The Assessment of Ecological Condition in South-East Queensland

Home , Abutilon grandifolium, Eucalyptus propinqua, Eucalyptus seeana, Leptospermum speciosum, Lomandra spicata, North Australia

The assessment of ecological condition in South-East Queensland,

Australia: An evaluation of reliability across variable environments

and surrogate efficacy for biodiversity values

David Ian Tucker

B. App. Sc. (Hons)

School of Earth, Environmental and Biological Sciences

Science and Engineering Faculty

Queensland University of Technology

This dissertation is submitted in fulfilment of requirements for the degree of

Doctor of Philosophy

2016

Abstract

As comprehensive biodiversity assessment is often costly and difficult to complete, ecological surrogates are employed to address gaps in taxonomic information. Whereas some surrogates are represented by the presence, absence or abundance of indicator species, others take the form of easily mapped environmental components such as land classes or vegetation types. One form of environmental surrogate involves the measurement of terrestrial ecological condition or integrity. Ecological condition surrogates are often multimetric indices and are represented by easily measured spatial, compositional, structural and functional attributes of native vegetation. Ecological condition is generally assessed by comparing attributes against benchmarks derived from reference sites that typify relatively unmodified examples of the same vegetation type. The utility of ecological condition approaches has seen them adopted by a number of Australian government agencies as part of a broader framework to measure impacts on terrestrial ecosystems, monitor ecological health and calculate vegetation offsets.

The BioCondition framework of Queensland is a typical Australian multimetric condition approach. BioCondition quantifies how well terrestrial ecosystems are functioning for biodiversity values by measuring a number of vegetation and landscape condition attributes. Important management decisions are based on the outcomes of ecological condition metrics and must therefore be repeatable, reliable and scientifically rigorous. Attributes must be applicable across multiple systems and scales and demonstrate significant associations with ecological patterns and processes. Since its inception, the BioCondition framework has been tested for attribute suitability, observer variability and appropriate reference site identification. However, the robustness of BioCondition and other multimetric condition approaches to variability linked to environmental gradients and disturbance regimes has not been assessed. The effects of variability may impair

surrogate efficacy and potentially result in poorly informed conservation policy and planning.

Therefore, this study is the first to consider the reliability of vegetation-based ecological condition assessment across variable environments, and the implications this variability may have for surrogate efficacy. Metric reliability was tested by measuring the influence of climatic, environmental and disturbance variables on selected vegetation communities, and on BioCondition attributes derived from condition assessments within these communities. In addition, the capacity for multimetric condition approaches to represent biodiversity values with respect to this variability was tested through comparative measurements of BioCondition and an independent acoustic index of ecological health.

The study focused on two ecological communities located in South-East Queensland, Australia. The two communities were defined by spotted gum (Corymbia citriodora ssp. variegata) open forest and scribbly gum (Eucalyptus racemosa ssp. racemosa) woodland and categorised under the regional ecosystem vegetation mapping framework as RE 12.11.5e and RE 12.5.3 respectively. A total of 62 sites were surveyed across the South-East Queensland study area, with each site subjected to a full floristic survey, ecological condition assessment, and a measurement of disturbance. In addition, acoustic recordings were taken within ten selected RE 12.11.5e sites and nine selected RE 12.5.3 sites.

Changes in vegetation composition and abundance, and vegetation condition attributes for RE 12.11.5e and RE 12.5.3 were related to latitudinal effects. Where these effects were explicitly tested across two distinct subregions for RE 12.11.5e there was a general trend of decreasing importance for latitudinally-related climatic gradients with progression from tree to ground layers. The influence of non-climatic environmental attributes and disturbance regimes were increasingly important for

shrub and ground strata. Where the same environmental and disturbance variables were examined for RE 12.5.3 it was shown that latitudinal variation was attributed to the geographical disjunction between two vegetation types and geographically- related disturbance processes linked to fire regimes and landscape fragmentation.

It was demonstrated that there was a significant relationship between BioCondition and the Normalised Difference Soundscape Index (NDSI) of ecological health. This relationship indicated that multimetric condition approaches can effectively measure biodiversity values with regard to the effects of environmental variability through comparisons with an independent index of similar purpose. Much of the variation in the soundscape as measured by NDSI was underpinned by landscape attributes, including patch size, context and connectivity. Although NDSI was related to a range of BioCondition vegetation attributes it was determined that their effects were likely connected to fragmentation processes.

The findings of this study have management implications for multimetric condition assessments. The strategic location of reference sites for benchmarking was fundamental to mitigating the effects of climate, environmental gradients and disturbance processes. Landscape-level variability was found to have an inordinate effect on vegetation composition and structure, vegetation condition attributes, bird species richness and composition, and the soundscape as measured by NDSI, suggesting that heavily weighted landscape attributes are important for metric efficacy. In addition, results suggest that multimetric design may be altered through the removal of vegetation species richness estimates, improving usability for non- specialists without compromising surrogacy potential. It was shown that BioCondition scores remained unaffected once adjusted, effectively providing the same measurement of ecological condition. Finally, it was concluded that multimetric condition approaches can be augmented with novel metrics of ecological health derived from acoustic recordings.

This study has shown that multimetric ecological condition indices can account for environmental variability through strategic application and design. Government agencies use ecological condition measurements to inform policy, and therefore it is essential that they operate effectively. If used in conjunction with other assessment methods, and if trialled and tested over multiple ecosystems, environmental gradients and disturbance regimes, then multimetric condition approaches may continue to serve as useful and scientifically rigorous tools for biodiversity management and conservation planning.

Keywords

Australia, benchmarks, biodiversity assessment, disturbance regimes, ecological condition, ecological surrogates, environmental gradients, eucalypt forest, , landscape fragmentation, metric efficacy, multimetric indices, normalised difference soundscape index, reference conditions, soundscape, South-East Queensland, vegetation attributes

Contents Statement of Original Authorship ...... 1

Acknowledgements ...... 2

Chapter 1: General introduction—biodiversity assessment and the utility of ecological condition metrics as ecological surrogates ...... 3

1.1 Introduction ...... 5

1.2 Research problems and project aims ...... 13

Chapter 2: General methods—an introduction to the study area, experimental design and vegetation survey approaches ...... 17

2.1 Broad study area ...... 19

2.2 RE 12.11.5e study areas ...... 21

2.3 RE 12.5.3 study area ...... 23

2.4 Site selection ...... 25

2.5 Ecological condition surveys...... 30

2.6 Vegetation composition and abundance surveys ...... 33

2.7 Disturbance survey ...... 34

2.8 Environmental and land use datasets ...... 36

Chapter 3: The study system—baseline analyses of vegetation patterns across environmental gradients and disturbance regimes in spotted gum open forest (RE 12.11.5e) and scribbly gum woodland (RE 12.5.3) ...... 39

3.1 Introduction ...... 41

3.1.1 Vegetation communities—patterns and processes ...... 41

3.1.2 Disturbance and succession ...... 42

3.1.3 The effects of fragmentation on vegetation communities in South-East Queensland ...... 44

3.1.4 Aims ...... 45

3.2 Materials and methods ...... 47

3.2.1 Field surveys ...... 47

3.2.2 Data analyses ...... 47

3.3 Results ...... 53

3.3.1 Vegetation patterns in spotted gum open forest (RE 12.11.5e) ...... 53

3.3.2 Vegetation patterns in scribbly gum woodland (RE 12.5.3) ...... 64

3.4 Discussion ...... 71

3.4.1 Factors influencing vegetation patterns in spotted gum open forest ...... 71

3.4.2 Factors influencing vegetation patterns in scribbly gum woodland ...... 75

3.5 Conclusion ...... 80

Chapter 4: The effects of environmental variability and disturbance regimes on vegetation condition in spotted gum open forest (RE 12.11.5e) and scribbly gum woodland (RE 12.5.3) ...... 83

4.1 Introduction ...... 85

4.1.1 Ecological surrogacy and condition assessments ...... 85

4.1.2 Aims ...... 87

4.2 Materials and methods ...... 89

4.2.1 Field surveys ...... 89

4.2.2 Data analyses ...... 89

4.3 Results ...... 93

4.3.1 Relationships between vegetation condition attributes, environmental gradients and disturbance regimes in spotted gum open forest (RE 12.11.5e) 93

4.3.2 Relationships between vegetation condition attributes, environmental gradients and disturbance regimes in scribbly gum woodland (RE 12.5.3) ..... 103

4.4 Discussion ...... 110

4.4.1 Effects of environmental gradients on vegetation condition attributes 110

4.4.2 Effects of disturbance regimes on vegetation condition attributes ...... 113

4.5 Conclusion ...... 117

Chapter 5: Testing the efficacy of BioCondition as an ecological surrogate— comparative analysis with a novel, independent acoustic index of ecological health in spotted gum open forest (RE 12.11.5e) and scribbly gum woodland (RE 12.5.3) ...... 119

5.1 Introduction ...... 121

5.1.1 Aims ...... 124

5.2 Materials and Methods ...... 126

5.2.1 Field surveys ...... 126

5.2.2 Data analyses ...... 130

5.3 Results ...... 134

5.3.1 Descriptive statistics ...... 134

5.3.2 NDSI and bird species richness ...... 134

5.3.3 BioCondition and NDSI...... 140

5.3.4 BioCondition vegetation and landscape attributes and NDSI ...... 141

5.4 Discussion ...... 150

5.4.1 NDSI and bird species richness and composition ...... 150

5.4.2 NDSI and BioCondition ...... 153

5.5 Conclusion ...... 157

Chapter 6: General discussion and conclusion ...... 159

6.1 General discussion ...... 161

6.1.1 Vegetation patterns and variability ...... 162

6.1.2 Implications for benchmarking ...... 164

6.1.3 Comparative index measurements ...... 169

6.1.4 Implications for management ...... 170

6.2 Conclusion and further studies ...... 173

References ...... 175

Supplementary Material ...... 227

List of Figures

Chapter 2 Figure 2.1: Bioregions of Queensland, with the South-East Queensland bioregion located at lower right...... 20

Figure 2.2: Map of the current extent of RE 12.11.5e within the South-East Queensland bioregion showing the two subregions around Brisbane and Gympie...... 22

Figure 2.3: Map of the current extent of RE 12.5.3 within the South-East Queensland bioregion...... 24

Figure 2.4: Map of RE 12.11.5e Brisbane survey sites...... 27

Figure 2.5: Map of RE 12.11.5e Gympie survey sites...... 28

Figure 2.6: Map of RE 12.5.3 survey sites...... 29

Chapter 3

Figure 3.1: UPGMA, using Bray-Curtis index, of native species presence/absence for RE 12.11.5e, indicating broad associations...... 55

Figure 3.2: UPGMA, using Bray-Curtis index, of native large tree species abundance for RE 12.11.5e, indicating broad associations...... 57

Figure 3.3: UPGMA, using Bray-Curtis index, of native tree species abundance for RE 12.11.5e, indicating broad associations...... 58

Figure 3.4: UPGMA, using Bray-Curtis index, of native shrub species abundance for RE 12.11.5e, indicating broad associations...... 60

Figure 3.5: UPGMA, using Bray-Curtis index, of native ground species cover abundance for RE 12.11.5e, indicating broad associations...... 62

Figure 3.6: UPGMA, using Bray-Curtis index, of native species presence/absence for RE 12.5.3, indicating broad associations...... 65

Figure 3.7: UPGMA, using Bray-Curtis index, of native large tree species abundance for RE 12.5.3, indicating broad associations...... 67

Figure 3.8: UPGMA, using Bray-Curtis index, of native tree species abundance for RE 12.5.3, indicating broad associations...... 67

Figure 3.9: UPGMA, using Bray-Curtis index, of all native shrub species abundance for RE 12.5.3, indicating broad associations...... 68

Figure 3.10: UPGMA, using Bray-Curtis index, of all native ground species abundance for RE 12.5.3, indicating broad associations...... 69

Figure 3.11: Photographs of RE 12.11.5e site managed with fire BRIS1 (D’Aguilar 1) and fire-excluded dry rainforest BRIS22 (Morrison Road)...... 72

Figure 3.12: Photographs of RE 12.5.3 heathland site 1253_3 (Caves Road) and woodland/open forest site 1253_8 (Sheepstation Creek)...... 76

Figure 3.13: Photographs of two small, fire-excluded RE 12.5.3 patches (1253_15 (Korawal Street) and 1253_17 (Komraus Court)...... 78

Chapter 4

Figure 4.1: Error bar plots of tree-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green)...... 99

Figure 4.2: Error bar plots of shrub-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green)...... 100

Figure 4.3: Error bar plots of weed-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green)...... 101

Figure 4.4: Error bar plots of ground-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green)...... 102

Figure 4.5: Error bar plots for tree-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3...... 107

Figure 4.6: Error bar plots for shrub-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3...... 108

Figure 4.7: Error bar plots for weed- and ground-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3...... 109

Figure 4. 8: Photograph of small, fire-excluded RE 12.5.3 patch (1253_15 Korawal Street) with high plant available water and a canopy dominated by Allocasuarina littoralis and Acacia disparrima...... 111

Figure 4. 9: Photographs of a small RE 12.11.5e Gympie patch (GYM17 Butler Street) with low levels of coarse woody debris and a large RE 12.11.5e Gympie patch (GYM7 Goomboorian) with elevated levels of coarse woody debris due to logging activity...... 116

Chapter 5

Figure 5.1: Map of RE 12.11.5e sites selected for soundscape recordings...... 127

Figure 5.2: Map of RE 12.5.3 sites selected for soundscape recordings...... 128

Figure 5.3: Scatterplot of Mean NDSI (24 hours) against bird species richness for RE 12.11.5e (blue) and RE 12.5.3 (green)...... 136

Figure 5.4: UPGMA of bird species composition for all sites across RE 12.11.5e and RE 12.5.3...... 137

Figure 5.5: Scatterplot of Mean NDSI (24 hours) against BioCondition scores for RE 12.11.5e (blue) and RE 12.5.3 (green)...... 141

Figure 5.6: Scatterplots of Mean NDSI (24 hours) against log10 patch size, patch connectivity, patch context for RE 12.11.5e (blue) and RE 12.5.3 (green)...... 144

Figure 5.7: Scatterplot of Mean NDSI (24 hours) against native shrub cover for RE 12.11.5e (blue) and RE 12.5.3 (green)...... 145

Figure 5.8: Scatterplots of Mean NDSI (24 hours) against coarse woody debris, litter cover and weed cover for RE 12.5.3 (green)...... 146

Figure 5.9: Scatterplot of Mean NDSI (24 hours) against tree species richness for RE 12.11.5e (blue)...... 147

List of Tables

Chapter 2 Table 2.1: Patch size descriptions for site stratification across RE 12.11.5e and RE 12.5.3. Medium patch size values were extended to 400 ha for RE 12.5.3 due to the scarcity of moderately sized forest fragments...... 25

Table 2.2: Patch connectivity descriptions for site stratification across RE 12.11.5e and RE 12.5.3...... 25

Table 2.3: The number of sites with large, medium and small patches separated according to connectivity and regional ecosystem...... 26

Table 2.4: BioCondition vegetation and landscape attributes and their respective weightings as used for all survey sites across the South-East Queensland bioregion. Attributes are described in the text...... 31

Table 2.5: Disturbance survey sheet used for recording disturbance impacts and management activities for RE 12.11.5e and RE 12.5.3...... 35

Table 2.6: Summary of GIS and text based digital datasets, including variable measured, file type, file source, and scale or resolution of the data where applicable...... 37

Chapter 3

Table 3.1: List of potential disturbance variables selected for CCA of vegetation composition and abundance for RE 12.11.5e and RE 12.5.3. Retained variables are indicated by X...... 51

Table 3.2: List of potential environmental variables selected for CCA of vegetation composition and abundance for RE 12.11.5e and RE 12.5.3. Retained variables are indicated by X...... 52

Table 3.3: Mantel tests (r statistics and p-values) of native vegetation composition and abundance, and geographical distance between sites, for RE 12.11.5e. Significant values (p<0.05) are indicated by grey shading...... 63

Table 3.4: Mantel tests (r statistics and p-values) of native vegetation composition and abundance, and geographical distance between sites, for RE 12.5.3. Significant values (p<0.05) are indicated by grey shading...... 70

Chapter 4 Table 4.1: Hierarchical partitioning of all vegetation attributes based on latitude (northing), patch size, disturbance and environmental variables for RE 12.11.5e. Highlighted values are those that significantly contribute to each vegetation attribute based on Monte Carlo permutations. Gradations of shading represent the relative importance of each variable, with darkest indicating the highest importance and lightest the least importance to each attribute...... 94

Table 4.2: ANCOVA results for RE 12.11.5e vegetation attributes and significant variables identified through hierarchical partitioning. Significant values (p<0.05) are indicated by grey shading...... 97

Table 4.3: Linear regression results for RE 12.11.5e vegetation attributes and variables where homogeneity of regression slopes was violated. Significant values (p<0.05) are indicated by grey shading...... 98

Table 4.4: Hierarchical partitioning of all vegetation attributes based on latitude (northing), patch size, disturbance and environmental variables for RE 12.5.3. Highlighted values are those that significantly contribute to each vegetation attribute based on Monte Carlo permutations. Gradations of shading represent the relative importance of each variable, with darkest indicating the highest importance and lightest the least importance to each attribute...... 104

Table 4.5: Pearson’s correlations for normally distributed data and Spearman rank correlations for non-normally distributed data for RE 12.5.3 vegetation attributes and significant variables identified through hierarchical partitioning. Significant values (p<0.05) are indicated by grey shading...... 106

Chapter 5 Table 5.1: Shapiro-Wilk tests for mean NDSI and bird species richness for RE 12.11.5e and RE 12.5.3...... 134

Table 5.2: Levene’s tests for mean NDSI and bird species richness for RE 12.11.5e and RE 12.5.3...... 135

Table 5.3: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of bird species presence/absence for RE 12.11.5e and RE 12.5.3...... 138

Table 5.4: Shapiro-Wilk tests for BioCondition scores for RE 12.11.5e and RE 12.5.3...... 140

Table 5.5: Levene’s test for BioCondition scores for RE 12.11.5e and RE 12.5.3. ... 140

Table 5.6: Hierarchical partitioning of mean NDSI (24 hours) based on BioCondition landscape and vegetation attributes for RE 12.11.5.e...... 142

Table 5.7: Hierarchical partitioning of mean NDSI (24 hours) based on BioCondition landscape and vegetation attributes for RE 12.5.3...... 142

Table 5.8: ANCOVA results for mean NDSI (24 hours) and BioCondition attributes identified through hierarchical partitioning for RE 12.11.5e and RE 12.5.3. Significant values (p<0.05) are indicated by grey shading...... 148

Table 5.9: Linear regression results for mean NDSI (24 hours) and BioCondition attributes identified through hierarchical partitioning where homogeneity of regression slopes was violated for RE 12.11.5e and RE 12.5.3. Significant values (p<0.05) are indicated by grey shading...... 149

List of Supplementary Material

Supplementary Figures

Chapter 2

Supplementary Figure 2.1: BioCondition survey plot layout...... 229

Chapter 3

Supplementary Figure 3.1: CCA of all species presence/absence and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons...... 238

Supplementary Figure 3.2: CCA of all large tree species abundance and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons...... 239

Supplementary Figure 3.3: CCA of all tree species abundance and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons...... 240

Supplementary Figure 3.4: CCA of all shrub species abundance and environmental and disturbance variables across RE 12.11.5e, with broad associations highlighted by black polygons...... 241

Supplementary Figure 3.5: CCA of all ground species cover abundance and environmental and disturbance variables across RE 12.11.5e, with broad associations highlighted by black polygons...... 242

Supplementary Figure 3.6: CCA of all species presence/absence and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons...... 243

Supplementary Figure 3.7: CCA of all large tree species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons...... 244

Supplementary Figure 3.8: CCA of all tree species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons...... 245

Supplementary Figure 3.9: CCA of all shrub species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons...... 246

Chapter 4 Supplementary Figure 4.1: Flowchart for ANCOVA and linear regression process for ecological condition attributes and environmental and disturbance variables for RE 12.11.5e and RE 12.5.3...... 298

Chapter 5

Supplementary Figure 5.1: Flowchart of statistical tests and outcomes for Chapter 5...... 314

Supplementary Tables

Chapter 2 Supplementary Table 2.1: Full site details including site code, site name, regional ecosystem, subregion (RE 12.11.5e), UTM coordinates, patch area, and patch size and connectivity descriptions for RE 12.11.5e and RE 12.5.3. * indicates a reference site used to calculate benchmarks...... 230

Supplementary Table 2.2: BioCondition survey sheet used for recording ecological condition data...... 233

Supplementary Table 2.3: Vegetation composition and abundance survey sheet. 236

Chapter 3 Supplementary Table 3.1: Pearson two-tailed correlations testing for collinearity of disturbance variables for RE 12.11.5e (N=42). Highlighted variables indicate those selected for analyses...... 247

Supplementary Table 3.2: Pearson two-tailed correlations testing for collinearity of disturbance variables for RE 12.5.3 (N=20). Highlighted variables indicate those selected for analyses...... 249

Supplementary Table 3.3: Pearson two-tailed correlations testing for collinearity of environmental variables for RE 12.11.5e (N=42). Highlighted variables indicate those selected for analyses...... 251

Supplementary Table 3.4: Pearson two-tailed correlations testing for collinearity of environmental variables for RE 12.5.3 (N=20). Highlighted variables indicate those selected for analyses...... 253

Supplementary Table 3.5: Botanical species list for RE 12.11.5e. * indicates exotic species...... 254

Supplementary Table 3.6: Botanical species list for RE 12.5.3. * indicates exotic species...... 261

Supplementary Table 3.7: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native species presence/absence for RE 12.11.5e...... 266

Supplementary Table 3.8: CCA eigenvalues and scores for significant axes for RE 12.11.5e species presence/absence data and environmental and disturbance variables. * indicates exotic species...... 268

Supplementary Table 3.9: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native large tree species abundance for RE 12.11.5e...... 271

Supplementary Table 3.10: CCA eigenvalues and scores for significant axes for RE 12.11.5e large tree species abundance data and environmental and disturbance variables...... 273

Supplementary Table 3.11: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native tree species abundance for RE 12.11.5e...... 274

Supplementary Table 3.12: CCA eigenvalues and scores for significant axes for RE 12.11.5e all tree species abundance data and environmental and disturbance variables...... 276

Supplementary Table 3.13: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native shrub species abundance for RE 12.11.5e ...... 277

Supplementary Table 3.14: CCA eigenvalues and scores for significant axes for RE 12.11.5e shrub species abundance data and environmental and disturbance variables. * indicates exotic species...... 279

Supplementary Table 3.15: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native ground species cover abundance for RE 12.11.5e...... 281

Supplementary Table 3.16: CCA eigenvalues and scores for significant axes for RE 12.11.5e ground species cover abundance data and environmental and disturbance variables. * indicates exotic species...... 283

