Population trends, habitat requirements and conservation recommendations for an endangered marsupial, the northern bettong (Bettongia tropica)

Whitehead, Tegan Carla (2018) Population trends, habitat requirements and conservation recommendations for an endangered marsupial, the northern bettong (Bettongia tropica). PhD thesis, James Cook University.

PDF (Thesis)
Download (3MB) | Preview
View at Publisher Website: https://doi.org/10.25903/5d0ab9e4fe2eb


This thesis examines the population trends and habitat requirements of the endangered northern bettong (Bettongia tropica) (Wakefield 1967) within wet to moist Eucalyptus woodland in their core population on the Lamb Range in north-eastern Queensland, Australia. This research has implications for managing habitat used by B. tropica and increasing their long-term population viability. Management recommendations to improve the conservation of B. tropica are presented.

Effective management strategies depend upon identifying and mitigating against the key threats to population stability. However, the ability to devise suitable management strategies is often impeded by a lack of data. This is frequently the case for endangered species, including B. tropica. In Chapter two I overcome this problem by using simulation models to make projections of the future impacts on B. tropica under various scenarios. The population viability and survival probability of B. tropica populations on the Lamb Range was modelled in response to 1) increased predation; 2) changes in drought and fire frequency predicted with anthropogenic climate change; and 3) synergistic effects of predation, fire, and drought. Population viability analysis (PVA) models suggest that populations were resilient to substantial declines (up to 75%) and recovered to carrying capacity within 10 years when no threats impacted upon the population.

However, modelling showed that a ≥40% increase in predation by cats, Felis catus, resulted in the population declining to extinction within 20 years. In contrast, populations were resilient to increases in droughts and fires. However, the impacts of predation could be more severe if predation and fire were to interact to increase the mortality of B. tropica. Interestingly, juvenile mortality was the main age class driving population viability, although mortality would need to double from the current rate before extinction was assured. To assist in maximising the long-term viability of B. tropica populations, it is recommended that the density of predators (especially cats and foxes) and B. tropica populations be regularly and consistently monitored. Predator control measures should be undertaken if high densities of predators are detected.

Bettongia tropica is a keystone species within Eucalyptus woodlands on the Lamb Range. The longterm viability of B. tropica is important for maintaining ecosystem function within these woodlands. The population density of B. tropica was previously assessed between 1994 and 1996 within the three main sub-populations (Davies Creek, Emu Creek and Tinaroo Creek) of the core population on the Lamb Range. During that previous study, the majority of sampling occurred at one sub-population, Davies Creek. No consistent monitoring of B. tropica's core population had been undertaken since 1996, although inconsistent monitoring between 2000 and 2009 at Davies Creek suggested a possible decline. In Chapter three, I re-assessed the population density, fitness (survival rates, body condition and females with young) and trap success of B. tropica in the three main sub-populations using a far more intensive trapping regime than used in previous assessments. Population parameters were quantified during both the wet and dry seasons from nine surveys over two years at each subpopulation. Substantially more sampling within each sub-population means that my estimates are more reliable than previous. Population density estimates appeared stable compared to the assessment 20 years prior. Population density was highest at the site with highest rainfall (Tinaroo Creek), with rainfall thought to indicate food abundance. Bettongia tropica sub-populations had similar density estimates and fitness during both seasons. Interestingly, trap success was higher during the dry season. Since long-term monitoring studies often use trap success to assess population trends, an accurate assessment of the population trend of B. tropica requires regular monitoring during both the wet and dry seasons. The population density varied with spatial distribution, with higher population densities occurring within preferred habitats. Population monitoring should therefore be conducted not only within preferred habitat, but also within more marginal habitats.

Knowledge of the spatial distribution of species can provide an insight into the habitat requirements and behaviours of species. This information can assist in devising management strategies to increase long-term habitat stability and thus population viability. In Chapter, four movement patterns, home range distribution and social interactions of B. tropica were investigated using data obtained approximately every 10 minutes from 41 Global Positioning System (GPS) collars. Bettongia tropica had home ranges of 20.90 ± 1.55 ha (mean ± SE), with core foraging and nesting areas of 5.53 ± 0.42 ha and 0.67 ± 0.10 ha respectively. An average of six nesting areas were used over an average of 25.43 ± 1.65 days. Bettongia tropica maintained separate core foraging and nesting areas, despite having largely overlapping home ranges. This suggests they defend areas with high resource density and are somewhat territorial, a trait not previously recorded for this species.

