Author Identifiers

ORCID: 0000-0003-3766-0849

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


School of Science

First Advisor

Dr Robert Davis

Second Advisor

Dr Eddie van Etten


Woodland restoration is a complex endeavour, and restoration ecology as a scientific discipline requires constant re-assessments and adjustments if it is to improve outcomes and better provide for biodiversity. The promise of effective restoration is often used to justify destructive processes that affect many of the world’s ecosystems. It is therefore imperative that those promises can be met, which comes down to restoration ecologists’ and land managers’ capacity to predict and facilitate desirable ecological changes in a timely and socio-economically responsible manner. As perspectives have changed, and knowledge has been gained over the past few decades there have been several fundamental shifts in how restoration is done. Efforts to ‘beautify’ degraded areas through the planting of fast growing non-native species is no longer thought of as responsible restoration practice. We have a better understanding of ecological thresholds, the creation of novel ecosystems and the ways ecosystems move between stable states through transitional processes. Yet many restoration projects still fail to deliver positive outcomes for certain taxonomic groups. Fauna are an important component of biodiversity, and yet ecological filters and traps remain common in restored habitats.

To date, the focus in restoration has been biased towards restoring flora, while fauna have been under-appreciated and under-utilised. This is likely due to a lack of clarity around how fauna can be used to assess restoration success. This study sought to address that issue by exploring ways fauna could be used to assess habitat quality, and evaluate whether they could fit into existing restoration management tools like a state-and-transition model. Variation in habitat quality was assessed using a number of biodiversity measures and behavioural patterns. This study used Rottnest Island in Western Australia, a mosaic landscape with a woodland restoration program that has been running for over 50 years. The Island’s woodland areas support a resident population of red-capped robins Petroica goodenovii, which was the focal species of this study. The robins are typically groundforaging insectivores that generally have been found to respond negatively to anthropocentric land use changes.

The suitability of the Island’s robin population as an indicator for the larger avian community was assessed to determine whether management and monitoring could simply focus on improving conditions for robins. Unfortunately, robins were found to be a poor indicator of the larger avian community. Factors that were positively correlated with estimated robin density, like woodland area and time since last fire, were negatively correlated with density of other avian species of conservation significance.

Invertebrate assembly was surveyed as a measure of food resource availability. There was a significant difference between woodland and heathland areas and to a lesser, but still significant, extent between restored woodland areas of different ages and remnant woodland. A major finding of this study was that Coleoptera were scarcely encountered in ground samples outside of remnant patches, but were among the most common orders in arboreal samples, specifically in old restoration. Given that this order is a major component of numerous insectivore diets, it is likely that this difference is influencing foraging habitat quality. This conclusion is supported by difference detected in the birds’ foraging behaviour, as birds in remnants foraged predominantly on the ground, while in restored areas birds were frequently observed collecting prey items from vegetation.

Aside from changing their foraging behaviour, the birds were also found to rarely display breeding related behaviours while in restored habitat. This mimicked a significant difference in juvenile robin population density between restored and remnant patches during the breeding season. As such, it appears robins readily use restored areas for feeding resources, but remnants remain a crucial component of their functional habitat requirements, providing important breeding habitat.

Behaviour was found to be a useful tool in explaining and verifying measured differences in habitat quality, and in this case, could easily be incorporated into pre-existing fauna monitoring programs. Robins weren’t found to be a suitable indicator species for the bird community, and given the small species pool on the island, management may need to consider all species of conservation significance separately.


Paper Location