Supplementary Table 3.17: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native species presence/absence for RE 12.5.3...... 285

Supplementary Table 3.18: CCA eigenvalues and scores for significant axes for RE 12.5.3 species presence/absence data and environmental and disturbance variables. * indicates exotic species...... 287

Supplementary Table 3.19: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native large tree species for RE 12.5.3...... 289

Supplementary Table 3.20: CCA eigenvalues and scores for significant axes for RE 12.5.3 large tree species abundance data and environmental and disturbance variables. * indicates exotic species...... 290

Supplementary Table 3.21: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native tree species abundance for RE 12.5.3...... 291

Supplementary Table 3.22: CCA eigenvalues and scores for significant axes for RE 12.5.3 all tree species abundance data and environmental and disturbance variables. * indicates exotic species...... 292

Supplementary Table 3.23: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native shrub species abundance for RE 12.5.3...... 293

Supplementary Table 3.24: CCA eigenvalues and scores for significant axes for RE 12.5.3 shrub species abundance data and environmental and disturbance variables. * indicates exotic species...... 294

Supplementary Table 3.25: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native ground species cover abundance for RE 12.5.3...... 296

Supplementary Table 3.26: CCA eigenvalues and p-values for RE 12.5.3 ground species cover abundance data and environmental and disturbance variables. .. 297

Chapter 4 Supplementary Table 4.1: Descriptive statistics for raw vegetation condition attribute values for RE 12.11.5e Brisbane and Gympie subregions...... 299

Supplementary Table 4.2: Descriptive statistics for raw vegetation condition attribute values for all RE 12.11.5e sites...... 300

Supplementary Table 4.3: Descriptive statistics for vegetation condition attributes for RE 12.5.3 sites...... 301

Supplementary Table 4.4: Results for Shapiro-Wilk tests untransformed vegetation attributes for RE 12.11.5e. Shaded values indicate non-normally distributed data...... 302

Supplementary Table 4.5: Results for Levene’s tests for untransformed vegetation attributes for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance...... 303

Supplementary Table 4.6: Results for Shapiro-Wilk tests for transformed vegetation attributes for RE 12.11.5e. Shaded values indicate non-normally distributed data...... 304

Supplementary Table 4.7: Results for Levene’s tests for transformed vegetation attributes for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance...... 305

Supplementary Table 4.8: Results for Shapiro-Wilk tests for untransformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate non-normally distributed data...... 306

Supplementary Table 4.9: Results for Levene’s tests for untransformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance...... 307

Supplementary Table 4.10: Results for Shapiro-Wilk tests for transformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate non-normally distributed data...... 308

Supplementary Table 4.11: Results for Levene’s tests for transformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance...... 309

Supplementary Table 4.12: Results for Shapiro-Wilk tests for untransformed vegetation attributes for RE 12.5.3. Shaded values indicate non-normally distributed data...... 310

Supplementary Table 4.13: Results for Shapiro-Wilk tests for transformed vegetation attributes for RE 12.5.3. Shaded values indicate non-normally distributed data...... 311

Supplementary Table 4.14: Results for Shapiro-Wilk tests for untransformed environmental and disturbance variables for RE 12.5.3. Shaded values indicate non-normally distributed data...... 312

Supplementary Table 4.15: Results for Shapiro-Wilk tests for transformed environmental and disturbance variables for RE 12.5.3. Shaded values indicate non-normally distributed data...... 313

Chapter 5

Supplementary Table 5.1: A summary of site codes, mean NDSI values, bird species richness and BioCondition attributes for RE 12.11.5e soundscape survey sites...... 315

Supplementary Table 5.2: A summary of site codes, mean NDSI values, bird species richness and BioCondition attributes for RE 12.5.3 soundscape survey sites. ... 316

Supplementary Table 5.3: Table of all bird species including common names, scientific names and site locations for soundscape survey sites in RE 12.11.5e...... 317

Supplementary Table 5.4: Table of all bird species including common names, scientific names and site locations for soundscape survey sites in RE 12.5.3. ... 324

Supplementary Table 5.5: Descriptive statistics for bird species richness and mean NDSI values for soundscape survey sites in RE 12.11.5e and RE 12.5.3...... 331

Supplementary Table 5.6: Hierarchical partitioning of Mean NDSI (24 hours) based on landscape and vegetation BioCondition attributes for RE 12.11.5e. Highlighted values are those that significantly contribute to each vegetation attribute based on Monte Carlo permutations...... 332

Supplementary Table 5.7: Hierarchical partitioning of Mean NDSI (24 hours) based on landscape and vegetation BioCondition attributes for RE 12.5.3. Highlighted values are those that significantly contribute to each vegetation attribute based on Monte Carlo permutations...... 333

Supplementary Table 5.8: Results for Shapiro-Wilk tests for untransformed landscape and vegetation BioCondition attributes for soundscape survey sites for RE 12.11.5e and RE 12.5.3. Shaded values indicate non-normally distributed data...... 334

Supplementary Table 5.9: Results for Levene’s tests for untransformed landscape and vegetation BioCondition attributes for soundscape survey sites for RE 12.11.5e and RE 12.5.3. Shaded values indicate data that violates homogeneity of variance...... 335

Supplementary Table 5.10: Results for Shapiro-Wilk tests for transformed landscape and vegetation BioCondition attributes for soundscape survey sites for RE 12.11.5e and RE 12.5.3. Shaded values indicate non-normally distributed data...... 336

Supplementary Table 5.11: Results for Levene’s tests for transformed landscape and vegetation BioCondition attributes for soundscape survey sites for RE 12.11.5e and RE 12.5.3. Shaded values indicate data that violates homogeneity of variance...... 337

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made.

Signed: QUT Verified Signature

David Ian Tucker

23/03/2016

Acknowledgements

I would like to thank my wonderful and dedicated supervisors, Dr Susan Fuller and Dr Ian Williamson, who without their guidance I would have crawled up under a scribbly gum and had a bit of a rest long ago. I would also like to thank my external advisor and the principal developer of BioCondition, Dr Teresa Eyre. I am forever grateful for your expertise, support and undying enthusiasm. Also, I would like to give many thanks to Professor Stuart Gage and Dr Jason Wimmer, two fantastic guides through the wilderness of bioacoustics and soundscape research. To my partner Clare, and step-daughters Evangeline, Véve and Orlando, you are amazing. Thank you for being a part of my life and putting up with me for the past five years. I would like to acknowledge my mum who, although in the midst of serious illness, was always there with words of encouragement. To my other family and friends, especially those who got roped into long days of field work (Matthew Tucker and Adam Stapleton), thank you. Things can be a lot easier with a helping hand. A special mention should be made for Peter Young; a godsend in the field, a mine of information and a great guy, your assistance was seriously appreciated. I would like to thank all of the staff and students at QUT and at the Institute for Future Environments who have, in some small or large way, helped me for the duration of this PhD. I would also like to acknowledge the Queensland Herbarium, the Queensland National Parks and Wildlife Service, the Sunshine Coast and Brisbane City Councils, and any other body I may have forgotten that has provided permits and resources and facilitated access to sites over the course of this research project.

Chapter 1: General introduction—biodiversity assessment and the

utility of ecological condition metrics as ecological surrogates

1.1 Introduction

The conservation and management of biodiversity is underpinned by how human societies relate to the natural world. This relationship may be defined by the precautionary principle, where the default position is to shield compositional, structural and functional diversity from uncertain anthropogenic disturbance (Hawksworth 2011; Perry 2013). Management decisions are often based on the perceived utilitarian benefits biodiversity provides, with protective measures ensuring the ongoing survival of the human species (Hooper et al. 2005; Alho 2008; Schneiders et al. 2012). Conversely, some authors have argued that plant and animal diversity has intrinsic value in of itself, and that all species have a basic right to exist independent of human needs or desires (Oksanen 1997; Ghilarov 2000; Norton 2000). Although the conservation of biodiversity can be guided by a variety of contrasting philosophical and ethical positions, a practical requirement for all is the demand for detailed ecological data. In many cases, it is imperative that these data are scientifically rigorous, as well as easily collected and understood by non- specialists, in order to make rapid yet informed management decisions regarding the potential impacts on natural systems.

These ecological data are commonly collected through the process of biodiversity assessment, and the monitoring of spatial and temporal changes in the composition, structure and function of ecosystems (Yoccoz et al. 2001; Andersen and Majer 2004; Lindenmayer and Likens 2010; Helson and Williams 2013). Many of these changes are precipitated by human activities, with key threats including habitat fragmentation (Fahrig 2003; Krauss et al. 2010; Bradshaw 2012), land use change (Haines-Young 2009; Trisurat et al. 2010; de Lima et al. 2013), climate change (Lepetz et al. 2009; Yates et al. 2010; Wiens et al. 2011), invasive species (Ferdinands et al. 2005; Powell et al. 2011; O’Donnell et al. 2012), grazing pressures (Cousins et al. 2003; Lunt et al. 2007; Fischer et al. 2010), and altered fire regimes (Gill and Williams 1996; Driscoll et al. 2010; Spies et al. 2012). With threatened ecosystems often susceptible to anthropogenic impacts (Hill et al. 2005; Saunders

5 et al. 2007; Hughes 2011), there is a critical need for biodiversity monitoring and assessment in order to support evidence-based conservation policy for their ongoing management.

As comprehensive biodiversity assessment is often costly and difficult to complete, ecological surrogates are employed to address gaps in taxonomic information. Surrogates serve as a proxy for all, or a significant subset of biodiversity within a defined geographical area (Lindenmayer et al. 2002; Lombard et al. 2003; Harborne et al. 2008; Bergeron et al. 2012). Ecological surrogates can be broadly defined as cross-taxon or environmental surrogates. Cross-taxon approaches measure correlations between surrogate and target species. Surrogate taxa are often defined by guilds rather than single species, familiar biological characteristics, and the ability to be easily sampled or observed (Caro and O’Doherty 1999; Sauberer et al. 2004; Lawler and White 2008; Heino et al. 2009; Di Minin and Moilanen 2014). Environmental surrogacy uses biotic and abiotic data to represent target species or broader biodiversity. Environmental surrogates are based on easily mapped environmental components such as land classes or vegetation types, with the composition and structure of these categories serving as ecological indicators (Ferrier and Watson 1997; Oliver et al. 2004; Sarkar et al. 2005).

Although both methods are commonly used for biodiversity assessment, evidence suggests that environmental surrogates may be more effective than species-based approaches (Oliver et al. 2004; Carmel and Stoller-Cavari 2007; Mandelik et al. 2012; Lindenmayer et al. 2014). This improved surrogacy potential may be due to the significant role vegetation attributes play in the design and implementation of environmental surrogates (Wright et al. 1994; Cowling and Heijnis 2001; Faith et al. 2001; Fraser et al. 2009; Barton et al. 2014). Vegetation is the most apparent and easily described component of terrestrial ecosystems. Plant species composition and structure is pivotal to ecosystem functioning (Guo et al. 2003; Callaghan et al. 2004; Corenblit et al. 2009), and closely linked to a range of environmental

6 gradients, ecological processes and land management practices (Bowman et al. 1991; Ehrenfeld et al. 1997; Neave and Norton 1998; Russell-Smith et al. 1998; Hutley et al. 2011; Shive et al. 2013; Halford and Fensham 2014; Kusumoto et al. 2015).

The capacity for vegetation to capture broad biological variation, combined with the ease and the relatively low financial cost of measurement, has made vegetation-centred environmental surrogates attractive as baseline tools for conservation planning. Despite vegetation extent being well established as an indicator of biological diversity (Ferris and Humphrey 1999; Wilson et al. 2002; Rooney et al. 2012; Polyakov et al. 2013), the use of vegetation attributes for measurements of terrestrial ecological condition are relatively recent developments (Gibbons et al. 2009; Yapp et al. 2010; Eyre et al. 2011b; Pert et al. 2012; Oliver et al. 2014). A combination of historical processes, contextual issues and technical impediments are responsible for their late implementation. Indices of ecological condition have traditionally focused on marine, coastal, estuarine and freshwater ecosystems. Karr (1981), and Kerans, Karr and Ahlstedt (1992) were instrumental in their early conception, designing the benthic index of biotic integrity. Their work has resulted in the generation of a robust body of scientific research centred on the health of aquatic and marine environments (Lopez and Fennessy 2002; Rothrock et al. 2008; Hawkins et al. 2010; Arnaiz et al. 2011), which has consequently influenced similar terrestrial approaches. Although the delay in the development of terrestrial condition metrics may be partly attributed to this historical legacy, the contextual constraints related to subjective variability are significant for the function of ecological condition indices across all ecosystem types.

Ecological condition is a value-laden concept (Gibbons and Freudenberger 2006). Perceptions of quality, viability and health can alter depending on the immediate observer. The problem of contextual interpretation where objective measurement is required may firstly be addressed by defining key ideas. Ecological condition is

7 commonly associated with ecosystem stability, resilience and naturalness. Stable systems may be defined as those that are resistant to environmental impacts and have the resilience to recover following disturbance (Andreasen et al. 2001). Stable systems typically have high natural values which are linked to the composition, structure and function of pre-disturbance states. In Australia and the Americas these natural values are often considered within the framework of European colonisation and its associated ecological consequences (Stephenson et al. 1999; Parkes et al. 2003; Gibbons et al. 2008).

Secondly, subjective variability may also be reduced through the development of rigorous conceptual models which describe ecosystems and provide a formal foundation for ecological condition assessment (Gibbons and Freudenberger 2006; Bleby et al. 2008). Reference models have shown to be particularly effective, enabling ecosystems to be compared across space and time, and ecological condition to be measured relative to reference or pre-disturbance states. In terms of vegetation-based condition indices, reference models typically include a selection of vegetation and landscape attributes, a clearly defined assessment method, and a comparative point of reference (Landsberg and Crowley 2004; Laughlin et al. 2004; Nangula and Oba 2004; Taylor 2004; McElhinny et al. 2005; Gibbons et al. 2008). However, the technical constraints of identifying and integrating each component may have a significant impact on metric efficacy.

Reference model attributes need to reflect the spatial arrangement, composition, structure and function of an ecosystem across a range of spatial and temporal scales. Attributes should ideally be sensitive to environmental variability, easily interpreted by end-users, applicable over multiple systems, stable under changing climatic and seasonal conditions, and cost-effective (Noss 1999; Andreasen et al. 2001; Gibbons and Freudenberger 2006; Bleby et al. 2008). The spatial arrangement of vegetation is often related to landscape fragmentation, where forest is broken up into smaller pieces due to land use change. Fragmentation can cause habitat

8 loss, create new edges, disconnect fragments from adjacent, continuous habitat, and reduce patch sizes (Fahrig 2003; Lafortezza et al. 2010). Although fragmentation effectively creates isolated pockets of vegetation within landscapes, the configuration of remaining patches is also important. Patch context, or the amount, position and connectivity of fragments, can regulate the movement and number of species within a region, affecting broad distribution patterns over time (Opdam 1991; Dunning et al. 1992; Taylor et al. 1993; Hanski 1998; McAlpine et al. 2002).

For reference models and other surrogate frameworks, variations in vegetation species composition and abundance are widely used to represent biodiversity and ecological condition (Parkes et al. 2003; Gibbons et al. 2008; Pomara et al. 2012; Barton et al. 2014; Imai et al. 2014). For example, high native plant species richness often equates with stable, healthy ecosystems (Parkes et al. 2003; Wulf and Kolk 2014), and reduced weed numbers suggest limited anthropogenic disturbance (Abensperg-Traun et al. 1998; Alofs and Fowler 2013). Similarly, vegetation structure can indicate broad ecological change. Intact ecosystems may be represented by high structural complexity (Lindenmayer et al. 2000; Khanaposhtani et al. 2012; Kaufmann et al. 2014), and the increased provision of foraging resources, shelter and breeding sites for fauna (Diaz 2006; Eyre et al. 2011b; Ikin et al. 2014). However, vegetation structural complexity may not always serve as a useful condition surrogate. Infrequent hot fires can result in dense understoreys, reduced amounts of coarse woody debris, increased weed presence and unstable ecosystem dynamics (Gilfedder and Kirkpatrick 1998; Ross et al. 2002). The ecological effects of varied fire regimes on vegetation highlight the complexities of ecosystem functioning. This complexity may be captured through the systematic selection of measured attributes, allowing vegetation to reveal the movement of energy and materials (Diaz et al. 2004), disturbance processes (Attiwill 1994a), and ecosystem change (Potter and Woodall 2012).

Significant spatial, compositional, structural and functional condition attributes are often identified through expert opinion informed by the use of multicriteria analyses and formalised guiding principles (Andreasen et al. 2001; Oliver et al. 2007). Multicriteria analyses employ analytical hierarchies in which expert opinion is used to assess and weigh potential attributes based on their feasibility and importance. Once selected, attributes must then be integrated into practicable measurements of ecological condition. Multimetric indices synthesise individual vegetation and landscape attributes through weighted or non-weighted additive scoring systems, multiplicative approaches, multivariate techniques and ecological modelling (Andreasen et al. 2001; Parkes et al. 2003; Bleby et al. 2008; Gibbons et al. 2009; Eyre et al. 2011b). Each method has its associated strengths and limitations related to tensions between utility and the potential loss of data. To serve as effective indices they must balance readily understood outputs such as scores or ranks, with the meaningful and accurate representation of ecological condition.

Reference-based multimetric condition approaches can serve as ecological surrogates by comparing selected attributes against benchmarks derived from reference sites. However, identifying intact reference sites can be difficult due to temporal and physical limitations (Gibbons et al. 2008). Benchmarks may be quantified by averaging reference values across sites of similar vegetation type and ecological condition, although shaped by different environmental gradients and disturbance regimes. By considering reference conditions as dynamic yet equivalent within a specified range of variation, biological communities may be represented with some degree of accuracy and effectively assessed over time.

The practicality of reference condition models has seen them adopted by a number of Australian government agencies. Ecological condition metrics form part of a broader assessment process in order to quantify impacts on terrestrial ecosystems, monitor ecological health and calculate vegetation offsets (Parkes et al. 2003; Dixon

10 et al. 2006; Gibbons et al. 2008; Eyre et al. 2011b). Prominent state-based examples include BioMetric in New South Wales (Gibbons et al. 2009), Habitat Hectares in Victoria (Parkes et al. 2003), and BioCondition in Queensland (Eyre et al. 2011b). Each multimetric condition index is represented by the spatial, compositional, structural and functional attributes of native vegetation, and combines weighted individual values into broad classes forming a final score for surveyed sites. These approaches aim to reduce subjectivity and standardise measurement by comparing attributes to those of the same vegetation type with little evidence of anthropogenic modification since European settlement. Benchmarks for each attribute are developed using quantitative data and expert knowledge. Condition assessments are conducted at site and landscape scales. Site- scale measurements are rapid, in-situ assessments based on easily recorded biophysical components. Landscape-scale assessments are based on expert knowledge, environmental predictors, or a combination of environmental predictors and remotely sensed data (Gibbons et al. 2006).

Although multimetric condition models are currently used across Australia, criticisms have arisen regarding their design and application. In particular, Habitat Hectares was evaluated by McCarthy et al. (2004), with suggested problems including multiple observer error, limitations of single benchmarks covering a range disturbance regimes, attribute combination problems, and practical concerns when applied to real-world systems. In response, Parkes et al. (2004) argued that the underlying tension between scientific rigour and metric applicability was a significant concern for ecological condition approaches. The authors suggested that the design of Habitat Hectares balanced these two issues without compromising metric integrity. Specifically, it was suggested that assessor variation may be mitigated by techniques such as training strategies and assessment aids. Observer variability was tested by Kelly et al. (2011) under the analogous BioCondition framework in Queensland. It was found that there was low variation in overall condition scores among assessors indicating that ecological condition may be consistently measured despite individual differences. Concerns related to

11 benchmarks and disturbance processes were also addressed by Parkes et al. (2004), who argued that reference models were underpinned by selecting appropriate benchmarks, and that identifying mature dominant growth forms for vegetation communities and defining relatively undisturbed sites were pivotal for metric efficacy. This would suggest that some degree of specialist training may be required for reference site selection. Issues identified by McCarthy et al. (2004) pertaining to problems of an additive attribute scoring system were considered by Parkes et al. (2004) as indicative of the conflict between metric precision and accuracy, and accessibility to non-specialists, with the benefits of using a simple additive approach outweighing any limitations with respect to loss of information. Finally, the practical application of multimetric condition approaches for conservation planning does bring with it the need for clear management objectives. Broadly, Parkes et al. (2004) addressed this concern by suggesting that Habitat Hectares would help identify the extent and range of ecological condition across vegetation communities, and ascertain the rates of change in conservation status of some of these vegetation types.

The BioCondition framework of Queensland is a typical Australian multimetric approach that it is broadly concerned with habitat and ecological integrity rather than individual species or suites of indicator taxa (Oliver et al. 2014). BioCondition quantifies how well terrestrial ecosystems are functioning for biodiversity values (Eyre et al. 2011b), and uses an assessment method that measures a number of vegetation and landscape condition attributes. Vegetation attributes are represented by the number of large trees, tree canopy height, recruitment of canopy species, tree canopy cover, native shrub cover, native plant species richness for four life forms, non-native plant cover, native perennial grass cover, coarse woody debris, and litter cover. Landscape attributes consist of patch size, context and connectivity for fragmented landscapes, or distance to permanent water for intact landscapes. Attributes have been selected because of their surrogacy potential, ease of measurement, and their demonstrated response to ecological change (Eyre et al. 2011b).

BioCondition combines attributes using a weighted, additive approach. The raw data is divided by benchmark values derived from reference sites belonging to specific ecological units known as regional ecosystems (REs). Regional ecosystems are the principal units for the measurement and monitoring of vegetation extent in the state of Queensland, and link this to biodiversity at local, regional and state levels (Sattler and Williams 1999). The regional ecosystem approach is based on land classification and is represented by a biotic (vegetation community) and an abiotic (land zone) component. Each regional ecosystem is defined by a code representing bioregion, land zone and vegetation type. For example, RE 12.5.3 represents the bioregion of South-East Queensland (12), remnant Tertiary surfaces (5), and Eucalyptus racemosa woodland (3). BioCondition data is summed and scored through comparisons to regional ecosystem specific benchmarks and designated attribute weightings (Eyre et al. 2011b).

1.2 Research problems and project aims

A principal focus for biodiversity management in Queensland is the mapping and monitoring of vegetation extent through the regional ecosystem framework (Sattler and Williams 1999), whereas the measurement of vegetation condition reflects the wider historical context, being established much later as a state-based assessment tool. Since its inception, BioCondition has become integral to conservation planning and policy for Queensland. The metric has become particularly important for decisions connected to vegetation offsetting and developmental approvals (Eyre et al. 2011b). Over the past five years BioCondition has undergone comprehensive testing with regard to surrogate efficacy, including an assessment of attribute selection, observer variability (Kelly et al. 2011) and reference site identification (Eyre et al. 2011a). However, no studies to date have evaluated the performance or rigour of the metric and there has been a significant absence of scientific research related to the effects of environmental gradients and disturbance processes on ecological condition attributes. If multimetric condition indices are inordinately

13 affected by variables such as temperature, rainfall, soil properties, fire and fragmentation, they may fail to operate effectively.