Across all sites, males had larger home ranges than females, with home ranges of both genders increasing during the dry season. Interestingly, home ranges were similar between sites for males and females. The distribution of males appeared influenced by the distribution of females (seeking mating opportunities) and food resources, whilst females were influenced only by the distribution of food resources. Bettongia tropica undertook rapid and direct movements between resource patches and then moved slowly at irregular angles whilst foraging. Fast, linear movements are effective for travelling quickly across areas with minimal resources or few mating opportunities, whilst slower movements maximised the time B. tropica spent within areas with high density of resources. From the movement patterns of B. tropica, the location of bettong nesting and foraging areas were determined.

In Chapter five, the microhabitat requirements of B. tropica were surveyed at nesting and foraging areas. Collared B. tropica were also radio-tracked to their nest location to determine the design and material used to construct nests. Bettongia tropica mainly constructed nests from grass (Poaceae spp.) or nested under the 'skirts' (leaves) of grass trees (Xanthorrhoea johnsonii). Different habitat parameters were important for nesting and foraging. Nests were situated in steep areas with high grass cover and an abundance of grass trees. Whilst foraging, B. tropica selected habitats with a higher density of cockatoo grass (Alloteropsis semialata) and a lower density of tree basal area. Predator pressure appeared to influence habitat selection by B. tropica. Nesting areas were chosen for camouflage while resting, whilst foraging areas were more open to allow rapid escape from predators. Camera trapping conducted for six sessions recorded capture rates of approximately one predator capture to 60 B. tropica and one potential competitor to around 16 B. tropica. The presence of invasive predators of B. tropica on the Lamb Range means it is vital to regularly and consistently monitor both B. tropica and predator populations to assess for changes in density that could impact on the future population viability of B. tropica.

The results of this study provide greater detail on the ecology of B. tropica and will assist in conserving the species. Food density appeared to have the greatest influence on B. tropica population density, which was reflected in how bettongs, especially females, moved throughout their habitat. Tinaroo Creek, which is the wettest site, had the highest population density. At Tinaroo Creek B. tropica had smaller home ranges (although not significantly) and females spent more time foraging (indicated by slow, angular movements) and less time travelling between resource patches (indicated by them undertaking rapid, linear movements). Higher rainfall would lead to higher resource density, enabling bettongs to travel shorter distances to access resources. This would allow more bettongs to occur within a given habitat area, with the habitat thus supporting a higher population density.

Camera trapping data shows that the current predation pressure was slightly higher at Emu Creek, with more camera captures of predator species per captures of B. tropica at the site. Emu Creek was the only site where both cattle and rufous bettongs co-occurred with B. tropica, with these species likely to compete with B. tropica for grass resources. Interestingly, current predation and competition pressure did not appear to significantly influence the fitness of B. tropica, with survival rates, body condition and number of females with young similar between sites. This was surprising since predation pressures appeared to strongly influence microhabitat selection by B. tropica. It is possible that the current predation and competitive pressure was not sufficiently different between sites to detect an influence on population density of B. tropica in this study. However, climate change may increase predation pressure, with PVA modelling showing that predation by feral species could have the greatest impact on the future viability of B. tropica populations. Managing the habitat to minimise the potential impacts of predators is thus of high conservation priority.

Current habitat management practices involve low-intensity mosaic burns undertaken every two to three years on the Lamb Range. The population on the Lamb Range is stable, indicating current fire regimes are generally adequate and not negatively impacting upon B. tropica. However, habitat management could be improved based on the results of this study. Specifically, it is recommended that burns be conducted at a 20 ha scale and at least six areas of approximately 0.70 ha be left unburned to provide sufficient post-fire nesting resources. Management practices should also focus on maintaining or increasing the density of the habitat parameters identified in this study that are important for both nesting (grass cover and grass trees) and foraging (cockatoo grasses and low tree basal area). This may assist in improving habitat quality and increasing the density and viability of B. tropica populations.

