which native and introduced species foxes were eating. Results showed that Swamp Wallaby was regularly eaten but many other mammals, birds, insects and reptiles were also consumed. The response of native mammals to seasonal fox removal was also determined.

Inferring activity budgets in wild animals to estimate the consequences of disturbances

Activity budgets can provide a direct link to an animal's bioenergetic budget and is thus a valuable unit of measure when assessing human-induced nonlethal effects on wildlife conservation status. However, activity budget inference can be challenging for species that are difficult to observe and require multiple observational variables. Here, we assessed whether whalewatching boat interactions could affect the activity budgets of minke whales (Balaenoptera acutorostrata). We used a stepwise modeling approach to quantitatively record, identify, and assign activity states to continuous behavioral time series data, to estimate activity budgets. First, we used multiple behavioral variables, recorded from continuous visual observations of individual animals, to quantitatively identify and define behavioral types. Activity states were then assigned to each sampling unit, using a combination of hidden and observed states. Three activity states were identified: nonfeeding, foraging, and surface feeding (SF). From the resulting time series of activity states, transition probability matrices were estimated using first-order Markov chains. We then simulated time series of activity states, using Monte Carlo methods based on the transition probability matrices, to obtain activity budgets, accounting for heterogeneity in state duration. Whalewatching interactions reduced the time whales spend foraging and SF, potentially resulting in an overall decrease in energy intake of 42%. This modeling approach thus provides a means to link short-term behavioral changes resulting from human disturbance to potential long-term bioenergetic consequences in animals. It also provides an analytical framework applicable to other species when direct observations of activity states are not possible.

Metrics to assess ecological condition, change, and impacts in sandy beach ecosystems

Complexity is increasingly the hallmark in environmental management practices of sandy shorelines. This arises primarily from meeting growing public demands (e.g., real estate, recreation) whilst reconciling economic demands with expectations of coastal users who have modern conservation ethics. Ideally, shoreline management is underpinned by empirical data, but selecting ecologically-meaningful metrics to accurately measure the condition of systems, and the ecological effects of human activities, is a complex task. Here we construct a framework for metric selection, considering six categories of issues that authorities commonly address: erosion; habitat loss; recreation; fishing; pollution (litter and chemical contaminants); and wildlife conservation. Possible metrics were scored in terms of their ability to reflect environmental change, and against criteria that are widely used for judging the performance of ecological indicators (i.e., sensitivity, practicability, costs, and public appeal). From this analysis, four types of broadly applicable metrics that also performed very well against the indicator criteria emerged: 1.) traits of bird populations and assemblages (e.g., abundance, diversity, distributions, habitat use); 2.) breeding/reproductive performance sensu lato (especially relevant for birds and turtles nesting on beaches and in dunes, but equally applicable to invertebrates and plants); 3.) population parameters and distributions of vertebrates associated primarily with dunes and the supralittoral beach zone (traditionally focused on birds and turtles, but expandable to mammals); 4.) compound measurements of the abundance/cover/biomass of biota (plants, invertebrates, vertebrates) at both the population and assemblage level. Local constraints (i.e., the absence of birds in highly degraded urban settings or lack of dunes on bluff-backed beaches) and particular issues may require alternatives. Metrics - if selected and applied correctly - provide empirical evidence of environmental condition and change, but often do not reflect deeper environmental values per se. Yet, values remain poorly articulated for many beach systems; this calls for a comprehensive identification of environmental values and the development of targeted programs to conserve these values on sandy shorelines globally.

Feeding wild birds in gardens: a test of water versus food

Bird feeding in residential gardens is an increasingly popular human-wildlife interaction. In Australia, the practice is discouraged by most government and nongovernment wildlife conservation agencies, although advice varies and the most common recommendation is to provide water and habitat for birds rather than supplementary food. This study compares bird abundance and diversity when residents in a Melbourne municipality provide water for birds versus food. Bird abundance was greater when food was provided compared with water, but avian assemblages did not differ.