Therefore, the goal of this PhD was to assess the robustness of multimetric ecological condition approaches to variability linked to climatic conditions, environmental gradients and disturbance regimes, using the BioCondition framework as a model. This assessment was achieved by addressing three key questions:

1. What are the key relationships between important sources of variability and the vegetation communities selected for the study area? 2. Are ecological condition attributes, derived from surveys within these selected vegetation communities, influenced by the same variability identified for Question One? 3. Do multimetric ecological condition approaches effectively measure biodiversity values with respect to the influence of variability?

An examination of the first question establishes a baseline for the study of condition metric reliability by describing the composition, abundance and structure of native vegetation and its relationship with climatic, environmental and disturbance variables. Furthermore, an assessment of the second question enables identification of the significant sources of variation for condition metrics, indicates whether this variability is shared across vegetation and ecological condition datasets, and suggests how multimetric condition approaches may mitigate the effects of variability. An analysis of the final question demonstrates whether multimetric condition assessments are effective surrogates and robust to the variability identified in the previous two questions.

This study involved the detailed analyses of two regional ecosystems located in South-East Queensland, Australia, classified as RE 12.11.5e and RE 12.5.3. RE

12.11.5e is described as Corymbia citriodora ssp. variegata open forest, usually including Eucalyptus crebra or Eucalyptus siderophloia, Eucalyptus propinqua and Eucalyptus acmenoides or Eucalyptus carnea with a mixed understorey of grasses, shrubs and ferns. This system is associated with Palaeozoic and older metamorphic and interbedded volcanic hills and ranges (Department of Environment and Heritage Protection 2014a). RE 12.11.5e was selected because it has a wide distribution, and it is concentrated within two distinct subregions. This geographical disjunction allowed for the effects of latitudinally-related climatic variability on ecological condition attributes to be examined. RE 12.5.3 is described as Eucalyptus racemosa ssp. racemosa woodland with Corymbia intermedia, Eucalyptus siderophloia +/- Eucalyptus tindaliae, Eucalyptus resinifera, Eucalyptus pilularis, Eucalyptus microcorys and Angophora leiocarpa. Melaleuca quinquenervia is often found on lower slopes. RE 12.5.3 occurs on old loamy and sandy plains defined as remnant Tertiary surfaces +/- Cainozoic and Mesozoic sediments (Department of Environment and Heritage Protection 2014b). RE 12.5.3 was selected because it is compositionally and structurally dissimilar to RE 12.11.5e. This contrast between RE 12.5.3 and RE 12.11.5e allowed for the effects of variability to be examined across two different vegetation communities.

This thesis has been formatted so that each data chapter examines a specific issue related to the principal goal of the PhD. Chapter 2 describes the general methods used in the study, including site selection and survey design. Chapter 3 explores the first key question, and examines the relationships between vegetation composition, abundance and structure, and a suite of climatic, environmental and disturbance variables for each regional ecosystem through the application of multivariate analyses. This process provided detailed vegetation descriptions and insights into the influence of variability for each regional ecosystem. Chapter 4 explores the second key question, and examines the effect of the same climatic, environmental and disturbance variables on BioCondition vegetation attributes through the application of hierarchical partitioning, ANCOVAs and linear regressions. Analyses identified significant sources of variation that may reduce metric efficacy.

Associations between condition attributes, floristic assemblages, climatic variables, environmental gradients and disturbance regimes were considered in order to understand the underlying ecological processes driving variability. Chapter 5 addresses the third key question, and assesses the capacity for multimetric indices to effectively measure ecological condition through a comparison of BioCondition with an independent and novel measure of similar purpose in the form of the Normalised Difference Soundscape Index. In this chapter the relationships between soundscape values, BioCondition scores, vegetation attributes and landscape attributes were examined. Finally, Chapter 6 provides a synthesis of these results, identifies potential modifications to multimetric condition surrogates, and positions this information within the broader context of ecological condition assessment.

Chapter 2: General methods—an introduction to the study area,

experimental design and vegetation survey approaches

This chapter will detail the general methods used throughout the study. The three key questions outlined in the introduction were examined within the broad geographical area described below. All data were collected within this area through multiple survey techniques and desktop datasets.

2.1 Broad study area

The broad study area was defined by the South-East Queensland bioregion, located on the central eastern coast of mainland Australia (Figure 2.1). Bioregions are biogeographic areas based on common climate, geology, landform, native vegetation and species information (Department of Environment 2014). The South- East Queensland bioregion extends from the Queensland-New South Wales border to the south, and terminates at the city of Gladstone in the north. The eastern margin of the bioregion is bound by the Queensland coastline, including coastal islands, and its western boundary broadly follows the Great Dividing Range. It has a moist sub-tropical climate, with seasonal, moderate to high rainfall (800 to 1500 mm per year) mainly falling during the summer months (Sattler and Williams 1999). However, rainfall and temperatures are modified by altitude on the western edge of the bioregion. South-East Queensland has many rare, threatened and endemic species, a number of which reach their northern and southern distributions within this area (Department of Environment 2009).

South-East Queensland has the fastest growing population in Australia (Department of Environment 2009). The region is characterised by many of the impacts associated with anthropogenic disturbance, including increased urban and peri- urban pressures, reduced native forest cover and habitat fragmentation.

Figure 2.1: Bioregions of Queensland, with the South-East Queensland bioregion located at lower right (Department of Environment and Heritage Protection 2015).

2.2 RE 12.11.5e study areas

RE 12.11.5e is a common regional ecosystem in South-East Queensland. It is listed as ‘least concern’ under the Vegetation Management Act 1999 with remnant extent greater than 10000 hectares and greater than 30% of the pre-clearing area. RE 12.11.5e is described as Corymbia citriodora ssp. variegata, or spotted gum open forest, usually including Eucalyptus crebra or Eucalyptus siderophloia, Eucalyptus propinqua and Eucalyptus acmenoides or Eucalyptus carnea with a mixed understorey of grasses, shrubs and ferns. This system is associated with hills and ranges of Palaeozoic and older moderately to strongly deformed and metamorphosed sediments and interbedded volcanics (Department of Environment and Heritage Protection 2014a). RE 12.11.5e was selected for this study because it has a wide distribution with two distinct subregions centred on Brisbane to the south and Gympie to the north (Figure 2.2). This subregional concentration allowed for the effects of latitudinally-related climatic gradients to be tested for vegetation communities and condition attributes. The Brisbane subregion was located between 56J 517000 6918504 and 56J 490609 6979048. This subregion was characterised by urban and peri-urban development. The Gympie subregion was located approximately 140 km north of Brisbane between 56J 471954 7096674 and 56J 472236 7124993. While the Gympie subregion was subjected to urban and peri- urban pressures, it was also impacted by forestry, grazing and rural residential land use.

Figure 2.2: Map of the current extent of RE 12.11.5e (Department of Environment and Heritage Protection 2010) within the South-East Queensland bioregion showing the two subregions around Brisbane and Gympie.

2.3 RE 12.5.3 study area

RE 12.5.3 was selected because its vegetation, soil and geological associations were distinct to that of RE 12.11.5e, allowing for the effects of climatic conditions, environmental gradients and disturbance regimes to be tested across different vegetation communities. RE 12.5.3 is described as Eucalyptus racemosa ssp. racemosa woodland with Corymbia intermedia, Eucalyptus siderophloia +/- Eucalyptus tindaliae, Eucalyptus resinifera, Eucalyptus pilularis, Eucalyptus microcorys and Angophora leiocarpa. Melaleuca quinquenervia is often found on lower slopes. This system occurs on a complex of remnant Tertiary surfaces +/- Cainozoic and Mesozoic sediments (Department of Environment and Heritage Protection 2014b). RE 12.5.3 has been extensively cleared for exotic pine plantation, horticulture and urban development, and is listed as ‘endangered’ under the Vegetation Management Act 1999. All sites were located between 56J 521289 6952426 and 56J 508240 7070170 (Figure 2.3). RE 12.5.3 sites were not divided into separate subregions due to their restricted geographical distribution compared to those of RE 12.11.5e. In addition, remnant patch sizes (small, medium and large) were unevenly distributed across the bioregion due to high levels of fragmentation. However, the effects of latitudinally-related climatic variables were tested because it was determined that the spread of sites across the bioregion was broad enough to discriminate between them.

Figure 2.3: Map of the current extent of RE 12.5.3 (Department of Environment and Heritage Protection 2010) within the South-East Queensland bioregion.

2.4 Site selection

Sites were selected based on measurements of patch size (small, medium and large) and patch connectivity (low and high) (Table 2.1 and Table 2.2) because of their significant associations with species numbers and distributions (Opdam 1991; Dunning et al. 1992; Taylor et al. 1993; Hanski 1998; McAlpine et al. 2002), and their considerable impact on the ecological condition of biological systems (Mitchley and Xofis 2005; Shanley et al. 2013). In addition, patch size and connectivity are readily measured and easily identified in the landscape.

Table 2.1: Patch size descriptions for site stratification across RE 12.11.5e and RE 12.5.3. Medium patch size values were extended to 400 ha for RE 12.5.3 due to the scarcity of moderately sized forest fragments.

Patch Size Description Patch Area (ha) Small 0-25 Medium >25-200* Large >200*

Table 2.2: Patch connectivity descriptions for site stratification across RE 12.11.5e and RE 12.5.3.

Patch Connectivity Description Patch Connectivity Low Connected with adjacent remnant vegetation >10 and <50% of its perimeter OR connected with remnant vegetation <10% of its perimeter AND is connected with adjacent native regrowth >25% of its perimeter. High Connected with remnant vegetation >75% of its perimeter OR includes >500ha of remnant vegetation.

Patch size and connectivity measurements were derived from aerial photographs and satellite imagery (Department of Environment and Heritage Protection 2010b;

ESRI 2013, Google Earth 2013), BioCondition V2.1 (Eyre et al. 2011b) and a GIS based tool developed by the Queensland Herbarium (Kelley and Kelly 2012). It was attempted, where possible, to establish an even number of sites across all patch sizes and connectivities. However, large forest fragments were typically associated with high levels of connectivity, and medium and small fragments with low levels of connectivity, making it difficult to determine the interactive effects of patch size and connectivity. (Table 2.3).

Table 2.3: The number of sites with large, medium and small patches separated according to connectivity and regional ecosystem.

Large patch Medium patch Small patch RE Low High Low High Low High connectivity connectivity connectivity connectivity connectivity connectivity 12.11.5e 0 14 3 1 6 0 Brisbane 12.11.5e 0 11 4 0 3 0 Gympie

12.5.3 1 5 6 2 6 0

Additionally, the number of different patch sizes and connectivities were determined by the specific parameters of each regional ecosystem, including levels of fragmentation, extent of pre-clearing vegetation, topographical associations, and ongoing land use pressures. Sites within these patches were selected based on access, availability, and their capacity to represent each regional ecosystem. As a result, a total of 42 sites were selected across the bioregion for RE 12.11.5e, with 24 associated with the Brisbane subregion (Figure 2.4) and 18 associated with the Gympie subregion (Figure 2.5). A total of 20 sites were selected across the extent of the study area for RE 12.5.3 (Figure 2.6). Full site details including site code, site name, regional ecosystem, subregion (RE 12.11.5e), UTM coordinates, patch size area, and patch size and connectivity descriptions are provided in Supplementary Table 2.1.

Figure 2.4: Map of RE 12.11.5e (Department of Environment and Heritage Protection 2010b) Brisbane survey sites.

Figure 2.5: Map of RE 12.11.5e (Department of Environment and Heritage Protection 2010b) Gympie survey sites.

Figure 2.6: Map of RE 12.5.3 (Department of Environment and Heritage Protection 2010b) survey sites.

2.5 Ecological condition surveys

BioCondition surveys were conducted between August 2011 and March 2012 (Brisbane) and in March 2013 (Gympie) for RE 12.11.5e, and between June 2013 and December 2013 for RE 12.5.3. Surveys were based on guidelines outlined in ‘BioCondition; A Condition Assessment Framework for Terrestrial Biodiversity in Queensland Assessment Manual’ (Eyre et al. 2011b). Each assessment site was delineated by a 100 m by 50 m (0.5 ha) area, within which 10 vegetation attributes and 3 landscape attributes were measured (Table 2.4).

Survey transects were located within remnant vegetation that was deemed representative of the associated forest fragment. This was based on a prior inspection of the pre-clearing and current vegetation mapping for each regional ecosystem, and a visual walkthrough of each patch, or the relevant section of each patch (i.e. target vegetation units). Survey transects were located at least 10 m from an edge so as to minimise edge effects, unless the fragment was too small to contain the 50 m by 10 m plot with a 10 m buffer. Under these circumstances, the plot was positioned in accordance with the spatial constraints of the fragment.

Vegetation Condition Attribute Weighting (%) Number of large trees 15 Tree canopy height 5 Recruitment of canopy species 5 Tree canopy cover (%) 5 Shrub layer cover (%) 5 Coarse woody debris 5 Native plant species richness for four life forms 20 Non-native plant cover 10 Native perennial grass cover (%) 5 Litter cover 5

Landscape Attribute Weighting (%) Size of patch 10 Context 5 Connectivity 5

Each site was marked out by a 100 m centre line following the contour. Site corners were flagged at 25 m either side of the centre line at 0 m and at 100 m, creating a 100 m by 50 m plot. The 10 vegetation attributes were assessed within a number of nested plots within the 100 m by 50 m site. A graphical representation of the BioCondition nested plots is provided in Supplementary Figure 2.1. A description of the survey approach is as follows:

1. 100 m by 50 m area: assessed for the number of large trees, recruitment of canopy species, tree canopy height and native tree species richness. Large trees were described as greater than 30 cm diameter at breast height (DBH) for Eucalyptus, Corymbia, Lophostemon and Angophora species, and greater than 20 cm DBH for all other species. Recruitment was recorded as the percentage of canopy species represented by seedlings in the understorey.

2. 100 m transect: assessment of tree canopy cover and native shrub canopy cover. Canopy and shrub covers were assessed by line intercept method using a 100 m tape extended down the centre line of each 100 m by 50 m plot.

3. 50 m by 10 m subplot, from the 25 m to the 75 m point along the centre line, and encompassing 5 m either side: assessed for non-native plant cover and native plant species richness of shrubs, grass and non-grass ground species.

4. 50 m by 20 m subplot, from the 25 m to the 75 m point along the centre line, and encompassing 10 m either side: assessed for coarse woody debris (CWD). The length of all pieces of CWD within the 50 m by 20 m subplot greater than 10 cm in diameter and greater than 0.5 m in length were measured.

5. Five 1 m by 1m quadrats, starting at the 35 m point and located 10 m apart along the 100 m centre line: assessed for native grass cover and organic litter cover (values are averaged over the five quadrats). Native non-grass ground cover (forbs and other species), non-native ground cover, non-native low shrub cover (less than 1 m in height), rock cover, bare ground cover, and cryptogam cover were also recorded.

Vegetation and landscape attributes were compared against benchmarks for each regional ecosystem. The Brisbane and Gympie subregions had separate benchmarks for RE 12.11.5e. Benchmarks were based on reference site mean attribute values, with reference sites being those considered to be the most intact examples or the ‘best-on-offer’ for each regional ecosystem and subregion. Benchmark calculation allowed for condition scoring across all survey sites. Brisbane RE 12.11.5e reference sites were BRIS1 (D’Aguilar National Park) and BRIS4 (Moggill Conservation Park). Gympie RE 12.11.5e reference sites were GYM2 (Gympie National Park) and GYM7 (Goomboorian State Forest). RE 12.5.3 reference sites were 1253_3 (Caves Road), 1253_4 (Murphys Road) and 1253_5 (Scientific Area 1), with all sites located within large, connected patches associated with the Glasshouse Mountains National Park. An example BioCondition survey sheet is provided in Supplementary Table 2.2.

2.6 Vegetation composition and abundance surveys

A vegetation survey was conducted for each transect based on the guidelines outlined in the ‘Methodology for Survey and Mapping of Regional Ecosystems and Vegetation Communities in Queensland’ (Neldner et al. 2005). Vegetation composition and abundance plots corresponded with the 50 m by 10 m nested plot used in the condition survey (Chapter 2.5), following the contour of each site. Location information, including a locality description, source of location data (GPS), UTM coordinates and the precision of the GPS recording was recorded for each plot. A site description of the dominant vegetation species in the upper and lower strata, broad classes of landform based on the Queensland Herbarium’s floristic records HERBRECS database and landform element, erosion pattern, and landform pattern derived from Speight (1990) was noted for each plot. Slope data, including slope type sourced from Speight (1990), slope angle derived from a laser range finder, and slope aspect were recorded. Site elevation was calculated by GPS and digital mapping. A Specht Structure Code was sourced from Neldner et al. (2005), reflecting the broad vegetation type (e.g. open forest, woodland). The median height, height range, total percentage cover, and key species percentage cover were recorded for each stratum, including trees, shrubs and ground cover.

A complete vascular plant species list by full botanical name was compiled for each site. Those species that could not be identified in the field were collected and identified later in the laboratory. A stem count by strata was carried out for each tree and shrub species. Basal area was recorded by species and stratum using a single, 360° sweep of a Bitterlich stick (Bitterlich 1948). A basal area factor of 1 from the 25 m mark was used, attributing each tree counted with a basal area of 1 m²/ha (Neldner et al. 2005). An example vegetation composition and abundance survey sheet is provided in Supplementary Table 2.3.

2.7 Disturbance survey

A disturbance survey was conducted for each site. Plot boundaries corresponded with the 50 m by 100 m condition survey plot. Surveys were partly based on the disturbance guidelines outlined in the ‘Methodology for Survey and Mapping of Regional Ecosystems and Vegetation Communities in Queensland’ (Neldner et al. 2005). Disturbance variables were storm damage proportion, storm damage age, roadworks proportion, roadworks age, fire proportion, fire age, fire height, grazing extent, logging, ringbarking, erosion extent, weed cover and feral presence/absence, and are listed in Table 2.5. Attributes were scored out of a total of 36 and then standardised to generate a value between 0 and 1.

Table 2.5: Disturbance survey sheet used for recording disturbance impacts and management activities for RE 12.11.5e and RE 12.5.3.

GENERAL INFORMATION Date: Collector: Bioregion: Time: Property: RE: Datum: Zone: AMGE: Site ID: Derivation: AMGN: Accuracy: Transect Bearing: Aspect: Description:

DISTURBANCE Disturbance Proportion Age Score Height Score Prop: 0=0, 1=<1, Type 1 Score 2=1-5, 3=>5% Storm 0 1 2 0 1 2 Age: 0=0, 1=>3, Damage 3 2=>1-2yrs Roadworks 0 1 2 0 1 2 3 Fire 0 1 2 0 1 2 0 1 2 3 Height: 0=0, 1=<1, 3 4 2=1-6, 3=6-12, 4=>12m

Disturbance Disturbance Type 2 Score Grazing 0 1 2 Grazing: 0=nil, 1=minor, 2=moderate, 3 3=severe Logging (# 0 1 2 Logging: 0=nil, 1=1-5, 2=5-10, stumps) 3 3=>10 Ringbarking (# 0 1 2 Ringbarking: 0=nil, 1=1-5, 2=5-10, 3=>10 stags) 3 Erosion 0 1 2 Erosion: 0=nil, 1=minor, 2=moderate, 3 3=severe Weed Cover 0 1 2 3 Cover: 0=0, 1=1-5, 2=5-20, 3=20-50, 4 3=>50% Feral 0 1 Feral Presence: 0=absent, 1=present Presence Total /36 Disturbance: Disturbance Score:

2.8 Environmental and land use datasets

GIS and text-based digital datasets were used to examine the effects of environmental variables, land use and road cover on vegetation composition and structure and vegetation condition attributes. A summary for each data set is provided in Table 2.6. Many of the data were GIS shapefiles or raster files. However, average annual rainfall and rainfall variability were calculated by locating the nearest weather stations within a 15 km radius of each survey point. Monthly rain measurements were recorded over a significantly representative time-scale, typically a minimum of 80 years, up until 2012. The closest weather stations were selected first, with progressively more distant stations being included in each dataset until ≥ 80 years had been accounted for. Those years with greater than three months of missing data were excluded. Those years that had three or fewer months of missing data had those measurements replaced by the mean monthly value. Annual average rainfall was based on the mean of the yearly average for each survey point. Rainfall variability was determined by calculating the coefficient of variation based on the average annual rainfall and rainfall standard deviation for each survey point.

Table 2.6: Summary of GIS and text based digital datasets, including variable measured, file type, file source, and scale or resolution of the data where applicable.

Variable File type Source Scale/resolution (if applicable)

Pre-clearing vegetation extent Shapefile feature class (ESRI GIS Department of Environment and 1:100000

data) Heritage Protection, Qld

Government (2010a)

Current vegetation extent Shapefile feature class (ESRI GIS Department of Environment and 1:100000

data) Heritage Protection, Qld

Government (2010b)

Average annual temperature File system raster (ESRI GIS data) WorldClim v1.4 (2004a) Spatial resolution: 1 km²

Temperature variability File system raster (ESRI GIS data) WorldClim v1.4 (2004b) Spatial resolution: 1 km²

Average annual rainfall Text file (Excel) Bureau of Meteorology, Australian N/A

Government (2013)

Rainfall variability Text file (Excel) Bureau of Meteorology, Australian N/A

Government (2013)

Variable File type Source Scale/resolution (if applicable)

Potential evapotranspiration File system raster (ESRI GIS data) CGIAR Consortium for Spatial Spatial resolution: 1 km²

Information (2009)

Soil description File system raster (ESRI GIS data) CSIRO Land & Water (2011a) 1:2000000

Plant available water File system raster (ESRI GIS data) CSIRO Land & Water (2011b) 1:100000

Bulk density File system raster (ESRI GIS data) CSIRO Land & Water (2011c) 1:100000

Distance from coast Calculated using ArcGIS proximity ArcGIS v10.1 (ESRI 2013) 1:100000

tool

Land use cover Shapefile feature class (ESRI GIS Department of Environment and 1:100000

data) Heritage Protection, Qld

Government (2012a)

Road cover Shapefile feature class (ESRI GIS Department of Environment and 1:100000

data) Heritage Protection, Qld

Government (2012b)

Chapter 3: The study system—baseline analyses of vegetation

patterns across environmental gradients and disturbance regimes in spotted gum open forest (RE 12.11.5e) and scribbly gum woodland (RE

12.5.3)

3.1 Introduction

3.1.1 Vegetation communities—patterns and processes

Our understanding of vegetation patterns has been shaped by over a century of scientific debate, stretching back to the inception of ecology as a formal discipline. For the first half of the twentieth century that debate was largely shaped by the observations of Frederic Clements (1916) and Henry Gleason (1926). Clements proposed that vegetation communities may be described as distinct, changeable entities with a final climax state; whereas Gleason perceived communities as existing on a continuum, with boundaries shaped by the tolerances of individual species to environmental gradients, stochastic events and species-species interactions.