It is important that habitat quality be improved throughout the species distribution. The distribution of B. tropica has previously been modelled based on the distribution of their food resources projected from environmental variables. After my study, camera trapping can be used to survey for the presence/absence of B. tropica throughout that modelled distribution. Monitoring within Eucalyptus and wet sclerophyll woodlands on steep slopes and comprising an abundance of grass cover, grass trees and cockatoo grass and low tree basal area will maximise the detectability of B. tropica. Previous researchers have found that vegetation thickening can reduce grass cover. I determined that grass cover and cockatoo grass are important resources for B. tropica. Bettong individuals constructed poorly camouflaged nests within thickets of lantana. These nesting structures were not observed throughout the rest of the habitat, indicating that weedy and thickened vegetation provides substandard habitat for B. tropica. If B. tropica occurs within areas where vegetation thickening is occurring, low-intensity burns should be conducted to reduce thickening and promote a grassy understorey. Low-intensity fire management may assist in improving the habitat quality for B. tropica throughout their distribution.

It is also recommended that the number of known B. tropica populations be increased. Bettongia tropica is now only recorded from two populations (the Lamb Range and Mt. Carbine). The Mt. Carbine population is very small, with little research conducted on that population. As determined from Chapter 2, the Lamb Range population is resilient to a reduction in the number of individuals, making translocations a viable option at this time. Establishing additional populations would increase the population viability of the species and provide a safeguard for the species' survival if the Lamb Range population suffered a large population decline or one or more of the sub-populations went locally extinct. The population viability may also be improved by a better understanding of the ecology and fate of juveniles, as juveniles were the main drivers of the population viability of B. tropica. In Chapter 3, it was shown that the survival rates of adult bettongs were high throughout the year across all sites and females carried pouch young during both the wet and dry season. However, the fate of juveniles and sub-adults was not measured. It is recommended that future research assess the survival rates of sub-adults and juveniles and determine the main factors affecting their survival. This will assist in improving the conservation of B. tropica.

Bettongia tropica provide important ecosystem services within Eucalyptus woodlands, including fungal spore dispersal and possible nutrient recycling. These services can improve habitat quality by improving the growth of certain plant species, which in turn affects vegetation community composition. Eucalyptus woodlands are habitat for a diversity of native species and protecting B. tropica may thus improve the health of an entire ecosystem. This conservation of healthy Eucalyptus woodlands should also assist in maintaining the population viability of other native species within this habitat.

The concepts from this study can also be applied to research into other small mammal species. This study highlights the importance of consistent monitoring during the wet and dry seasons. My research also demonstrates that studying animals' movement patterns can determine their microhabitat requirements. Many small mammal species have cryptic behaviours and their microhabitats are often poorly understood. Previous studies have determined the habitat preferences of small mammals by comparing trap capture rates within different habitats. However, this method may be biased by many factors, including animals being attracted to an area due to baiting of traps or less sampling effort occurring within difficult to access habitats. Using species movement patterns thus provides a more accurate method and is recommended for ascertaining the microhabitat requirements of other small mammal species. This information is crucial for species conservation, as it enables management to focus on protecting important microhabitats.

Item ID: 58748
Item Type: Thesis (PhD)
Keywords: climate change, population viability analysis, pro-active management, survival rates, Bettongia tropica, eucalyptus woodlands, endangered species, bettongs, North Queensland, predation, habitat use
Related URLs:
Copyright Information: Copyright © 2018 Tegan Carla Whitehead
Additional Information:

Publications arising from this thesis are available from the Related URLs field. The publications are:

Chapter 2: Whitehead, Tegan, Vernes, Karl, Goosem, Miriam, and Abell, Sandra E. (2018) Invasive predators represent the greatest extinction threat to the endangered northern bettong (Bettongia tropica). Wildlife Research, 45 (3). pp. 208-219.

Date Deposited: 19 Jun 2019 23:14
FoR Codes: 05 ENVIRONMENTAL SCIENCES > 0502 Environmental Science and Management > 050202 Conservation and Biodiversity @ 50%
05 ENVIRONMENTAL SCIENCES > 0502 Environmental Science and Management > 050211 Wildlife and Habitat Management @ 50%
SEO Codes: 96 ENVIRONMENT > 9608 Flora, Fauna and Biodiversity > 960806 Forest and Woodlands Flora, Fauna and Biodiversity @ 100%
Downloads: Total: 42
Last 12 Months: 34
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page