Empirical research during the 1940s and 1950s (Whittaker 1956; Curtis 1959) found that vegetation patterns were closely aligned to environmental gradients, suggesting that the continuum model was largely correct. However, although plant species were seen to be distributed individually, distinct changes in vegetation composition and structure were still observed. Clusters were found to be associated with ecotones represented by abrupt abiotic boundaries, and with ecoclines described by incremental environmental gradients (Van Leeuwen 1966; Van der Maarel 1990). These associations suggested that the concepts were not mutually exclusive, and that conflict was a product of two contradictory frames of reference; the continuum described as a theoretical abstraction, and communities as a function of the landscape (Austin and Smith 1989). Current interpretations favour an integration between Clementsian and Gleasonian models (Van der Maarel 2005), linking the continuum concept to an environmental space defined by landscape- specific gradients (Austin and Smith 1989).

This integration has been facilitated by the perception of vegetation distribution as a product of both pattern and process. Pattern considers how species and communities are distributed over the landscape, and process is concerned with what processes (e.g. competition, herbivory, history) function in biological communities and how they shape patterns (Anand 2000; Pickett et al. 2009). Process, in particular, has allowed for the conception of vegetation communities as, if not distinct then, uniform entities with a similar appearance and physiognomy over prescribed areas (Van der Maarel 2005). Communities are components of continuous gradients, but also show some directionality and predictability. However, this does not necessarily infer that vegetation communities develop towards a stable climax state. Communities are strongly influenced by environmental changes over time, are dynamic, and are affected by disturbance (Kent et al. 1997; Anand 2000).

3.1.2 Disturbance and succession

Vegetation communities are shaped by a range of factors, including human-induced disturbance, natural variation, species interactions and stochastic events. It has been recognised that, as a consequence, successional change may not necessarily proceed to a single Clementsian climax state. Vegetation communities are considered to be dynamic, and various theoretical and practical models have been developed in parallel to explain observed patterns and processes (Grime 1973; Connell and Slatyer 1977; Austin and Smith 1989; Pickett et al. 2009). Typically these models conceptualise change and its effects on vegetation composition, structure and function as multidimensional, reflecting a complexity at odds with simple unilinear explanations. State and transition models provide one such framework indicating how persistent vegetation communities can exist in a range of alternative stable states (Westoby et al. 1989; Stringham et al. 2003; Briske et al. 2005). Transitions, or the trajectories between states, are often caused by multiple disturbances including natural events (e.g. climatic events or fire) and management activities (grazing, farming, burning etc.).

It has been suggested that state and transition models are complementary to earlier Clementsian succession approaches, and therefore should be used as widely applicable, adaptive tools for understanding and managing ecosystems rather than an alternative theory of vegetation dynamics (Herrick et al. 2006; Quétier et al. 2007). The strength of state and transition models is derived from their applicability to a range of community processes, their ability to incorporate ecosystem parameters other than plant species composition (e.g. soil characteristics, climate variation), their lack of scale-dependence and their inclusion of site-time interactions that can influence community responses to environmental conditions or managerial effects (Jackson and Bartolome 2002).

State and transition approaches to vegetation change also address the ability of plant species to respond to disturbance. This response is often defined by the capacity for ecosystems to preserve the same function and structure during disturbance and subsequently recover (Westman 1978; Walker and Salt 2006). Stable ecosystems are resistant to change and resilient, while allowing for changes in plant composition over time (Stringham et al. 2003). Within state and transition models, ecological resistance and resilience are tied to the concept of thresholds, or points in space and time at which one or more ecological processes responsible for maintaining equilibrium degrade beyond the point of self-repair. Processes need to be restored before returning to the previous state, otherwise a new state supporting a different suite of plant communities and a new threshold is formed (Stringham et al. 2003). Thresholds and state transitions often indicate that a system has undergone change beyond its capacity to resist or accommodate disturbance, and may be driven by modifications in biotic structure and community interactions (e.g. altered population and competitive dynamics) (Briske et al. 2005). Additionally, transitions between stable states may occur in response to long-term abiotic changes that modify site characteristics (e.g. climate change, severe soil erosion, water table variation) (Briske et al. 2005).

Although patterns and processes provide a framework for conceptualising vegetation communities, they do not address the fundamental question of whether plant communities exist as defined entities. It has been demonstrated that patterns in plant composition are driven by abrupt changes in the physical environment (Gleason 1926; Baker and Weisberg 1995; Kent et al. 1997; Walker et al. 2003), and may therefore negate the very concept of community. However, if considered in the context of processes that significantly structure ecosystems, a community may be regarded as an integrated entity with unique, emergent properties (Grime 1973; Margules et al. 1987; Kent 1997; Jiang et al. 2012). Comprehending the landscape as an assimilation of uniform entities and underlying ecological processes is important for this study. Regional ecosystems are an amalgam of vegetation cover, soil, geology and landform; suggesting that there are readily identifiable floristic associations, or communities, organised along a range of environmental gradients. In addition, regional ecosystems may be subjected to a range of disturbance processes associated with natural phenomena or human activity.

3.1.3 The effects of fragmentation on vegetation communities in South-East

Queensland

Models of vegetation change provide some insight into the role and effects of human induced disturbance across natural systems. The South-East Queensland bioregion has been altered by thousands of years of indigenous land management, and approximately two hundred years of European settlement. The impacts have been profound, with pastoral and agricultural development driving early European clearing (McAlpine et al. 2002) and more recently, growing urban populations contributing to further landscape change. From 2008 to 2009, 1544 hectares of native vegetation were cleared from the South-East Queensland Natural Resource Management region, including the Gold Coast, Brisbane, Caboolture and Sunshine Coast areas, for the expansion of urban development (Field et al. 2012). These processes have culminated in the fragmentation of forest, habitat loss, the creation of new edges, and a disconnection of ever-smaller fragments from adjacent,

44 continuous habitat (Fahrig 2003; Lafortezza et al. 2010). The effect on native species cannot be overstated. The disruption of spatially structured populations, increasingly separated by space, and connected by more tenuous dispersal pathways, results in shifting distribution patterns with reduced diversity and local extinctions (Opdam 1991; Hanski 1998; McAlpine et al. 2002).

3.1.4 Aims

To gain an understanding of the effects of fine-scale environmental gradients, local patterns of disturbance, and climatic variability on vegetation communities in South-East Queensland, two regional ecosystems (RE 12.11.5e spotted gum open forest and RE 12.5.3 scribbly gum woodland) were studied. These communities were chosen because of their close geographical proximity within the South-East Queensland bioregion, the geographical disjunction between RE 12.11.5e subregions for testing latitudinally-related climatic effects, and their unique biological assemblages, edaphic and geological characteristics, and management histories. Both regionals ecosystems have been shaped by the interaction between natural ecological processes, indigenous management practices and the impacts of European colonisation.

The overall aim of this chapter was to undertake a baseline analysis of the vegetation patterns of RE 12.11.5e and RE 12.5.3 across environmental gradients and disturbance regimes. In order to resolve fine-scale patterns, a comprehensive analysis of the composition, abundance and physiognomy of vegetation was undertaken. This required full structural partitioning of the vegetation communities. This approach accounts for the fact that species diversity has been shown to vary independently between strata, and enables a better determination of the underlying ecological processes responsible for the observed vegetation patterns (Peet 1978).

The following questions were examined:

1. What factors influence the compositional components and species abundance within each structural layer in the two vegetation communities, and 2. What commonalities and differences exist across the two vegetation communities?

It was important to develop this detailed baseline understanding because the composition and structure of native and introduced vegetation underpin ecological condition attributes. While the current extent, composition and broad abiotic associations (e.g. soil, geology) of RE 12.11.5e and RE 12.5.3 are well understood, as are the predominant anthropogenic impacts on each system, fine-scale patterns have not been previously examined.

3.2 Materials and methods

3.2.1 Field surveys

Two regional ecosystems were examined in this study. For RE 12.11.5e, 42 sites were selected, and for RE 12.5.3, 20 sites were selected as per methods detailed in Chapter 2.4. RE 12.11.5e has been heavily impacted by disturbance associated with European colonisation, although it still retains greater than 30 percent pre-clearing extent. Existing remnants centre on the Brisbane subregion to the south and Gympie to the north, although these roughly represent the pre-clearing cover of this unit. This regional ecosystem has undergone extensive fragmentation, particularly in the heavily populated Brisbane area (Bradshaw 2012), and has been subjected to persistent logging pressures in the Gympie subregion. RE 12.5.3 has a relatively limited extent, but within its restricted distribution it is structurally and compositionally diverse. This regional ecosystem has been extensively cleared for human habitation, agriculture and pine plantation forestry, with less than 10 percent of pre-clearing extent remaining. What remains is typically found in reserves and is highly fragmented. Field surveys were conducted as per methods detailed in Chapters 2.5 (DBH measurements) and 2.6.

3.2.2 Data analyses

3.2.2.1 UPGMA numerical analyses

UPGMA numerical analyses were performed on presence/absence and abundance data for RE 12.11.5e and RE 12.5.3 sites using PATN version 3 for Windows (Belbin 2008). Numerical analyses were conducted for (1) presence/absence data for native species across all sites; (2) abundance data, based on DBH measurements of native large trees, across all sites; (3) abundance data, based on stem counts of native tree species (T1 and T2 layers), across all sites; (4) abundance data, based on stem counts of native shrub species (S1 and S2 layers), across all sites; and (5) abundance data, based on percentage native ground cover measurements, across all sites. Non-native species were removed from the analyses because floristic associations

47 between native species were considered diagnostic of the biogeographical and environmental relationships for RE 12.11.5e and RE 12.5.3, and non-native species are not a component of the reference ecological condition state upon which regional ecosystems are based. Vegetation layers were analysed separately.

Raw presence/absence data remained untransformed prior to analyses. Large tree abundance data, tree abundance data and shrub abundance data were square root transformed prior to analyses. Percentage ground cover data were arcsine transformed prior to analyses. The Bray-Curtis association measure (Bray and Curtis 1957) was used to quantify the floristic dissimilarity between sites based on these representative datasets. A two-step function was used to explore the relationship between plant species and sites. Data classification was carried out using agglomerative hierarchical classification and flexible unweighted pair group arithmetic averaging (UPGMA). A beta value of -0.1 was used, slightly dilating the ‘ecological space' defined by the study area’s species (Belbin 2004). Principal component correlation (PCC) performed multiple linear regressions on selected variables, and was used to calculate the correlation between species and sites. A Monte-Carlo permutation test was used to test the ‘robustness’ of the PCC (Belbin 2004). The species lists generated by Monte-Carlo attributes in an ordination (MCAO) were examined for those species that were statistically significant (MCAO values ≤1%). The statistically significant species, and their relationships with survey sites and regional ecosystems, were analysed through the interpretation of two- way tables.

3.2.2.2 Canonical correspondence analysis

A suite of disturbance variables were selected in order to test their effects on the vegetation composition, abundance and structure of RE 12.11.5e and RE 12.5.3. Variables for each regional ecosystem were subjected to Pearson two-tailed correlations to test for collinearity. Retained variables either demonstrated a lack of

48 collinearity or, if correlated, were considered potentially significant predictors relative to the associated correlated variables, which were subsequently removed from further analyses (Table 3.1). A similar process was applied to a suite of environmental variables, with remaining attributes reflecting either their lack of collinearity or relative importance (Table 3.2). All correlations are provided in Supplementary Table 3.1 and Supplementary Table 3.2 for RE 12.11.5e and RE 12.5.3 disturbance variables and Supplementary Table 3.3 and Supplementary Table 3.4 for RE 12.11.5e and RE 12.5.3 environmental variables.

A Canonical correspondence analysis (CCA) approach was used similar to that implemented by Holdaway et al. (2011) to explore vegetation patterns and relationships. In the current study, CCA using PAST3 (Hammer 2013) was applied to untransformed presence/absence data. All species, including non-native species, were included in the analysis. RE 12.11.5e sites, presence/absence data, and a suite of 15 disturbance and environmental variables identified through Pearson two- tailed correlations (Table 3.1 and Table 3.2) were used to examine the relationships between species and selected variables. Prior to analysis all singleton species were removed from the dataset. Using Legendre and Legendre’s (1998) eigenanalysis algorithm, ordinations were presented as site scores, and environmental variables were plotted as correlations with site scores (Hammer 2013). Axes were inspected for their contribution to overall variance, and tested for significance using Monte Carlo permutations (999). Significant axes and associated ordination plots were examined in order to determine the relationships between species, environmental gradients and disturbance regimes. The procedure was repeated for RE 12.11.5e square root transformed large tree abundance data, square root transformed tree abundance data, square root transformed shrub abundance data and arcsine transformed percentage ground cover data. The procedure was also applied to all RE 12.5.3 datasets based on 12 disturbance and environmental variables identified through Pearson two-tailed correlations (Table 3.1 and Table 3.2).

3.2.2.3 Mantel tests

Mantel tests were conducted for geographical distance versus native species presence/absence data, large native tree abundance, all native tree abundance, native shrub abundance and native ground cover for RE 12.11.5e and RE 12.5.3 using R studio (R Core Team 2013) and R package “ade4” (Dray et al. 2014). Association matrices for vegetation datasets were generated through previous numerical analyses using PATN (Belbin 2004). The association matrix for geographical distance was based on UTM coordinates for each site, and was generated using “ade4”.

Table 3.1: List of potential disturbance variables selected for CCA of vegetation composition and abundance for RE 12.11.5e and RE 12.5.3. Retained variables are indicated by X.

Variable Variable code RE 12.11.5e RE 12.5.3

Patch size (ha) PS X X

Roadworks proportion (%) RP

Roadworks age (years) RA

Fire proportion (%) FP X

Fire age (years) FA

Fire height (m) FH X

Logging (# stumps classes) L X X

Conservation cover (m²) C

Forestry cover (m²) F

Residential cover (m²) R

Rural residential cover (m²) RR X

Grazing natural environment cover (m²) GNE X

Horticulture cover (m²) H

Main road cover (m) MR

Local road cover (m) LR X X

Vehicle track cover (m) VT

Table 3.2: List of potential environmental variables selected for CCA of vegetation composition and abundance for RE 12.11.5e and RE 12.5.3. Retained variables are indicated by X.

Variable Variable Code RE 12.11.5e RE 12.5.3

Latitude (UTM northing) N X X

Slope (°) S

Aspect (cosine transformed) ASP X X

Altitude (m) ALT X X

Distance from coast (km) DC

Average annual rainfall (mm) AAR X

Rainfall variability (C.V.) RCV X X

Potential evapotranspiration PET (mm)

Average annual temperature (°C) AAT X

Temperature variability (C.V.) TCV X X

Bulk density 0-30 cm (mg/m³) BD X X

Plant available water 0-1 m (mm) PAW X X

3.3 Results

A total of 298 plant species were identified for RE 12.11.5e, of which 78 were trees, 102 were shrubs, 132 were forbs and 50 were graminoids. A total of 197 plant species were identified for RE 12.5.3, of which 44 were trees, 80 were shrubs, 86 were forbs and 38 were graminoids. Multiple species were represented by both tree and shrub growth forms. Species lists for RE 12.11.5e and RE 12.5.3 are provided in Supplementary Table 3.5 and Supplementary Table 3.6.

Overall, the main factor influencing vegetation patterns in both vegetation communities was latitude. The UPGMA analyses provided a clear visual representation of the relationship of sites based on geographical location and consequently, in the following results section only dendrograms have been provided. The CCA analyses provided valuable insight into environmental and disturbance factors, but for the sake of brevity and clarity, a verbal summary of results has been given, while the corresponding figures and associated tables can be found in the supplementary material section.

3.3.1 Vegetation patterns in spotted gum open forest (RE 12.11.5e)

3.3.1.1 Composition UPGMA of native species presence/absence for RE 12.11.5e (Figure 3.1 and Supplementary Table 3.7) indicated a strong relationship between latitude and species composition, with the majority of sites splitting along subregional affiliations. This pattern was supported by CCA of species presence/absence data (Supplementary Figure 3.1 and Supplementary Table 3.8), with a large proportion of the ordination variance attributed to climatic components associated with a latitudinal gradient, including average annual temperature and rainfall variability. Eucalyptus carnea and Eucalyptus acmenoides were significantly related to climate and geographical location due to the restricted distribution of Eucalyptus carnea

53 among Brisbane sites and Eucalyptus acmenoides being confined to the Gympie subregion.

CCA of species presence/absence showed that the first six significant axes (p<0.05) explained approximately 65% of the variance. Much of the variation among sites and species was related to environmental rather than disturbance gradients. The first axis indicated a relationship between increased average annual temperature and latitude, and a suite of species in Gympie National Park (GYM2). This site, although structurally defined as open forest, contained several rainforest/wet sclerophyll (Astrotricha latifolia, Cyperus tetraphyllus, Marsdenia sp.) and regionally restricted (Eucalyptus cloeziana) species. Additionally, increased rainfall variation at lower latitudes was related to several Brisbane sites including BRIS3, BRIS6 and BRIS22. These sites were also defined by the presence of rainforest/wet sclerophyll plants (Cupaniopsis parvifolia, Hippocratea barbata, Nyssanthes diffusa) and a suite of exotic weed species including Ageratum conyzoides, Cardiospermum grandiflorum and Desmodium uncinatum. Altitude (GYM 1 and GYM2) and rural residential cover (BRIS6 and BRIS22) were significant for the same group of sites and associated weeds. Increased temperature variability was negatively related to several sites including GYM1, GYM2 and BRIS20 and an assemblage of species found in wet sclerophyll and rainforest (Astrotricha latifolia, Cyperus tetraphyllus and Pittosporum revolutum) and riparian areas (Plectranthus parviflorus and Velleia paradoxa). A relationship was identified between species, sites and fragmentation processes, with decreased local road cover and increased patch size being linked to a suite of species (Chrysocephalum apiculatum, Crotalaria montana, Gompholobium pinnatum, Senecio hispidulus and Xanthorrhoea latifolia) found in Tamborine National Park (BRIS 7 and BRIS8). The CCA revealed a strong positive relationship between average annual rainfall, BRIS1, BRIS2 and GYM7, and Arundinella nepalensis, a dry woodland and grassland species often found in drainage ways (National Herbarium of NSW 2014a). Grazing pressures were significant for several Gympie sites, including GYM10 and GYM12 and a suite of native species. Eragrostis

54 sororia and Pratia concolor exhibited strong negative relationships with grazing, with Eragrostis sororia being a common component of native pasture.

Figure 3.1: UPGMA, using Bray-Curtis index, of native species presence/absence for RE 12.11.5e, indicating broad associations.

3.3.1.2 Abundance Trees

UPGMA analysis of native large tree species abundance (Figure 3.2 and Supplementary Table 3.9) indicated similar patterns to those exhibited for native species presence/absence data, with the primary split occuring between the Brisbane and Gympie subregions. CCA of all large tree species abundance (Supplementary Figure 3.2 and Supplementary Table 3.10) indicated that Eucalyptus acmenoides, Eucalyptus cloeziana, Eucalyptus exserta and Eucalyptus longirostrata were significantly related to increased average annual temperatures and plant available water in the Gympie subregion (GYM2, GYM4, GYM9, GYM10 and GYM14). Logging was the predominant disturbance driver of large tree numbers, particularly Eucalyptus acmenoides among Gympie sites. Grazing had a significant positive relationship with several tree species (Allocasuarina torulosa, Corymbia intermedia, Cupaniopsis parvifolia and Eucalyptus tereticornis) for sites in the Brisbane and Gympie subregions. High local road cover and decreased soil bulk density were linked to reduced Eucalyptus cloeziana abundance.

UPGMA analysis of native tree species abundance (Figure 3.3 and Supplementary Table 3.11) also revealed a split between the Brisbane and Gympie subregions due to the restricted distribution of Eucalyptus carnea to Brisbane sites and higher numbers of ironbark species (Eucalyptus crebra and Eucalyptus siderophloia) in the Brisbane subregion. The first axis for the CCA of all tree species (Supplementary Figure 3.3 and Supplementary Table 3.12) explained approximately 35% of the variance, and suggested that climatic variables linked to latitude were related to changes in tree abundance for RE 12.11.5e. Angophora woodsiana and Corymbia trachyphloia numbers decreased in the Gympie subregion with increased average annual temperatures and reduced rainfall variability. The second and third axes explained a further 31% of the variance. Grazing was significantly related to GYM13 and Allocasuarina torulosa, and high local road cover was related to several small

56 urban Brisbane patches (BRIS18, BRIS20 and BRIS23) possessing high numbers of Acacia disparrima.

Figure 3.2: UPGMA, using Bray-Curtis index, of native large tree species abundance for RE 12.11.5e, indicating broad associations.

Figure 3.3: UPGMA, using Bray-Curtis index, of native tree species abundance for RE 12.11.5e, indicating broad associations.

Shrubs

UPGMA analysis (Figure 3.4 and Supplementary Table 3.13) indicated that geographical location had an important yet reduced effect on shrub species abundance compared to overstorey layers. CCA suggested that RE 12.11.5e shrub species abundance (Supplementary Figure 3.4 and Supplementary Table 3.14) was primarily related to surrounding land use and fire activity, although latitudinally- related climatic conditions were still significant. High rural residential cover and reduced fire height were related to increased numbers of dry rainforest (Alyxia ruscifolia, Cupaniopsis parvifolia and Guioa semiglauca) and weed species (Ochna serrulata, Senna pendula var. glabrata and Solanum mauritanum) across several Brisbane and Gympie sites (BRIS6, BRIS22 and GYM17). Increased temperature and rainfall variability were also significant for the same suite of species, particularly in the Brisbane subregion. Sites located in large Brisbane patches were related to variable temperatures and species such as Indigofera australis, Persoonia sericea and Xanthorrhoea johnsonii. Logging activity and increased average annual temperatures were significant for Pittosporum revolutum in the Gympie subregion. A geographical disjunction between Brisbane and Gympie was also found with Acrotriche aggregata and Zieria smithii more abundant in the Gympie subregion. Fragmentation processes, represented by increased local road cover, were related to increased numbers of Corymbia trachyphloia in shrub layers for a number of Brisbane sites (BRIS10 and BRIS16).

Figure 3.4: UPGMA, using Bray-Curtis index, of native shrub species abundance for RE 12.11.5e, indicating broad associations.

Ground

UPGMA analysis of native ground cover abundance (Figure 3.5 and Supplementary Table 3.15) indicated that geographical associations were less pronounced when compared to other vegetation strata. CCA of ground cover abundance (Supplementary Figure 3.5 and Supplementary Table 3.16) suggested that ground species responded primarily to changes in land use intensity and disturbance processes. Increased rural residential cover and decreased fire height were positively related to invasive weeds (Passiflora suberosa, Lantana montevidensis) and shade tolerant plants (Oplismenus hirtellus ssp. imbecillis, Ottochloa gracillima, Pseudoranthenum variabile) across a number of Brisbane a Gympie sites (BRIS3, BRIS6, BRIS22, GYM1, GYM17 and GYM18). Increased patch size was significant for species such as Xanthorrhoea johnsonii and negatively related to plants associated with reduced fire height (e.g. Pseudoranthenum variabile). Temperature variability was a significant climatic variable for species such as Smilax australis, particularly in the Brisbane subregion. Altitude was positively related to species including Arundinella nepalensis, Carissa ovata and Smilax australis, and negatively related to invasive weed species such as Lantana montevidensis and Urochloa decumbens.

Figure 3.5: UPGMA, using Bray-Curtis index, of native ground species cover abundance for RE 12.11.5e, indicating broad associations.

3.3.1.3 Association with geographic distance Mantel tests indicated that most matrices were significantly and positively correlated, except for native large tree abundance and native ground cover (Table 3.3). Results suggested that native presence/absence data was strongly related to native ground cover, with strength diminishing for each successive vegetation stratum. Adjacent vegetation layers (e.g. shrub and ground strata) were more closely related than those further removed. Geographical distance was strongly correlated with large native tree abundance, native tree abundance and native presence/absence, and weakly related to shrub and ground strata.

Native Large All native Native Native presence/ native tree tree Shrub ground absence abundance abundance abundance cover

Large native tree r=0.270 abundance p=0.001

All native tree r=0.274 r=0.631 abundance p=0.001 p=0.001

Native Shrub r=0.488 r=0.171 r=0.219 abundance p=0.001 p=0.023 p=0.001

Native ground r=0.570 r=0.126 r=0.124 r=0.281 cover p=0.001 p=0.073 p=0.035 p=0.002

Geographical r=0.320 r=0.501 r=0.553 r=0.167 r=0.153 distance p=0.001 p=0.001 p=0.001 p=0.002 p=0.001

3.3.2 Vegetation patterns in scribbly gum woodland (RE 12.5.3)

3.3.2.1 Composition UPGMA analysis of native species presence/absence for RE 12.5.3 (Figure 3.6 and Supplementary Table 3.17) indicated that sites clustered due to the emergence of distinct heathland and woodland/open forest vegetation groups, and geographical location. CCA of species presence/absence (Supplementary Figure 3.6 and Supplementary Table 3.18) indicated that sites 1253_1, 2, 3, 5 and 20 were positively related to patch size and fire proportion, and negatively related to climatic variables (temperature and rainfall variability) with latitudinal associations. These sites were described by a suite of species including Austromyrtus dulcis, Dampiera stricta, Epacris obtusifolia, Petrophile shirlyae and Strangea linearis, reflecting heathland vegetation assemblages in large, northern patches. Conversely, sites such as 1253_18 and 1253_19 were located in small, isolated woodland/open forest patches, negatively related to fire proportion and associated with a number of weed species including Desmodium uncinatum and Senna pendula var. glabrata. Sites including 1253_7, 1253_9, 1253_11, 1253_17 and species such as Eucalyptus propinqua were related to climatic variability and increased soil bulk density.

Figure 3.6: UPGMA, using Bray-Curtis index, of native species presence/absence for RE 12.5.3, indicating broad associations.

3.3.2.2 Abundance Trees

UPGMA analysis of native large tree species abundance for RE 12.5.3 (Figure 3.7 and Supplementary Table 3.19) suggested that geographical location, landform and substrate were important for canopy vegetation. CCA of all large tree species abundance for RE 12.5.3 (Supplementary Figure 3.7 and Supplementary Table 3.20) indicated that low fire proportion was significant for Angophora leiocarpa, Eucalyptus propinqua and exotic Pinus species and a single site (1253_18) located in a small, fire-excluded patch. Geographical location had a significant effect on large tree species numbers. Higher latitudes were linked to the same site and species negatively related to fire proportion. Lower latitudes were related to species such as Eucalyptus crebra and Eucalyptus resinifera. Some southern sites (1253_15 and 1253_16) and species (Acacia disparrima, Banksia integrifolia and Melaleuca sieberi) were associated with increased plant available water and local road cover. Several sites were related to altitude and patch size. 1253_1 and 1253_6 were located in large patches at low elevations, and were associated with species such as Callitris columellaris, a species found in wallum woodland, and Banksia serrata.

UPGMA analysis of native tree species abundance for RE 12.5.3 (Figure 3.8 and Supplementary Table 3.21) indicated that sites clustered based on heathland and woodland/open forest associations, geographical location and the impacts of fire exclusion. CCA of all tree species (Supplementary Figure 3.8 and Supplementary Table 3.22) exhibited similar relationships to those displayed by large trees. Reduced fire proportion was related to increased numbers of Lophostemon suaveolens, Melaleuca quinquenervia and Pinus sp., and 1253_18. Increased local road cover and plant available water corresponded with an increased abundance of Allocasuarina littoralis, Banksia integrifolia, Melaleuca sieberi, Corymbia trachyphloia and Eucalyptus microcorys for 1253_7, 1253_15 and 1253_16.

Figure 3.7: UPGMA, using Bray-Curtis index, of native large tree species abundance for RE 12.5.3, indicating broad associations.

Figure 3.8: UPGMA, using Bray-Curtis index, of native tree species abundance for RE 12.5.3, indicating broad associations.

Shrubs

UPGMA analysis of native shrub species abundance for RE 12.5.3 (Figure 3.9 and Supplementary Table 3.23) indicated that latitudinal gradients had less of an effect on shrub species compared to overstorey layers, with sites from northern, central and southern locations clustering together. However distinct northern heathland and southern woodland sites were also evident. A single axis was significant for the CCA of all shrub abundance (Supplementary Figure 3.9 and Supplementary Table 3.24), explaining 9.74% of the variance. Increased fire proportion and altitude were related to high numbers of Hakea florulenta, and a lower abundance of exotic species such as Pinus sp., Schefflera actinophylla and Ochna serrulata. Although not significant, vegetation associations linked to geography were suggested by the ordination.

Figure 3.9: UPGMA, using Bray-Curtis index, of all native shrub species abundance for RE 12.5.3, indicating broad associations.

Ground UPGMA analysis of ground cover abundance for RE 12.5.3 (Figure 3.10 and Supplementary Table 3.25) indicated that species assemblages were related to vegetation associations (heathland or woodland/open forest). Substrate moisture, particularly among seasonally wet heaths, was important for a range of species such as Ptilothrix deusta and Leptocarpus tenax. Both the absence of fire and recent burning activity had similar effects on understorey species by reducing native grass cover. Latitudinal effects related to vegetation associations (heathland in the north and woodland/open forest to the south) were observed. CCA results indicated that all axes were non-significant for RE 12.5.3 ground layers (Supplementary Table 3.26).

Figure 3.10: UPGMA, using Bray-Curtis index, of all native ground species abundance for RE 12.5.3, indicating broad associations.

3.3.2.3 Association with geographic distance Mantel tests indicated that presence/absence data were positively correlated with other vegetation matrices, but were not significantly related to geographical distance (Table 3.4). Native large trees were correlated with all native tree abundance. Shrub and ground matrices were significantly correlated, however understorey strata were not related to either tree dataset. Geographical distance had strong significant relationships with large native tree abundance, all native tree abundance and native shrub abundance matrices.

Native Large All native Native Native presence/ native tree tree Shrub ground absence abundance abundance abundance cover

Large native tree r=0.210 abundance p=0.026

All native tree r=0.272 r=0.542 abundance p=0.006 p=0.001

Native Shrub r=0.486 r=0.110 r=0.191 abundance p=0.001 p=0.145 p=0.054

Native ground r=0.598 r=0.130 r=0.129 r=0.310 cover p=0.001 p=0.072 p=0.112 p=0.002

Geographical r=0.200 r=0.342 r=0.319 r=0.507 r=0.098 distance p=0.054 p=0.007 p=0.011 p=0.001 p=0.191

3.4 Discussion

Tree, shrub and ground species composition and abundance for RE 12.11.5e and RE 12.5.3 were linked to multiple environmental gradients and disturbance regimes. Observed interactions between vegetation communities and abiotic patterns within the study area suggest that both regional ecosystems conform to a continuum model, but exhibit differential responses to deterministic and stochastic ecological processes. While many of these responses were restricted to each vegetation community, latitudinal effects were universally important and represent an interplay between environmental niche-based processes and disturbance variables linked to relative landscape position and plant distributions (Austin et al. 1985; Reed et al. 1993; Austin et al. 1996). These effects were evident from analyses of presence/absence data for both regional ecosystems, and broadly supported by Mantel tests of geographical distance.

3.4.1 Factors influencing vegetation patterns in spotted gum open forest

RE 12.11.5e sites split along geographical lines defined by the Brisbane and Gympie subregions. Much of this variation may be attributed to environmental components related to plant-water balance (Stephenson 1990; Currie 1991; Austin et al. 1996). Changes in temperature and rainfall represented broad climatic envelopes which can significantly impact plant species persistence (Austin et al. 1985; Margules et al. 1987; Specht and Specht 1989a; Rohde 1992; Egan and Williams 1996; Williams et al. 1996; Fang et al. 2012). Results indicated that the physiological tolerances of individual species to climatic variables, allied with fundamental and realised niche conditions, were principal drivers of vegetation patterns (Austin et al. 1990). Wet sclerophyll and rainforest plants (Astrotricha latifolia, Cyperus tetraphyllus and Pittosporum revolutum) and riparian species (Plectranthus parviflorus and Velleia paradoxa) linked to higher rainfall and temperatures in the Gympie subregion were shown to be adversely affected by climatic variability. These relationships suggested that unpredictable precipitation, coupled with thermal stress, excluded or reduced the abundance of certain species across the study area. However, it may

71 also be argued that the restricted distributions of common species such as Pittosporum revolutum were due to reduced sampling effort in patches with high numbers of understorey rainforest plants. Alternatively, stochastic demographic and dispersal processes may have been instrumental in determining the persistence of these species, particularly within small, isolated patches of remnant vegetation (Chase and Myers 2011).

Similarly, the role of temperature and precipitation variability in the geographical confinement of several rainforest species (Hippocratea barbata, Nyssanthes diffusa and Cupaniopsis parvifolia) to the Brisbane subregion may also be linked to sampling effort and stochastic effects. However, the presence of these and other rainforest plants, and an assemblage of exotic weeds including Ageratum conyzoides, Cardiospermum grandiflorum and Desmodium uncinatum was also related to altered fire regimes. Changes in vegetation composition and structure, including increased canopy cover, were indicative of fire exclusion and the transition from open eucalypt systems to rainforest (Figure 3.11).

Figure 3.11: Photographs of RE 12.11.5e site managed with fire BRIS1 (D’Aguilar 1) and fire-excluded dry rainforest BRIS22 (Morrison Road).

Fire age data resolution was low across most sites due to missing or inaccurate historical fire records, with fire height the only significant disturbance variable related to burning of native vegetation. Conspicuous transitional changes in vegetation provided a broad, yet adequate, estimation of time since last burn. Although fire patterns may partly explain changes in vegetation composition, the effects of climatic variability cannot be completely discounted, and may in itself be considered a form of disturbance (Pausas and Austin 2001). Increased temperature and rainfall variability can create favourable conditions for weed growth in disturbed habitat, or in fragments where fire has been removed over long periods of time (Flory and Clay 2009; Lewis and Debuse 2012).

Variability in the tree layer was mainly driven by latitudinally-related climatic components. This relationship may be attributed to the deep roots of large trees mitigating the impacts of fine-scale, episodic processes attributed to natural variation or human disturbance. The restricted distributions of Eucalyptus carnea to the Brisbane subregion, and Eucalyptus acmenoides and Eucalyptus cloeziana to the Gympie subregion were primary indicators of geographical disjunction, and may be partly linked to elevated average annual temperatures at higher latitudes. Results indicated that the increased abundance of Angophora woodsiana among Brisbane sites was also linked to subregional climatic envelopes. Fraser Island, located approximately 50 km north of Gympie, represents the northern limit for Angophora woodsiana (EUCLID Eucalypts of Australia 2014a), with its southern extent terminating along the central coast of New South Wales (National Herbarium of NSW 2014c). Observed vegetation patterns suggested that, although restricted by adverse temperature gradients, Angophora woodsiana may also be competitively displaced in the Gympie subregion by species adapted to increased thermal pressures (Margules et al. 1987).

Although climate had a significant impact on RE 12.11.5e tree strata, overstorey species also responded to changes in substrate properties and disturbance

73 processes. High numbers of Eucalyptus acmenoides, typically found on deeper soils of moderate fertility and regular moisture (National Herbarium of NSW 2014b), were related to increased plant available water within the Gympie subregion. In addition, logging activity in Gympie forests resulted in reduced numbers of favoured timber species such as Corymbia citriodora and Eucalyptus crebra. The compositional and structural simplification of native vegetation, as indicated in many of these actively managed forests, is a standard forestry practice used to optimise timber production for commercial harvest (Horner et al. 2012).

Results for shrub strata indicated a strong association with changes in abiotic conditions and disturbance processes, although latitudinal gradients had a reduced, yet still significant effect. Climatic attributes were found to be important drivers of shrub abundance. The influence of increased average annual rainfall in the Gympie subregion, and high temperature variability in the Brisbane subregion mirrored those relationships observed in the presence/absence data. However, land use intensification had the strongest effect on shrub composition and numbers. Rural residential cover was indicative of landscape fragmentation and altered fire regimes for RE 12.11.5e, particularly on the margins of large contiguous patches (eg. BRIS6, BRIS22 and GYM17). These sites retained many native species, but were also characterised by an absence of eucalypt recruitment, and the transition from open forest to communities dominated by dry rainforest (Alyxia ruscifolia, Cupaniopsis parvifolia, Guioa semiglauca, Jagera pseudorhus) and exotic (Ochna serrulata, Senna pendula var. glabrata, Solanum mauritanum) species. Low fire activity across these sites resulted in shade-tolerant midstorey vegetation dominating shrub strata (Close et al. 2009; Lewis and Debuse 2012). Consequently, developing shrub layers can compete with eucalypt canopy species for water resources, as well as adversely affecting eucalypt-ectomycorrhizal relationships through modified microclimatic conditions (Close et al. 2009; Jurskis and Walmsley 2012). Eucalypt recruitment can be limited under these circumstances due to specific adaptations to low nutrient soils, drought and moderate to high fire environments. Where burning is removed

74 or is infrequent and of high intensity, then fire intolerant and shade-adapted species become established (Margules et al. 1987; Close et al. 2009).

Fire patterns and fragmentation were also responsible for much of the variation in RE 12.11.5e ground species composition and abundance, with latitudinal gradients showing less of an effect relative to shrub and tree layers. This effect may be due to plant-plant interactions moderating relationships between vegetation and the abiotic environment at fine scales where relatively shallow rooted ground species are in close contact or competing for the same resources (Reed et al. 1993). As with RE 12.11.5e shrub strata, decreased fire activity was linked to a variety of exotic species, including Passiflora suberosa and Lantana montevidensis. Many common forbs and graminoids were absent from these sites, with high covers of shade tolerant plants, including Ottochloa gracillima and Claoxylon australe, found under closed canopies of dry rainforest and introduced tree species (Peet 1978, Specht and Specht 1993).

Rural residential cover was the predominant indicator of land use intensification for RE 12.11.5e ground layer vegetation. Rural residential cover was associated with weeds such as Desmodium uncinatum, a cattle fodder crop and common invasive species now naturalised across South-East Queensland (Weeds Australia 2014). High local road cover, indicative of landscape fragmentation, was also linked with an increased weed presence among ground layers, including the introduced forage grass Urochloa decumbens. This species has a persistent seed bank typical of many exotic grasses, often replacing native species where disturbance occurs (Odgers 1999).

3.4.2 Factors influencing vegetation patterns in scribbly gum woodland

Compositional data for RE 12.5.3 also reflected a north-south gradient. However, this gradient was primarily related to geographically-restricted vegetation

75 assemblages, patterns of fragmentation and land use change. Heathland vegetation was predominantly found in the northern extent of RE 12.5.3 and woodland/open forest to the south (Figure 3.12).

Figure 3.12: Photographs of RE 12.5.3 heathland site 1253_3 (Caves Road) and woodland/open forest site 1253_8 (Sheepstation Creek).

While not indicated in CCA results, north-south associations may be linked to substrate conditions, with heath species generally found on shallow, less productive soils and woodland/open forest species on deeper soils of moderate to high fertility (Specht and Specht 1989b). In addition, northern areas, associated with forestry, agriculture and rural residential development, were subjected to reduced levels of fragmentation and land use intensity when compared to the suburban and peri- urban landscapes defining southern sites. Latitudinally-related climatic variables also had some effect on RE 12.5.3 vegetation. Several heathland species (Austromyrtus dulcis, Dampiera stricta, Petrophile shirlyae and Strangea linearis) were absent from areas associated with increased temperature stress and variable rainfall conditions. This link with predictable climatic patterns may be underpinned by high soil moisture requirements for these species, compounded by their presence on shallow substrates with low water holding capacities (Specht and Specht 1989b).

While RE 12.5.3 tree layers varied with latitude, this variation was primarily attributed to the disjunction between heathland and woodland/open forest and geographically-related disturbance regimes rather than climatic conditions. It has been demonstrated that the reduced tree species richness and cover associated with heathland vegetation is related to the limited water-holding capacity and low nutrient levels of associated soils (Specht and Specht 1989b). Heathland substrates can significantly restrict annual shoot growth for overstorey trees when compared to woodland eucalypt species on more fertile soils under the same regional climatic conditions (Specht and Specht 1989b). However, soils with high moisture levels may also affect plant growth. Sites with high plant available water were shown to support fewer eucalypt species and greater numbers of wet heath trees such as Melaleuca sieberi.

Although vegetation type and soil properties were important for RE 12.5.3 overstorey species, CCA results suggested that reduced fire activity was the principal driver of vegetation change for trees within this regional ecosystem. Although the impacts of fire exclusion were limited to several small patches, the effects of removing fire from scribbly gum woodland were profound, and instrumental in the replacement of eucalypt canopies and subcanopies with Allocasuarina littoralis, Acacia disparrima and exotic Pinus species. This process was particularly evident in 1253_15 (Korawal Street, Redland), 1253_17 (Komraus Court, Caboolture) and 1253_18 (New Settlement Road, Caboolture), with Eucalyptus racemosa found in low numbers or undergoing widespread dieback (Figure 3.13). In most cases, this transition was exacerbated by landscape fragmentation. Small patches, particularly in southern suburban and peri-urban areas, were often characterised by an absence of active fire management and subsequent vegetation change.

Figure 3.13: Photographs of two small, fire-excluded RE 12.5.3 patches (1253_15 (Korawal Street) and 1253_17 (Komraus Court).

Shrub abundance was shown to be aligned with geographical distance. This result was broadly supported by the UPGMA analysis of RE 12.5.3 shrub presence/absence data, with some sites clustered partly due to northern, heathland and southern, woodland associations. This outcome suggests that the relationship between latitude and vegetation type was the predominant driver of shrub species composition and abundance for RE 12.5.3, irrespective of other effects. However, as indicated in the CCA of RE 12.5.3 shrub abundance, vegetation type was also linked to disturbance processes, with large, northern heathland patches subjected to increased fire activity when compared to smaller woodland remnants to the south.

Fire exclusion also had a dramatic effect on RE 12.5.3 understorey species (Specht and Specht 1987b; Lunt 1997; Lunt 1998; Close 2009; Jurskis and Walmsley 2012). The increased dominance of Allocasuarina littoralis among unburnt RE 12.5.3 sites suppressed shrub and ground strata through shading and the development of thick litter layers (Specht and Specht 1989b; Lunt 1998; Close 2009). Similarly, Gill and Williams (1996) reported that where exotic Pinus species have become established, thick canopies replace relatively open tree layers, shading understorey plants and

78 possibly reducing faunal abundance and species richness. The potential for pine to invade remnant forest patches is high throughout the northern and central extent of RE 12.5.3 due to the wide distribution of commercial plantations within this part of the bioregion. Pine is readily dispersed by wind and birds, with seedlings found up to four kilometres from parent plants (Gill and Williams 1996; Williams and Wardle 2005). In this study, fire exclusion in scribbly gum forest was often linked to landscape fragmentation. Reduced burning of shrub strata in small patches was related to a suite of weed species, including Schefflera actinophylla, Ochna serrulata, Pinus species and Lantana camara, with native plant cover represented primarily by depauperate and structurally simple shrub layers (Specht and Specht 1989b; Costello et al. 2000). As with tree species, fire reduction or exclusion favoured fire-averse and shade tolerant assemblages. In particular, invasive species such as Lantana camara and exotic Pinus species have been shown to respond favourably to disturbance driven by landscape fragmentation, altered fire regimes and habitat degradation, to the detriment of remaining native vegetation (Williams and Wardle 2005; Gooden et al. 2009; Lewis and Debuse 2012).

Additionally, some fire-excluded RE 12.5.3 sites were partially composed of shrub species associated with wetter soil conditions, including Melaleuca quinquenervia and Lophostemon suaveolens. Melaleuca quinquenervia is found on lower eroded slopes in this regional ecosystem (Department of Environment and Heritage Protection 2014b), and although fire tolerant, is often associated with high water tables. Most fires in Melaleuca forest are of low to moderate intensity due to decreased fuel loads linked to low nutrient soils (Skull and Adams 1996). It may be expected that small forest remnants on moist substrates, and dominated by Melaleuca quinquenervia, exotic species and fire-averse native shrubs and trees, would be subjected to reduced fire activity.

Ground species also reflected patterns exhibited by corresponding shrub layers. Species composition was related to vegetation type, fire activity and soil moisture.

Large northern heathland remnants typically contained species such as Ptilothrix deusta, Daviesia umbellulata, Strangea linearis, Leucopogon juniperinus and Pultenaea myrtoides, with Imperata cylindrica, Pteridium esculentum and Lobelia purpurascens often found in areas dominated by woodland/open forest. Fire exclusion and recent burning had similar effects on ground species composition for several sites, where many common grasses were absent from assemblages where either ground layers were shaded and litter levels high in long unburnt patches, or where fire had consumed large quantities of flammable plant material. Ground species growing on damp substrates and dominated by sedges such as Leptocarpus tenax, were notably subjected to reduced fire activity, particularly in small, isolated patches.

3.5 Conclusion

Changes in vegetation composition and abundance for RE 12.11.5e and RE 12.5.3 were related to latitudinal effects. Where these effects were explicitly tested across two distinct subregions for RE 12.11.5e, there was a general trend of decreasing importance for latitudinally-related climatic gradients with progression from tree to ground layers. The influence of non-climatic environmental attributes and disturbance regimes were increasingly important for shrub and ground strata, with soil conditions, fire activity, land use intensification and habitat fragmentation shaping understorey vegetation patterns. However, where the same environmental and disturbance variables were examined for a different ecosystem type (RE 12.5.3), it was shown that latitudinal variation was attributed to the geographical disjunction between heathland and woodland/open forest and geographically- related disturbance processes linked to fire regimes and landscape fragmentation. These findings align with a Gleasonian model, suggesting that environmental gradients underpinned vegetation patterns. However, the two regional ecosystems may also be perceived as distinct communities arising as a function of the landscape. Management practices and disturbance factors were instrumental in generating ecological change, with regional ecosystems exhibiting a range of

80 vegetation states. Although commonalities were apparent across regional ecosystems, it was shown that each community was defined by distinct biological assemblages, biotic-abiotic relationships and anthropogenic impacts. Based on these results, and in answer to the first key question of the study, changes in the composition and structure of RE 12.11.5e and RE 12.5.3 were related to a range of environmental and disturbance variables. Therefore, the question must then be asked whether related ecological condition attributes exhibit similar responses to the same suite of variables.

Chapter 4: The effects of environmental variability and disturbance

regimes on vegetation condition in spotted gum open forest (RE

12.11.5e) and scribbly gum woodland (RE 12.5.3)

4.1 Introduction

4.1.1 Ecological surrogacy and condition assessments

Ecological surrogates are used to assess and simplify complex biological associations (Feld 2010). The design and utility of ecological surrogates as tools for conservation management are based on several key components. These components include the selection of species or environmental features used to represent broader biodiversity, establishing the scale and geographical region over which surrogates are applied, and determining the form, scope and significance of relationships between indicator taxa or environmental features and important ecological counterparts.

Ecological surrogates, of which measurements of ecological condition are a subset, can be defined by either a cross-taxon (Pearson and Cassola 1992; Rodriguez et al. 1998; Ricketts et al. 2002; Lawler et al. 2003; Sverdrup-Thygeson and Lindenmayer 2003; Schmit et al. 2005; Tognelli 2005) or an environmental focus (Ferrier and Watson 1997; Saetersdal et al. 2004; Faith et al. 2004; Juutinen and Monkkonen 2004; Sarkar et al. 2005; Buxbaum and Vanderbilt 2007; Rodrigues and Brooks 2007; Juutinen et al. 2008). Cross-taxon condition metrics use species-level associations to gauge the quality or condition of an ecosystem, with the presence or abundance of certain organisms measuring the extent of human impact. Cross- taxon approaches typically take the form of environmental health (Avidano et al. 2005; Beketov and Liess 2008; Borja and Dower 2008; Hampel et al. 2009), management (Landres et al. 1988; Simberloff 1998) or biodiversity indicators (Pearson 1994; Ferrier and Watson 1997; Blair 1999; Warman 2004). Cross-taxon indicators require surrogate species to be sensitive to pollution, disturbance or habitat change, have a well-known biology, be easily observed and measured, and be broadly distributed over a variety of habitats (Caro and O’Doherty 1999; Lawler and White 2008).

Although cross-taxon indicators are often used as surrogates for biodiversity and ecological condition, their surrogacy potential remains uncertain (Chase et al. 2000; Bonn et al. 2002; Ricketts et al. 2002; Lawler et al. 2003; Sauberer et al. 2004; Tognelli 2005). Contradictory results have been linked to spatial and temporal limitations, poor indicator validation, and ambiguous conceptual frameworks that fail to define represented ecological states or processes (Niemi and McDonald 2004). Saetersdal et al. (2004) have argued that the use of surrogate species is limited by some fundamental assumptions about species-specific patterns of community composition, particularly that patterns are comparable across multiple taxa. Cross-taxon indicators tend to have a narrow focus, and when concerned with broad measures of ecological health or biodiversity, fail to capture adequate levels of variation

Environmental surrogates may address the limitations of cross-taxon approaches. Environmental surrogates can combine biotic and abiotic data over different spatial scales for multiple geographical areas (Andreasen et al. 2001; Rodrigues and Brooks 2007; Moreno-Rueda and Pizarro 2009). Environmental surrogates are based on readily mapped environmental components such as land classes or vegetation types (MacNally et. al 2002a; Singh et al. 2010; Capotorti et al. 2012), with composition, structural characteristics, abiotic features and landscape-level associations serving as ecological indicators (Faith et al. 2004; Ferrier 2005; Rodrigues and Brooks 2007; Torras et al. 2008).

Ecological condition surrogates often incorporate environmental attributes into multimetric indices that are easier to understand and compare than individual indicators alone (Bleby et al. 2008). Expert opinion is frequently used to assess and weigh potential attributes based on their feasibility and perceived ecological importance. Once assessed, attributes may then be adopted and combined into condition indices (Oliver et al. 2007). The simplest indices add attribute data as equally valued scores, recorded as either presence/absence or ranked based on a

86 predetermined category. However, weighted approaches, where each attribute is assigned a certain level of ecological priority before it is summed or averaged, are considered to more accurately reflect changes in ecosystem condition (Parkes et al. 2003; Eyre et al. 2011b). Regardless, there is an underlying assumption that strategic attribute selection addresses the shortcomings of cross-taxon indicators, maximises the capacity of environmental surrogates to accurately measure target biodiversity or ecological condition, and captures biotic and abiotic variation across regions.

Most terrestrial ecological condition metrics are dominated by measurements of vegetation characteristics (Noss 1999; Lopez and Fennessy 2002; Parkes et al. 2003; Oliver 2004; Standovar et al. 2006; Gibbons et al. 2009; Wehenkel et al. 2009; Winter et al. 2010; Yapp et al. 2010). Vegetation is fundamental to ecosystem functioning. Vascular plant composition is often associated with wider species diversity (Juutinen and Monkkonen 2004; Saetersdal et al. 2004). Furthermore, vegetation communities are perceived to be robust to seasonal changes in climate, and variation among plant community composition and structure represents important environmental gradients and disturbance regimes (Ferrier and Watson 1997; Saetersdal et al. 2004; Bleby et al. 2008; Juutinen et al. 2008). Although rigorous design and field testing have been used to assess the capacity for vegetation condition attributes to represent broader biodiversity values (Eyre and Neldner 2011; Kelly et al. 2011; Eyre et al. 2011b), no studies have examined the effects of environmental gradients and disturbance regimes on vegetation condition attributes.

4.1.2 Aims

It was shown in Chapter 3 that environmental gradients and disturbance regimes influenced vegetation composition, abundance and structure in spotted gum open forest (RE 12.11.5e) and scribbly gum woodland (RE 12.5.3) communities in South-

East Queensland. Therefore, it was necessary to consider that the same sources of variability may also influence ecological condition measures. Specifically, the aim of chapter 4 was to determine and quantify changes in vegetation condition attributes in RE 12.11.5e and RE 12.5.3 in response to environmental variability and disturbance regimes.

It was critical to establish this because the effective operation of multimetric ecological condition surrogates depends on these measures being robust to environmental variability, so that they can be employed across broad geographical areas, and under a range of climatic conditions and disturbance regimes. Finally, the implications of these results on metric efficacy will be discussed, and in particular, strategies to mitigate the effects of variability through changes in metric application and design will be considered.

4.2 Materials and methods

4.2.1 Field surveys

The two regional ecosystems examined in Chapter 3 were also used for this component of the study. Forty-two sites for RE 12.11.5e (spotted gum open forest) and twenty sites for RE 12.5.3 (scribbly gum woodland) were selected as per methods detailed in Chapter 2.4. RE 12.11.5e sites were located within the Brisbane subregion to the south and the Gympie subregion to the north (Figures 2.4 and 2.5). RE 12.5.3 sites were located between Brisbane to the south and Noosa to the north, and centred on the Sunshine Coast hinterland (Figure 2.6).

Field surveys were conducted as per methods detailed in Chapter 2.5. Each site was subjected to a BioCondition assessment (Eyre et al. 2011b), with all vegetation and landscape attributes recorded. Benchmarks were derived from reference site values. Reference sites were BRIS1 (D’Aguilar National Park) and BRIS4 (Moggill Conservation Park) for the RE 12.11.5e Brisbane subregion; GYM2 (Gympie National Park) and GYM7 (Goomboorian State Forest) for the RE 12.11.5e Gympie subregion; and 1253_3 (Caves Road), 1253_4 (Murphys Road) and 1253_5 (Scientific Area 1) for RE 12.5.3.

4.2.2 Data analyses

4.2.2.1 Descriptive statistics Descriptive statistics, including mean and median values, standard deviations, and minimum and maximum values were generated for vegetation attribute values for each RE 12.11.5e subregion, all RE 12.11.5e sites combined and all RE 12.5.3 sites using SPSS V.21 (IBM Corporation 2012). Descriptive statistics are provided for RE 12.11.5e subregions in Supplementary Table 4.1, all RE 12.11.5e sites in Supplementary Table 4.2, and RE 12.5.3 in Supplementary Table 4.3.

4.2.2.2 Hierarchical partitioning Hierarchical partitioning was used to identify significant sources of variation for ecological condition vegetation attributes. Hierarchical partitioning is a multiple regression approach that can identify the independent effects of predictor variables on condition attribute variation (Chevan and Sutherland 1991; MacNally 2002b). Hierarchical partitioning averages the influence of variables across all possible models, therefore mitigating the effects of multicollinearity (MacNally 2002b). The disturbance variables selected for canonical correspondence analyses in Chapter 3 (Table 3.1) were also used as predictor variables for the hierarchical partitioning of raw vegetation attribute values for RE 12.11.5e and RE 12.5.3 (Table 2.3). Predictor variables and assessable BioCondition vegetation attributes were standardised using the formula: Z=(X-μ)/σ, where Z is the standardised value, X is the unstandardized value, μ is the mean, and σ is the standard deviation. Standardised disturbance values (predictor variables) and raw vegetation attribute values (response variables) were subjected to hierarchical partitioning using R studio (R Core Team 2013) and R package: hier.part (Walsh and MacNally 2013). A randomisation approach using Monte Carlo permutation tests was used to compute Z-scores and the statistical significance (p<0.05) of each disturbance variable (MacNally 2002b) using the R package: coin (Hothorn et al. 2014).

Standardised RE 12.11.5e and RE 12.5.3 environmental values, also identified for canonical correspondence analyses in Chapter 3 (Table 3.2), were used as predictor variables. Environmental predictor variables and raw vegetation attributes values (response variables) were subjected to the same processes of hierarchical partitioning and Monte Carlo permutation tests.

The disturbance and environmental values identified as significant predictor variables for RE 12.11.5e and RE 12.5.3 from the previous, separate analyses were

90 combined for each regional ecosystem and used for the hierarchical partitioning of raw vegetation attributes values using the R package: hier.part (Walsh and MacNally 2013). Monte Carlo permutation tests were used to compute Z-scores and the statistical significance of each variable using the R package: coin (Hothorn et al. 2014).

4.2.2.3 ANCOVAs and linear regressions ANCOVAs and linear regressions were used to further examine the relationships between ecological condition vegetation attributes and significant environmental and disturbance variables identified through hierarchical partitioning. Vegetation attributes, disturbance and environmental variables for RE 12.11.5e sites were tested for normality using the Shapiro-Wilk test and homogeneity of variance using Levene’s test in SPSS V.21 (IBM Corporation 2012). Those variables that violated normality and/or homogeneity of variances were transformed. Results for Shapiro- Wilk tests and Levene’s tests are provided in Supplementary Tables 4.4 to 4.11 for RE 12.11.5e. However, if normality continued to be violated post-transformation, ANCOVAs were still used as they are robust to non-normal data (Barrett 2011). Additionally, if homogeneity of variances were violated and yet the ratio of the largest group variance was not more than three times that of the smallest group, ANCOVAs were also used (Glass et al. 1972). Where ANCOVA homogeneity of regression slopes was not met, and there was a significant interaction between subregion (fixed factor) and environmental or disturbance variable (covariate), linear regressions were used to explore the relationships between dependent and independent variables. All analyses were conducted using SPSS V.21 (IBM Corporation 2012). A flow chart summarising these steps is provided in Supplementary Figure 4.1.

4.2.2.4 Correlations Vegetation attributes, and disturbance and environmental variables for RE 12.5.3 sites were tested for normality using the Shapiro-Wilk test in SPSS V.21 (IBM

Corporation 2012), and transformed if normality was violated. Results for Shapiro- Wilk tests are provided in Supplementary Tables 4.12 to 4.15 for RE 12.5.3. ANCOVAs were still applied if normality continued to be violated post- transformation based on the same assumptions used for RE 12.11.5e. Pearson and Spearman correlations were used to explore the relationships between vegetation attributes and significant variables identified by hierarchical partitioning for RE 12.5.3 using SPSS V.21 (IBM Corporation 2012).

4.2.2.5 Error bar plots Error bar plots (+/- 1 standard error) of vegetation condition attributes and significant predictor variables for RE 12.11.5e and RE 12.5.3 were generated using SPSS V.21 (IBM Corporation 2012). This approach was favoured over scatterplots because error bar plot categories illustrated the relationships between environmental gradients, disturbance processes and vegetation condition attributes more clearly. Variable categories were based on median values, the number of individual data points, and the range for each variable.

4.3 Results

4.3.1 Relationships between vegetation condition attributes, environmental gradients and disturbance regimes in spotted gum open forest (RE 12.11.5e)

Descriptive statistics for RE 12.11.5e Brisbane sites, RE 12.11.5e Gympie sites and all RE 12.11.5e sites combined are provided in Supplementary Table 4.1 and Supplementary Table 4.2. Patch size varied widely for RE 12.11.5e, with the maximum patch area for Brisbane sites (44110.12 ha) more than double that for Gympie (17328.75 ha). Descriptive statistics for several other ecological condition vegetation attributes, including the number of large trees, coarse woody debris, native grass cover and litter cover also varied markedly across subregions. There were relatively high numbers of large trees in the Brisbane subregion, with a median value of 39.5 compared to 30 for Gympie. Increased levels of coarse woody debris were found in the Gympie subregion, with a median value of 59.39 m and a maximum value of 157 m, compared to a median of 30.49 m and a maximum of 91.72 m for Brisbane. Native grass cover was higher among Brisbane sites, with a median value of 33.8% versus 23.5% for the Gympie subregion. However, litter cover was higher for Gympie, with a median value of 49% compared to 35.6% for Brisbane. Other attributes were comparable across subregions, although maximum native shrub cover was high for Brisbane with a value of 116.10 m compared to 79.15 m for the Gympie.

Table 4.1: Hierarchical partitioning of all vegetation attributes based on latitude (northing), patch size, disturbance and environmental variables for RE 12.11.5e. Highlighted values are those that significantly contribute to each vegetation attribute based on Monte Carlo permutations. Gradations of shading represent the relative importance of each variable, with darkest indicating the highest importance and lightest the least importance to each attribute.

Northing/L Patch Disturbance Environmental atitude Size Attribute Northing/ Patch Fire Logging Rural Res. Grazing Local Aspect Altitude Average Rainfall Average Temp. Bulk Plant Latitude Size (PS) Height (L) Cover Cover Road (ASP) (ALT) Annual Variab. Annual Variab. Density Avail. (N) (FH) (RR) (GNE) Cover Rainfall (RCV) Temp. (TCV) (BD) Water (LR) (AAR) (AAT) (PAW) No. of large Z=-1.950 Z=0.467 Z=-0.461 Z=0.225 Z=-0.544 Z=0.601 Z=-0.554 Z=0.599 Z=-0.552 Z=0.316 Z=1.790 Z=-3.025 Z=-0.429 Z=-0.039 Z=-2.735 trees p=0.052 p=0.645 p=0.658 p=0.831 p=0.593 p=0.550 p=0.590 p=0.560 p=0.595 p=0.757 p=0.073 p=0.00 p=0.662 p=0.968 p=0.00 %v=43.28 %v=56.72 Canopy Z=0.473 Z=0.725 Z=1.045 Z=-1.395 Z=-1.117 Z=-1.223 Z=-0.469 Z=0.643 Z=1.486 Z=0.443 Z=0.449 Z=-0.182 Z=-0.105 Z=2.070 Z=-1.139 height p=0.645 p=0.481 p=0.305 p=0.165 p 0.268 p=0.220 p=0.649 p=0.529 p= 0.146 p=0.661 p=0.653 p=0.861 p=0.916 p=0.038 p=0.258 %v=34.45 CWD Z=1.72 Z=2.118 Z=-0.364 Z=1.012 Z=-0.234 Z=1.647 Z=-2.812 Z=-2.260 Z=1.673 Z=2.126 Z=-0.417 Z=0.551 Z=-1.234 Z=1.219 Z=0.453 p=0.082 p=0.034 p=0.715 p=0.314 p=0.810 p=0.102 p=0.003 p=0.024 p=0.097 p=0.033 p=0.681 p=0.587 p=0.222 p=0.225 p=0.660 %v=6.44 %v=49.59 %v=22.57 %v=21.39 Native Z=-0.530 Z=-0.224 Z=1.230 Z=-0.454 Z=-0.456 Z=-2.580 Z=1.602 Z=0.319 Z=-0.672 Z=-0.096 Z=0.754 Z=-0.192 Z=0.277 Z=-0.543 Z=0.183 grass p=0.602 p=0.822 p=0.226 p=0.651 p=0.653 p=0.008 p=0.110 p=0.756 p=0.509 p=0.927 p=0.454 p=0.850 p=0.787 p=0.586 p=0.858 cover %v=53.83 Litter Z=1.241 Z=-1.614 Z=-0.080 Z=2.202 Z=-0.324 Z=1.280 Z=0.966 Z=-0.712 Z=-1.375 Z=-0.427 Z=-2.160 Z=1.823 Z=-1.829 Z=-0.358 Z=1.479 cover p=0.227 p=0.106 p=0.935 p=0.026 p=0.749 p=0.208 p=0.338 p=0.478 p=0.170 p=0.669 p=0.031 p=0.069 p=0.065 p=0.726 p=0.145 %v=49.77 %v=50.23

Attribute Northing/ Patch Size Fire Logging Rural Res. Grazing Local Aspect Altitude Average Rainfall Average Temp. Bulk Plant Latitude (N) (PS) Height (L) Cover Cover Road (ASP) (ALT) Annual Variab. Annual Variab. Density Avail. (FH) (RR) (GNE) Cover Rainfall (RCV) Temp. (TCV) (BD) Water (LR) (AAR) (AAT) (PAW) Weed Z=-0.598 Z=-2.039 Z=-0.703 Z=0.537 Z=2.211 Z=0.326 Z=1.921 Z=0.340 Z=-0.727 Z=0.31 Z=0.924 Z=-0.294 Z=0.918 Z=-2.280 Z=-0.659 cover p=0.557 p=0.040 p=0.491 p=0.598 p=0.026 p=0.746 p=0.052 p=0.731 p=0.477 p=0.767 p=0.359 p=0.776 p=0.360 p=0.020 p=0.515 %v=3.10 %v=80.00 %v=16.91 Native Z=0.742 Z=1.249 Z=-2.742 Z=0.296 Z=-0.699 Z=0.762 Z=-0.367 Z=-0.662 Z=0.62 Z=0.654 Z=-0.137 Z=-0.758 Z=0.270 Z=0.181 Z=-1.302 canopy p=0.463 p=0.211 p=0.004 p=0.775 p=0.493 p=0.463 p=0.713 p=0.518 p=0.543 p=0.519 p=0.894 p=0.457 p=0.794 p=0.857 p=0.197 cover %v=74.92 Native Z=-1.036 Z=-3.294 Z=0.585 Z=-0.390 Z=0.134 Z=-0.282 Z=3.568 Z=0.484 Z=-2.384 Z=-2.202 Z=-0.453 Z=0.888 Z=-0.310 Z=0.414 Z=0.18 shrub p=0.308 p=0.001 p=0.570 p=0.706 p=0.896 p=0.784 p=3e-04 p=0.634 p=0.012 p=0.025 p=0.660 p=0.374 p=0.756 p=0.680 p=0.857 cover %v=19.01 %v=59.66 %v=8.56 %v=12.77 Canopy Z=1.740 Z=0.827 Z=-0.254 Z=-0.088 Z=-0.312 Z=0.295 Z=-1.085 Z=-0.201 Z=1.846 Z=1.183 Z=0.385 Z=0.145 Z=0.850 Z=0.966 Z=-0.697 richness p=0.083 p=0.418 p=0.802 p=0.931 p=0.762 p=0.770 p=0.283 p=0.841 p=0.063 p=0.245 p=0.706 p=0.888 p=0.404 p=0.343 p=0.486

Tree Z=-0.512 Z=-1.240 Z=-0.28 Z=0.812 Z=-0.591 Z=-0.040 Z=1.608 Z=0.251 Z=-0.658 Z=1.210 Z=2.070 Z=0.491 Z=-1.476 Z=0.841 Z=-0.111 richness p=0.604 p=0.223 p=0.780 p=0.424 p=0.561 p=0.967 p=0.110 p=0.810 p=0.514 p=0.231 p=0.039 p=0.628 p=0.142 p=0.418 p=0.911 %v=50.25 Shrub Z=0.618 Z=-1.127 Z=-1.864 Z=-0.101 Z=0.115 Z=-1.804 Z=0.483 Z=1.559 Z=1.179 Z=-0.761 Z=0.245 Z=-0.197 Z=0.511 Z=1.160 Z=-0.302 richness p=0.544 p=0.271 p=0.061 p=0.921 p=0.913 p=0.070 p=0.635 p=0.120 p=0.237 p=0.451 p=0.811 p=0.847 p=0.612 p=0.250 p=0.761

Grass Z=0.628 Z=0.190 Z=-0.812 Z=0.393 Z=0.104 Z=0.949 Z=-0.434 Z=-1.021 Z=-0.890 Z=-8e-04 Z=-1.555 Z=1.700 Z=-0.249 Z=-1.937 Z=1.17 richness p=0.530 p=0.852 p=0.423 p=0.695 p=0.917 p=0.349 p=0.671 p=0.305 p=0.376 p=0.999 p=0.122 p=0.093 p=0.803 p=0.053 p=0.241

Forbs Z=1.680 Z=2.814 Z=-0.859 Z=-0.685 Z=1.067 Z=1.747 Z=-4.021 Z=-1.032 Z=3.752 Z=0.707 Z=-0.846 Z=-0.894 Z=2.001 Z=0.803 Z=-0.584 other p=0.091 p=0.004 p=0.398 p=0.504 p=0.291 p=0.082 p=1e-04 p=0.307 p<0.001 p=0.487 p=0.398 p=0.390 p=0.042 p=0.421 p=0.563 richness %v=13.07 %v=51.46 %v=27.99 %v=7.48

Significant relationships identified through hierarchical partitioning were examined further using ANCOVAs and linear regressions. Shapiro-Wilk tests for normality and Levene’s tests for homogeneity of variance for untransformed and transformed RE 12.11.5e vegetation attributes and significant disturbance and environmental variables are provided in Supplementary Tables 4.4 to 4.11.

ANCOVAs for RE 12.11.5e vegetation attributes and significant variables identified through hierarchical partitioning are provided in Table 4.2. For attributes and variables where homogeneity of regression slopes was violated, linear regressions are provided in Table 4.3. The process for identifying condition attributes and variables for ANCOVAs and linear regressions is illustrated in a flow chart in Supplementary Figure 4.1. Subregions were separated out for regression analyses in order to examine potentially significant relationships between vegetation attributes and geographically-restricted environmental gradients and disturbance regimes.

Table 4.2: ANCOVA results for RE 12.11.5e vegetation attributes and significant variables identified through hierarchical partitioning. Significant values (p<0.05) are indicated by grey shading.

Attribute Significant variable Homogeneity of Levene’s test Fixed factor Covariate sig. Adjusted R² (Covariate) regression (Subregion) PAW (log) F=0.075, p=0.786 F=0.247, p=0.622 F=0.708, p=0.405 F=8.290, p=0.006 0.239 No. of large trees Average annual temp. F=9.735, p=0.003 Linear regressions Canopy height (sqrt) Bulk density Attribute violates homogeneity of variances (See Supplementary Table 4.7) Local road cover (log) Variable violates homogeneity of variances Aspect (cos) F=0.013, p=0.909 F=1.836, p=0.183 F=2.062, p=0.159 F=2.443, p=0.186 Not sig. CWD (log) Average annual rainfall F=0.435, p=0.513 F=4.556, p=0.039 ANCOVA violates homogeneity of variance Patch size (log) F=8.562, p=0.006 Linear regressions Native grass cover (arcsin) Grazing cover (arcsin) Variable violates homogeneity of variances (See Supplementary Table 4.11) Rainfall variability F=0.002, p=0.969 F=5.064, p=0.030 ANCOVA violates homogeneity of variance Litter cover (arcsin) Logging (sqrt) Variable violates homogeneity of variances (See Supplementary Table 4.11) Rural res. (arcsin) F=0.783, p=0.382 F=1.204, p=0.279 F=0.744, p=0.394 F=21.540, p<0.001 0.326 Weed cover (arcsin) Bulk density F=2.125, p=0.153 F=1.834, p=0.183 F=0.078, p=0.781 F=4.947, p=0.032 0.072 Patch size (log) F=0.608, p=0.440 F=0.281, p=0.599 F=0.391, p=0.536 F=12.616, p=0.001 0.209 Native canopy cover Fire height (sqrt) F=3.098, p=0.086 F=0.358, p=0.553 F=0.219, p=0.642 F=6.325, p=0.016 0.099 Local road cover (log) Variable violates homogeneity of variances (See Supplementary Table 4.11) Patch size (log) F=0.050, p=0.824 F=0.808, p=0.374 F=0.394, p=0.534 F=14.316, p=0.001 0.237 Native shrub cover (arcsin) Average annual rainfall F=0.012, p=0.913 F=0.276, p=0.602 F=0.000, p=0.990 F=2.019, p=0.163 Not sig. Altitude (log) F=0.139, p=0.711 F=1.229, p=0.744 F=0.040, p=0.843 F=2.397, p=0.130 Not sig. Tree richness Rainfall variability F=0.818, p=0.372 F=0.010, p=0.920 F=1.189, p=0.282 F=5.405, p=0.025 0.101 Local road cover (log) Variable violates homogeneity of variances (See Supplementary Table 4.11) Altitude (log) F=0.983, p=0.328 F=1.167, p=0.286 F=0.034, p=0.854 F=20.094, p<0.001 0.341 Forbs other richness Patch size (log) F=0.571, p=0.455 F=2.896, p=0.097 F=2.922, p=0.095 F=9.101, p=0.004 0.191 Temp. variability F=13.030, p=0.001 Linear regressions

Attribute Significant variable Subregion Β value t-test Significance Adjusted R² Average annual Brisbane β = -0.035 t(22) = -0.166 p =0.870 Not significant No. of large trees temp. Gympie β = 0.745 t(16) = -4.466 p<0.001 R² = 0.527, F(1, 16) = 19.994 Brisbane β = 0.272 t(22) = 1.326 p =0.198 Not significant CWD (log) Patch size (log) Gympie β = 0.679 t(16) = 3.702 p=0.002 R² = 0.428, F(1, 16) = 13.704 Brisbane β = 0.642 t(22) 3.930 p =0.001 R² = 0.386, F(1, 22) = 15.444 Forbs other richness Temp. variability Gympie β = -0.308 t(16) = -1.294 p=0.214 Not significant

There were significant negative relationships between the number of large trees, plant available water and average annual temperature. However, the relationship with average annual temperature was restricted to the Gympie subregion (Table 4.2 and 4.3, and Figure 4.1). Canopy cover was significantly reduced with increased fire height for both Brisbane and Gympie sites (Table 4.2 and Figure 4.1). There was a significant positive relationship between tree species richness and rainfall variability. This relationship was evident for the Gympie subregion, particularly between sites with the lowest and moderately variable rainfall (Table 4.2 and Figure 4.1).

Figure 4.1: Error bar plots of tree-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green).

There was a significant negative relationship between native shrub cover and patch size for both the Brisbane and Gympie subregions, particularly between small/medium and large patches (Table 4.2 and Figure 4.2).

Figure 4.2: Error bar plots of shrub-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green).

100

There was a significant negative relationship between weed cover and patch size for both subregions (Table 4.2 and Figure 4.3). Similarly, weed cover decreased with higher bulk density values for Brisbane and Gympie (Table 4.2 and Figure 4.3). However, weed cover increased in areas with more rural residential development (Table 4.2 and 4.3).

Figure 4.3: Error bar plots of weed-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green).

101

The quantity of coarse woody debris increased significantly with patch size, but was restricted to the Gympie subregion (Table 4.3 and Figure 4.4). Similarly, forbs and other species richness increased with patch size and altitude. Although the relationships between forbs, patch size and altitude were significant across the study area, error bar plots suggested that they were confined to Brisbane sites (Table 4.2 and Figure 4.4). There was a significant positive relationship between forbs and other species richness and temperature variability for the Brisbane subregion (Table 4.3 and Figure 4.4).

Figure 4.4: Error bar plots of ground-related vegetation attributes and significant disturbance and environmental variables for the RE 12.11.5e Brisbane subregion (blue) and Gympie subregion (green).

102

4.3.2 Relationships between vegetation condition attributes, environmental gradients and disturbance regimes in scribbly gum woodland (RE 12.5.3)

Descriptive statistics for RE 12.5.3 sites are provided in Supplementary Table 4.3. Patch size ranged between 1.70 ha and 13501.07 ha for RE 12.5.3, although this was considerably less compared to values for RE 12.11.5e. Mean (2140.75 ha) and median (178.88 ha) RE 12.5.3 patch size values were also markedly lower than those exhibited by RE 12.11.5e (12560.72 ha mean patch size, 4038.98 ha median patch size). The number of large trees, litter cover, native shrub cover, tree richness and native forb richness were similar to RE 12.11.5e values. Canopy height, coarse woody debris, native grass cover, weed cover, native canopy cover, canopy richness and native grass richness were all lower for RE 12.5.3 sites compared to RE 12.11.5e. However, RE 12.5.3 native shrub richness was higher relative to RE 12.11.5e values.

Hierarchical partitioning values, including Z scores and p-values for all vegetation attributes, and percent variation for significant disturbance (including patch size as an indicator of fragmentation) and environmental (including northing as a measurement of latitudinal gradients) variables for RE 12.5.3 are provided in Table 4.4. There were fewer significant attributes for RE 12.5.3 compared to RE 12.11.5e. This result may be due to a reduced sample size of 20 sites for RE 12.5.3 compared to a total of 42 sites for RE 12.11.5e.

103

Northing Patch Disturbance Environmental /Latit Size ude Attribute Northing Patch Fire Prop. Fire Logging Local Aspect Altitude Rainfall Temp. Bulk Plant /Latitude Size (PS) (FP) Height (L) Road (ASP) (ALT) Variability Variability Density Available (N) (FH) Cover (LR) (RCV) (TCV) (BD) Water (PAW) Number of Z=0.479 Z=1.300 Z=1.765 Z=0.338 Z=0.564 Z= - 0.505 Z=0.820 Z= - 1.729 Z= - 1.096 Z=0.886 Z=0.345 Z= - 2.376 large trees p = 0.649 p=0.194 p=0.077 p=0.743 p=0.588 p=0.621 p=0.422 p 0.085 p=0.289 p=0.385 p=0.741 p=0.014 %v=47.76 Canopy Z=1.350 Z=1.675 Z=1.952 Z=1.546 Z=0.935 Z= - 0.927 Z=0.131 Z=1.161 Z= - 0.868 Z=0.443 Z=0.627 Z= - 0.425 height p=0.183 p=0.093 p=0.045 p=0.128 p=0.354 p=0.365 p=0.896 p=0.252 p=0.399 p=0.664 p=0.549 p=0.683 %v=67.38 CWD Z=-0.931 Z=-0.416 Z=0.755 Z= - 1.025 Z=0.232 Z= - 1.127 Z=0.190 Z= - 0.341 Z=1.079 Z=0.774 Z= - 0.261 Z=1.359 p=0.055 p=0.690 p=0.479 p=0.317 p=0.821 p=0.266 p 0.858 p=0.736 p=0.292 p=0.451 p=0.800 p=0.188

Native grass Z=0.426 Z=1.407 Z=1.193 Z=1.831 Z= - 0.461 Z= - 0.518 Z=0.026 Z=1.334 Z=1.178 Z= - 0.115 Z=0.596 Z= - 0.061 cover p=0.680 p=0.164 p=0.249 p=0.069 p=0.654 p=0.627 p=0.982 p=0.183 p=0.246 p=0.911 p=0.567 p=0.953

Litter Z=-0.190 Z=-0.941 Z=0.121 Z= - 0.919 Z=1.478 Z=0.190 Z= - 1.337 Z= - 0.649 Z= - 1.102 Z=1.160 Z= - 0.04 6 Z= - 1.343 cover p=0.859 p=0.351 p 0.915 p=0.379 p=0.140 p=0.856 p=0.192 p=0.525 p=0.277 p=0.266 p=0.964 p=0.185

104

Attribute Northing Patch Fire Prop. Fire Logging Local Aspect Altitude Rainfall Temp. Bulk Plant /Latitu Size (FP) Height (L) Road (ASP) (ALT) Variabili Variabili Densit Available de(N) (PS) (FH) Cover ty (RCV) ty (TCV) y (BD) Water (LR) (PAW) Weed cover Z=-0.992 Z=-2.080 Z= - 0.781 Z= - 0.068 Z=1.568 Z=1.3 51 Z=0.742 Z=0.599 Z=0.630 Z=1.552 Z= - 1.313 Z= - 0.156 p=0.332 p=0.034 p=0.442 p=0.954 p=0.118 p=0.180 p=0.472 p=0.560 p=0.539 p=0.119 p=0.190 p=0.877 %v=3.72 Native Z=0.9082 Z=1.020 Z=0.581 Z=0.350 Z=1.512 Z= - 1.016 Z=1.829 Z= - 0.180 Z= - 1.158 Z=1.875 Z=0.036 Z= - 1.813 canopy cover p=0.378 p=0.312 p=0.583 p=0.745 p=0.138 p=0.319 p=0.071 p=0.858 p=0.256 p=0.063 p=0.974 p=0.069

Native shrub Z=1.8918 Z=2.682 Z=0.121 Z=0.033 Z= - 1.551 Z= - 1.049 Z= - 0.819 Z= - 0.013 Z= - 1.879 Z= - 0.098 Z=0.336 Z= - 1.547 cover p=0.055 p=0.004 p=0.910 p=0.977 p=0.124 p=0.305 p=0.418 p=0.990 p=0.058 p=0.929 p=0.742 p=0.128 %v=46.97 Canopy Z=-0.713 Z= - 1.482 Z= - 0.003 Z=0.910 Z=1.458 Z= - 0.238 Z= - 0.070 Z=1.323 Z=0.167 Z=2.822 Z= - 0.925 Z= - 1.350 richness p=0.493 p=0.145 p = 1.000 p=0.374 p=0.146 p=0.815 p=0.943 p=0.195 p=0.868 p=0.003 p=0.367 p=0.186 %v=67.00 Tree richness Z=-0.449 Z=-1.816 Z=0.761 Z=0.685 Z=1.763 Z=0.304 Z=0.311 Z=0.448 Z=0.759 Z=2.456 Z= - 0.110 Z=0.707 p=0.152 p=0.069 p=0.457 p=0.513 p=0.079 p=0.767 p=0.759 p=0.661 p=0.460 p=0.013 p=0.917 p=0.496 %v=69.69 Shrub Z=2.429 Z=2.2 06 Z=0.815 Z=2.295 Z= - 0.845 Z= - 0.599 Z= - 0.384 Z=2.4 88 Z= - 1.087 Z= - 0.036 Z= - 1.202 Z= - 0.319 richness p=0.013 p=0.024 p=0.433 p=0.017 p=0.409 p=0.557 p=0.713 p=0.011 p=0.290 p=0.975 p=0.233 p=0.763 %v=34.04 %v=8.50 %v=23.97 %v=33.49 Grass Z=-1.906 Z= - 0.591 Z= - 0.668 Z= - 1.285 Z= - 0.547 Z= - 1.308 Z= - 0.777 Z= - 0.598 Z=2.195 Z=1.258 Z=1.393 Z= - 0.555 richness p=0.054 p=0.565 p=0.550 p=0.211 p=0.597 p=0.202 p=0.452 p=0.561 p=0.023 p=0.210 p=0.169 p=0.590 %v=54.61 Forbs other Z=-0.942 Z=0.131 Z= - 0.759 Z= - 1.413 Z= - 0.447 Z= - 0.767 Z=1.401 Z=0.184 Z=1.611 Z=0.800 Z=0.578 Z= - 0.760 richness p=0.361 p=0.894 p=0.489 p=0.168 p=0.663 p=0.454 p=0.177 p=0.859 p=0.108 p=0.441 p 0.580 p=0.458

105

Shapiro-Wilk tests for normality for untransformed and transformed RE 12.5.3 vegetation attributes and significant variables identified through hierarchical partitioning are provided in Supplementary Tables 4.12 to 4.15. Pearson’s correlations for normally distributed data and Spearman rank correlations for non- normally distributed data for RE 12.5.3 vegetation attributes and significant variables identified through hierarchical partitioning are provided in Table 4.5.

Attribute Significant Correlation Degrees of Correlation Significance variable type freedom coefficient No. of large PAW (log) Spearman df=18 rs= -0.545 p=0.013 trees Canopy Fire Spearman df=18 rs= 0.448 p=0.048 height proportion Weed Patch size Spearman df=18 rs= -0.477 p=0.033 cover (log) Native Patch size Spearman df=18 rs= 0.615 p=0.004 shrub cover (log) Canopy Temp. Spearman df=18 rs= 0.647 p=0.002 richness variability Tree Temp. Spearman df=18 rs= 0.564 p=0.010 richness variability Northing Pearson df=18 rs= 0.620 p=0.004

Fire height Spearman df=18 rs= 0.526 p=0.017 Shrub richness Patch size Pearson df=18 rs= 0.533 p=0.016 (log) Altitude Pearson df=18 rs= 0.400 p=0.084 (sqrt) Grass Rainfall Pearson df=18 rs= 0.602 p=0.005 richness variability

106

There was a significant negative relationship between the number of large trees and plant available water (PAW) (Table 4.5 and Figure 4.5). There were significant positive relationships between canopy species richness, tree species richness and temperature variability. Although variation among canopy species richness was evident for low/moderate and high variability sites, the error bar plot indicated a weaker relationship for trees (Table 4.5 and Figure 4.5). A Spearman rank correlation demonstrated that canopy height increased with increased fire proportion (Table 4.5 and Figure 4.5).

Figure 4.5: Error bar plots for tree-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3.

107

There was a significant increase in shrub cover and shrub species richness with patch size for RE 12.5.3 (Table 4.5 and Figure 4.6). There was a significant relationship between shrub species richness and latitude, particularly between northern, central north and southern sites (Table 4.5 and Figure 4.6). A Spearman rank correlation suggested that there was a significant increase in shrub species richness for moderate to high fire heights. However, this result was not reflected in the error bar plot. A Pearson’s correlation indicated that the relationship between shrub species richness and altitude was not significant for RE 12.5.3.

Figure 4.6: Error bar plots for shrub-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3.

108

There was a significant negative relationship between weed cover and patch size (Table 4.5 and Figure 4.7). A Pearson’s correlation indicated that there was a significant increase in grass species richness with more variable rainfall (Table 4.5 and Figure 4.7).

Figure 4.7: Error bar plots for weed- and ground-related vegetation attributes and significant disturbance and environmental variables for RE 12.5.3.

109

4.4 Discussion

The composition, abundance and structure of native and non-native vegetation underpin BioCondition vegetation attributes. It was shown in this component of the study that vegetation condition attributes were significantly affected by environmental gradients and disturbance regimes, and that their responses to variability broadly corresponded with those exhibited by the vegetation compositional and structural data in Chapter 3.

4.4.1 Effects of environmental gradients on vegetation condition attributes

Environmental variables linked to plant-water balance were important for tree- related attributes. Reduced numbers of large trees in RE 12.11.5e were linked to high plant available water. Mesic conditions and competition from native and introduced species adapted to wetter substrates excluded many eucalypts from spotted gum sites. The dominant canopy species Corymbia citriodora is found on well-drained, light or skeletal soils on ridges and steep hills (EUCLID 2014b; Florabank 2014). It has been demonstrated that the roots of large Eucalyptus and Corymbia species can move into deeper than expected substratum layers (Eberbach and Burrows 2006; Falkiner et al. 2006). Although high soil moisture restricted eucalypt distribution, the stony ridges defining RE 12.11.5e may store water deeper in the soil profile, effectively sustaining remnant vegetation where subsurface water levels are low.

Elevated soil moisture was also linked to reduced numbers of large eucalypts for RE 12.5.3. Adverse environmental gradients had a direct effect on native vegetation. Furthermore, damp soils may have exacerbated fire exclusion in fragmented scribbly gum forests. Under these conditions open woodland communities dominated by Eucalyptus racemosa developed high canopy covers, with upper strata defined by smaller species (e.g. Allocasuarina littoralis and Acacia disparrima) usually associated with shrub layers (Figure 4.8).

110

Figure 4. 8: Photograph of small, fire-excluded RE 12.5.3 patch (1253_15 Korawal Street) with high plant available water and a canopy dominated by Allocasuarina littoralis and Acacia disparrima.

Conversely, climatic variables linked to thermal stress and water deficits were negatively related to the number of large trees for RE 12.11.5e, and had a significant impact on native tree species richness for both regional ecosystems. Higher temperatures in the Gympie subregion potentially increased seedling mortality, preventing recruitment in microhabitats suitable for mature vegetation (Osmond et al. 1987), and restricting the distribution and dominance of some Eucalyptus and Corymbia species. However, it was demonstrated in this study that rainfall and temperature variability, which can also be a threat to species persistence (Boyce 1992; Menges 2000), were associated with increased tree and canopy species richness for RE 12.11.5e and RE 12.5.3.

111

Rainfall variability can affect growth rates and survival through regulating soil moisture, supressing or promoting competition and triggering reproductive or rapid growth events (Mitchell and Csillag 2001; Levine et al. 2008; Travers and Eldridge 2013; Medina et al. 2014), effectively stabilising or destabilising species interactions over time (Adler et al. 2006). Many studies have shown that highly variable rainfall is detrimental for vegetation growth and may reduce species richness (Hoffman and Jackson 2000; Ni and Zhang 2000: McAlpine et al. 2009; Giesecke et al. 2010; Alvarez-Cansino et al. 2013). However, the positive relationship between variable rainfall and large tree species richness observed in the current study may be explained by processes attributed to coexistence theory (Chesson and Huntley 1989), the storage effect and compensatory dynamics (Chesson 2000; Gonzalez and Loreau 2009).

Coexistence theory and the storage effect propose that competing species can coexist within a community if there is a diversity of asynchronous responses to climatic variability, ensuring intraspecific competition under challenging conditions and interspecific competition during favourable times (Chesson 2000; Higgins et al. 2000; Gardner 2006). Variability effectively stabilises competition by increasing the number of available temporal niches. Competing organisms need to be long-lived in order to effectively store reproductive potential and bolster populations during adverse climatic periods. Subsequently, compensatory dynamics dictate that rainfall and temperature variability may result in the formation of diverse and environmentally partitioned floristic assemblages through the maintenance of total community density or biomass (Gonzalez and De Feo 2007; Ives and Carpenter 2007; Ranta et al. 2008; Gonzalez and Loreau 2009). The canopies and subcanopies of RE 12.11.5e are dominated by Myrtaceae species. Large Eucalyptus, Corymbia and Lophostemon trees possess life history traits, including longevity and the ability to resprout following damage or disturbance (Benson and McDougall 1998; Burrows 2002; Mathews and Bonser 2005) that can enable the storage of reproductive potential across generations.

112

Similarly, the relationship between increased canopy species richness and temperature variability suggested that similar processes may also be operating in RE 12.5.3. However, many scribbly gum sites associated with thermal variability were described by woodland rather than heathland understoreys. Woodland species are found on deeper soils of moderate to high fertility, and heath species are primarily associated with shallow, less productive soils (Specht and Specht 1989b). Although not measured in the current study, patterns suggest that edaphic conditions may underpin variation in RE 12.5.3 tree species richness independent of climatic factors.

4.4.2 Effects of disturbance regimes on vegetation condition attributes

Disturbance factors did influence tree-related condition attributes. High and presumably intense fires reduced RE 12.11.5e canopy cover in the Gympie subregion, opening up dominant tree layers through significant post-fire mortality (Cochrane and Schulze 1999; Knox and Clarke 2012). Single or ongoing events of relative severity have the potential to cause pyrogenic feedbacks, further increasing the flammability of previously burnt forest through increased light penetration and ground fuel desiccation (Knox and Clarke 2012). However, the increase in RE 12.5.3 canopy height with fire proportion illustrated the effects of fire exclusion rather than gradients of disturbance. The canopy height of a single site (1253_18 New Settlement Road) was shown to be greatly reduced (15 m compared to a mean of 23.7 m and a median of 25 m) where fire proportion was less than one percent, and indicated how small, fire-excluded scribbly gum patches develop thick but low canopy layers dominated by Allocasuarina littoralis to the detriment of Eucalyptus racemosa.

Shrub and ground strata were significantly altered by disturbance processes linked to human activity. Native shrub cover responded to landscape fragmentation across RE 12.11.5e and RE 12.5.3 sites. However, responses were specific to each regional

113 ecosystem, with native shrub cover increasing in small RE 12.11.5e patches and decreasing in corresponding RE 12.5.3 patches. This disparity suggests that distinctive vegetation characteristics and associated ecological processes determine how ecosystems change under similar types of impact. Small patches were often associated with limited management activity, and were defined by fire exclusion or suppression, particularly within urban and peri-urban landscapes (Ross et al. 2002; Close et al. 2009). Long unburnt spotted gum forest was shown to transition from open forest to communities dominated by dry rainforest species. Although light penetration was reduced, shade-tolerant native plants were able to dominate and increase their cover relative to shrub strata under open eucalypt canopies. In comparison, when RE 12.5.3 was subjected to equivalent reductions in burning Eucalyptus racemosa was replaced by Allocasuarina littoralis as the dominant canopy species. Increased shading and high litter accumulation, perhaps combined with low soil fertility, resulted in decreased native shrub cover and reduced shrub species richness (Specht and Specht 1989b; Lunt 1998; Close 2009). The results indicate that native shrub cover is a significant, cross-community indicator of landscape-level impacts, and as such, an important component of terrestrial ecological condition metrics.

Causes of vegetation change were not always well-defined. Although higher latitudes were linked to increased native shrub species richness for RE 12.5.3, the relationship was determined by fragmentation processes rather than climatic gradients. Large patches, dominated by species-rich heathland understoreys, were found mainly in the north, with smaller, southern fragments containing depauperate shrub layers.

Fragmentation of native vegetation has a profound effect on the presence and abundance of invasive species. This effect is heavily driven by the interplay between landscape spatial structure (Saunders et al. 1991; Fahrig and Merriam 1994; Bradshaw 2012), the form and evolution of surrounding land use (Gill and Williams

114

1996; Forman and Alexander 1998; Hansen et al. 2005; Catford et al. 2011), and changes in the composition, structure and function of remaining native vegetation (Ross et al. 2002; Catford et al. 2011). Increased weed cover among small RE 12.11.5e and RE 12.5.3 patches, and with rural residential development, was indicative of these processes. Heightened edge effects and the subsequent increase in light penetration, altered wind profiles, and variable soil moisture and nutrient levels facilitate the establishment of invasive plant species within small patches (Theoharides and Dukes 2007; Prober and Wiehl 2012), as does the increase in exotic propagule pressure driven by land use intensification (Hansen et al. 2005). It has been suggested that rural residential development affects native plant and animal species through its proximity to intact, remnant vegetation, and its longevity within the landscape when compared to other non-urban land uses (Hansen et al. 2005). Historical processes underpin these impacts where productive land appropriated for private agriculture becomes interspersed with protected, less fertile areas (Gude et al. 2006). Farmland is often subdivided for low-density housing. Unlike grazing practices which often incorporate soil or pasture rest periods, the effects of rural residential development are comparatively permanent and ongoing, and may facilitate further land use intensification and weed dispersal within the landscape (Hansen et al. 2005).

In addition to landscape fragmentation, the establishment and persistence of exotic plant species may also be determined by favourable environmental conditions. Degraded and fragmented habitats are generally characterised by disturbed soil profiles (Lindsay and French 2005). The relationship between low bulk density values and high weed cover suggested that increased soil porosity and resource availability promoted weed growth across affected RE 12.11.5e sites (Davis et al. 2000; Kyle et al. 2007). Once established, weeds may alter soil properties, reducing compaction through the increase of organic matter, nutrient loading and other transformative mechanisms (Sperber et al. 2003; Lindsay and French 2005). Ultimately, natural variation in the physical and chemical components of soils supporting intact native vegetation may exclude or reduce invasive species

115 numbers where bulk density is high. Spotted gum sites with thin, rocky substrates and low fertility potentially favour well adapted native plants to the detriment of exotic species, particularly where human impacts are reduced.

The structure and composition of ground strata were also affected by disturbance processes. Small RE 12.11.5e patches in the Gympie subregion retained or produced less coarse woody debris, potentially due to altered fire regimes, timber collection and edge effects (Gould et al. 2008) (Figure 4.9). Reduced fire management in urban and peri-urban areas can result in hot, sporadic wildfires often depleting coarse woody debris greater than 10 cm in diameter (Woldendorp et al. 2002). Additionally, large spotted gum remnants in the Gympie subregion were historically and currently managed as state forests. Elevated levels of coarse woody debris were generated where defective timber was left on-site due to the demand for high quality logs (Harmon et al. 1986; Woldendorp and Keenan 2005; Collins et al. 2012) (Figure 4.9).

116

Ground species were susceptible to the effects of disturbance and found in reduced numbers among small, isolated RE 12.11.5e patches. This response to fragmentation conforms to species-area relationships (Arrhenius 1921) and island biogeography theory (MacArthur 1967), and suggests that landscape fragmentation is detrimental to forb dispersal and species persistence in small remnants over time (Soons et al. 2005; Alados et al. 2008). Although forb species richness was shown to be sensitive to anthropogenic impacts, it was positively related to climatic variability. This result parallels those patterns observed for RE 12.5.3 grass species richness, and may indicate that variable temperature and rainfall promote understorey diversity through temporal niche partitioning (Adler et al. 2006; Letten et al. 2013).

4.5 Conclusion

Results of this study indicate that environmental gradients and disturbance regimes influenced vegetation condition attributes. Although the effects of variability were often unique to each regional ecosystem, broad patterns, and sometimes specific relationships were shared across vegetation communities. Soil moisture, fire exclusion and latitudinal effects were important for tree strata. Disturbance processes related to fragmentation were significant for shrub and ground layers with weed and shrub cover responding to changes in patch size.

Relationships were also shared across condition attribute and vegetation datasets, demonstrating that vegetation composition, structure and function underpin multimetric condition indices. The effects of plant available water were significant for RE 12.11.5e tree-related attributes and measurements of large tree composition and abundance. Similarly, latitudinally-related climatic gradients were important for tree layers across RE 12.11.5e datasets, as were their decreasing influence on understorey layers and the corresponding increase in the impacts of landscape- level fragmentation on shrub and ground strata. Plant available water and

117 latitudinal effects were also important for RE 12.5.3 attribute and floristic datasets. However, unlike RE 12.11.5e, latitude was related to the geographical position of heathland and woodland/open forest, and disturbance processes linked to landscape fragmentation.

The significant relationships between condition attributes, vegetation composition and structure, and climate, environmental gradients and disturbance processes has certain implications for metric application and design. Reference condition approaches such as BioCondition may effectively account for this variability through strategic reference site selection. Climatic effects may be absorbed through subregional benchmarking. Environmental gradients and disturbance regimes may be mitigated by locating reference sites based on significant associations between vegetation, soil, geology and landform, and historical and current patterns of land use, management and disturbance. The implications for multimetric condition assessment will be examined in further detail in Chapter 6.

118

Chapter 5: Testing the efficacy of BioCondition as an ecological surrogate—comparative analysis with a novel, independent acoustic index of ecological health in spotted gum open forest (RE 12.11.5e)

and scribbly gum woodland (RE 12.5.3)

5.1 Introduction

Condition metrics, as a subset of ecological surrogates, are based upon an inherent trade-off between utility and scientific rigour. Although measurements of ecological condition have to be easily interpreted and understood, they must also accurately represent environmental change. The assessment of surrogate efficacy can often be constrained by complex ecological relationships and the large number of species conceivably present. However, the capacity for surrogates to reflect target biodiversity may be determined by measuring the level of agreement between the geographical cover, presence or arrangement of environmental surrogates (e.g. land classes, vegetation groups, habitat components) or the distributions of biodiversity indicators, with the distributions and diversity of target species (Lombard et al. 2003; Warman et al. 2004; Barton et al. 2014; Pierson et al. 2015).

Previous research suggests that multimetric ecological condition approaches perform better for some biological groups than for others. Oliver et al. (2014) examined the relationships between condition scores derived from several multimetric approaches and species-level biodiversity, and found that vascular plants and birds were the only taxa, out of a total of 11, where species richness was significantly and positively correlated with condition scores. Jansen and Robertson (2005) demonstrated that increased grazing intensity was significantly related to a decline in riparian condition for south-eastern Australian eucalypt forests, corresponding with a decrease in woodland and threatened bird species and tadpole numbers. Furthermore, an assessment of multimetric condition approaches by Weinberg et al. (2008) suggested that although they were generally poor indicators for many vertebrate species, multimetric indices provided reasonable richness estimates for taxa dependent upon structurally diverse vegetation such as woodland birds, arboreal mammals and reptiles. In addition, the direct measurement of habitat quality using indicator species suggests that many taxonomic groups respond strongly to patterns of disturbance. Fragmentation processes and reduced levels of ecological condition are often associated with

121 adverse impacts on vertebrate (Lampila et al. 2005; Shanley et al. 2013) and invertebrate populations (Thomas et al. 2001; Driscoll and Weir 2005; Bauerfeind et al. 2009), and vascular plant species (Tang et al. 2011).

The gathering of sufficient biological data for testing surrogate efficacy can be labour intensive and time consuming. Sampling methods require slow and involved field observation and collection, often by experts (Depraetere et al. 2012). However, advances in acoustic monitoring systems provide a novel alternative to traditional field based methods. The recording and analysis of acoustic signals produced by audible organisms has been used to assess ecosystem health and biodiversity (Sueur et al. 2008; Laiolo 2010; Blumstein et al. 2011; Depraetere et al. 2012; Kuehne et al. 2013; Proppe et al. 2013), and underpins the development of soundscape ecology (Pijanowski et al. 2011a, b). Soundscape ecology is the study of biological (biophony), anthropogenic (anthrophony) and geophysical (geophony) sounds produced from the landscape, their spatial and temporal variation, and their relationships with ecological processes and human activities (Pijanowski et al. 2011a, b).

Acoustic technologies are relatively inexpensive tools for examining biodiversity when compared to traditional field approaches and have shown to be more reliable than field observation and collection alone (Sueur et al. 2008; Celis-Murillo et al. 2009; Wimmer et al. 2013b). Recently developed acoustic indices may be broadly categorised as those that maintain a species or morphospecies focus, including rapid biodiversity appraisal approaches (Rempel et al. 2005; Celis-Murillo et al. 2009) and those that measure the soundscape at a community level (Gage et al. 2001; Qi et al. 2008; Sueur et al. 2008). Although complete species inventories, or representative subsets, are attractive, their collation may be problematic. Acoustic signatures for individual species must first be identified and analyses can be complex and time intensive (Sueur et al. 2008).

122

The use of acoustic indices (e.g. Acoustic Entropy (H), Dissimilarity (ADI) (Sueur et al. 2008), Richness (Depraetere et al. 2012), Acoustic Complexity (ACI) (Pieretti et al. 2011) and Normalised Difference Soundscape Index (NDSI) (Gage et al. 2001; Kasten et al. 2012)) can avoid these difficulties. It has been theorised that communities with more audible species have a greater acoustic diversity, that biodiversity will correlate positively with acoustic diversity, and that in some cases, measures of acoustic diversity may provide an index of ecological health (Gage et al. 2001; Qi et al. 2008). High levels of human generated sound (anthrophony) may impact on the reproductive success and habitat use of fauna (Francis et al. 2011) and the ecological integrity of habitat (Pijanowski et al. 2011b). The relationship between biological sound, or biophony, and ecological condition is founded on the premise that vegetation structure, particularly complexity, influences species richness and therefore biophony (Pijanowski et al. 2011a). For example, natural, rural, peri-urban and urban habitats exhibit a successive reduction in vegetation complexity, paralleled by a similar reduction in natural acoustic diversity (Joo et al. 2011).

In this study, the efficacy of BioCondition as a representative multimetric ecological condition index was assessed by measuring the level of agreement between condition scores and an independent acoustic metric of ecological health. NDSI was selected as the comparative acoustic index because it has the capacity to discriminate between human generated and biological sound, and therefore function effectively in urban and peri-urban landscapes (Joo et al. 2011). NDSI measures the soundscape at a community level, and therefore negates the need to identify individual species. However, in this study it was necessary to establish a connection between NDSI and the number of audible organisms based on the assumption that biodiversity will positively correlate with acoustic diversity. Birds, insects and anurans are the dominant sound producing taxa between 3000 and 11000 Hz. These three taxonomic groups, besides being a major component of the biological soundscape, are also commonly used as more general indicators of environmental health. Birds, insects and anurans are easily observed and sampled

123

(Blair 1999; Ricketts et al. 2002; Pineda et al. 2005; Drever et al. 2008; Favero et al. 2011; Eglington et al. 2012). Birds are varied in habitat selection and life history (Chace and Walsh 2006; Carrascal et al. 2008; Schuster and Arcese 2013), as are insects, which exhibit habitat specificity at lower taxonomic levels and are highly susceptible to habitat change (Rodriguez et al. 1998; Cabrini et al. 2013). Anuran species are considered useful indicators of disturbance and pollution because of their dual lifecycle, semi-permeable skin and microhabitat requirements (Welsh and Ollivier 1998; Price et al. 2007). Birds respond to variation in vegetation composition, structure and condition (Catterall et al. 1998; Mörtberg 2001; Dures and Cumming 2010; Skroblin and Legge 2012), and birds and insects in particular have been shown to serve as useful biodiversity surrogates for unrelated taxa (Pearson and Cassola 1992; Rodriguez et al. 1998; Sauberer et al. 2004; Bazelet and Samways 2012; Eglington et al. 2012; Larsen et al. 2012).

Of the three taxonomic groups, birds were used to determine whether changes in NDSI were related to changes in biodiversity. In addition, measuring the relationship between NDSI and ecological condition scores and attribute values positioned avian species as a representative subset of broader biodiversity with which BioCondition may or may not conform. Birds were selected here because they were found in high numbers, were easily identified and their distributions and diversity are often related to those of other organisms (Sauberer et al. 2004; Eglington et al. 2012; Larsen et al. 2012) and to measurements of ecological condition (Jansen and Robertson 2005; Weinberg et al. 2008; Oliver et al. 2014).

5.1.1 Aims

The aim of this chapter was to test the capacity for BioCondition, as a representative multimetric condition approach, to reflect target biodiversity values by measuring the level of congruence between condition scores, attributes and an independent acoustic metric of ecological health (NDSI). The relationship between

124

NDSI and bird species richness was first examined to establish whether NDSI reflected avian biodiversity. Secondly, the relationship between NDSI and ecological condition was measured to determine the potential for cross-index congruence. Finally, vegetation and landscape condition attributes that were significantly related to soundscape values were identified in order to examine specific ecological relationships between condition and acoustic metrics.

125

5.2 Materials and Methods

5.2.1 Field surveys

5.2.1.1 Site selection Field surveys were conducted across ten selected sites for RE 12.11.5e (Figure 5.1) and nine for RE 12.5.3 (Figure 5.2). RE 12.11.5e sites were located within a 35 km radius of Brisbane so as to minimise climatic (rainfall, temperature) and physiographical (elevation, broad landform patterns) variability. RE 12.5.3 sites ranged over approximately 40 km, and were centred on a region north of Brisbane defined by the Glasshouse Mountains between Burpengary and Beerwah. Sites were selected based on patch size and patch connectivity as described in Chapter 2.4, to encompass landscape variation within the study area.

5.2.1.2 Ecological condition surveys Ecological condition surveys were conducted as per methods detailed in Chapter 2.5. Vegetation and landscape attributes were measured for each site and compared to benchmarks based on reference values from BRIS1 (D’Aguilar National Park) and BRIS4 (Moggill Conservation Park) for RE 12.11.5e and 1253_3 (Caves Road), 1253_4 (Murphys Road) and 1253_5 (Scientific Area 1) for RE 12.5.3.

126

Figure 5.1: Map of RE 12.11.5e sites selected for soundscape recordings (Department of Environment and Heritage Protection 2010).

127

Figure 5.2: Map of RE 12.5.3 sites selected for soundscape recordings (Department of Environment and Heritage Protection 2010).

128

5.2.1.3 Soundscape surveys Soundscape surveys measured the anthropogenic and biological sound at each site. Acoustic measurements were made using Song Meter SM2 recording devices (Wildlife Acoustics 2013), producing recordings in WAV file format. Recordings were written to SD cards based on a programmable schedule and each file was tagged with a location and date-time stamp.

Song Meters were positioned near the centre of each site. Recording devices were deployed for a one month period for 10 selected sites (BRIS1, BRIS2, BRIS3, BRIS4, BRIS6, BRIS17, BRIS18, BRIS19, BRIS20 and BRIS21) between September 1, 2012 and October 4, 2012 for RE 12.11.5e, and for 9 selected sites (1253_3, 1253_4, 1253_5, 1253_12, 1253_13, 1253_14, 1253_18, 1253_19 and 1253_20) between September 10, 2013 and October 16, 2013 for RE 12.5.3. Sites were selected to encompass the range of different available patch sizes and associated patch connectivities for each regional ecosystem, with patch dimensions described in Tables 2.1 and 2.2. Each mono-aural recording was made at 22050 Hz for one minute in length at a recording interval of thirty minutes. Upon collection, the data were transferred to a server at Michigan State University. The audio files were checked for integrity, linked to project metadata (project, habitat, location etc.) and archived in the Remote Environmental Assessment Laboratory’s Digital Acoustics Library System (Kasten et al. 2012). All acoustic files are publicly available and accessible through the Remote Environmental Assessment Laboratory’s website (http://www.real.msu.edu).

Acoustic metrics derived from each recording were based on the computation of Power Spectral Density (PSD) values from recorded sounds (Welch 1967) using MATLAB (2012) code developed by Gage (2008). Bird species richness was calculated through the expert analysis of calls and spectrograms using an online acoustic environmental workbench (http://sensor.mquter.qut.edu.au; Wimmer et al. 2013a). Twenty randomly selected minutes were sampled and analysed from the

129 dawn chorus (between 0430 hours and 0600 hours) over twenty separate days for each site. Recent studies in South-East Queensland forest systems have shown that targeted sampling of recorded bird call data from the dawn chorus detected the highest number of species (Wimmer et al. 2013b).

A summary of site codes, mean NDSI values, bird species richness and BioCondition attributes are provided in Supplementary Table 5.1 for RE 12.11.5e and Supplementary Table 5.2 for RE 12.5.3. A list of identified bird species and their site locations are provided in Supplementary Table 5.3 for RE 12.11.5e and Supplementary Table 5.4 for RE 12.5.3.

5.2.2 Data analyses

A flowchart detailing the key statistical approaches used for Chapter 5 and potential outcomes is provided in Supplementary Figure 5.1.

Acoustic metrics derived from each recording were based on the computation of Power Spectral Density (PSD) values from recorded sounds (Welch 1967). This technique was applied to each 1 kHz interval in a recording, providing ten values (22 kHz/2) for each recording. The PSD values (watts/kHz) were normalised to produce a value between 0 and 1 for each frequency interval (Kasten et al. 2012). Normalised values represent the relative soundscape power (RSP) for each frequency interval in each recording made at each landscape position, enabling comparison of RSP values across recordings. In addition to RSP values, indices based on these values were computed for two soundscape components: biophony (RSP values between 3–11 kHz) and anthrophony (RSP values between 1–2 kHz) (Qi et al. 2008; Pijanowski et al. 2011a). PSD values were converted to the Normalised Difference Soundscape Index (NDSI) (Gage 2001; Kasten et al. 2012) based on these components, where NDSI = [biophony - anthrophony] / [biophony + anthrophony].

130

NDSI values were averaged over a 24 hour period over 25 days of recording. Mean NDSI, bird species richness data (Table 5.1 and Table 5.2) and BioCondition scores (Table 5.4 and Table 5.5) were tested for normality using Shapiro-Wilk tests and homogeneity of variance using Levene’s tests for RE 12.11.5e and RE 12.5.3 sites in SPSS V.21 (IBM Corporation 2012).

Descriptive statistics, including mean and median values, standard deviations, and minimum and maximum values, were generated for bird species richness and mean NDSI values using SPSS V.21 (IBM Corporation 2012). Descriptive statistics are provided in Supplementary Table 5.5 for RE 12.11.5e and RE 12.5.3.

Analyses of covariance (ANCOVAs) were conducted for mean NDSI values (24 hours) as the dependent variable, bird species richness or BioCondition as the covariate and regional ecosystem as the fixed factor using SPSS V.21 (IBM Corporation 2012). Because ANCOVAs are robust to non-normal data, they were used where normality was violated (Barrett 2011). Additionally, if homogeneity of variances was violated and yet the ratio of the largest group variance was not more than three times that of the smallest group, ANCOVAs were also used (Glass et al. 1972). Where ANCOVA homogeneity of regression slopes was not met, and there was a significant interaction between regional ecosystem (fixed factor) and bird species richness (covariate), linear regressions were used to explore the relationships between dependent and independent variables. All analyses were conducted using SPSS V.21 (IBM Corporation 2012). A flow chart summarising these steps is provided in Supplementary Figure 4.1. Scatterplots were generated for mean NDSI and bird species richness for RE 12.11.5e and RE 12.5.3. Similarly, a scatterplot was used to illustrate the relationship between mean NDSI and BioCondition scores.

131

5.2.2.1 Bird species composition UPGMA analysis was completed for presence/absence bird data across all sites using PATN version 3 for Windows (Belbin 2008). Raw presence/absence data remained untransformed prior to analyses. The Bray-Curtis association measure (Bray and Curtis 1957) was used to quantify the dissimilarity between sites. A two- step function was used to explore the relationship between bird species composition and sites and two-way tables were produced as a result. Data classification was carried out using agglomerative hierarchical classification and flexible unweighted pair group arithmetic averaging (UPGMA). The species lists generated by Monte-Carlo attributes in an ordination (MCAO) were examined for those species that were statistically significant (MCAO values ≤1%).

5.2.2.2 BioCondition vegetation and landscape attributes and NDSI BioCondition attributes were used as predictor variables for the hierarchical partitioning of mean NDSI (24 hours) in RE 12.11.5e and RE 12.5.3. Predictor variables and mean NDSI (24 hours) values were standardised using the following formula: Z=(X-μ)/σ, where Z is the standardised value, X is the unstandardized value, μ is the mean, and σ is the standard deviation.

Standardised predictor variables (BioCondition attributes) and response variables (mean NDSI values (24 hours)) were subjected to hierarchical partitioning using the R package: hier.part (Walsh and MacNally 2013). A randomisation approach using Monte Carlo permutation tests was used to compute Z-scores and the statistical significance (p<0.05) of each disturbance variable for RE 12.11.5e (MacNally 2002b) using the R package: coin (Hothorn et al. 2014). This procedure was repeated for RE 12.5.3.

BioCondition attributes identified through hierarchical partitioning were tested for normality using the Shapiro-Wilk test and homogeneity of variance using Levene’s

132 test for RE 12.11.5e and RE 12.5.3 sites in SPSS V.21 (IBM Corporation 2012). Area data and cover data that violated normality were log and arcsine transformed respectively. Transformed BioCondition attribute values were then tested for normality and homogeneity of variance.

ANCOVAs and linear regressions were used to explore the relationships between NDSI and significant BioCondition attributes identified through hierarchical partitioning for RE 12.11.5e and RE 12.5.3. Analyses of covariance (ANCOVAs) were conducted on either untransformed or transformed BioCondition data based on the results of the Shapiro-Wilk and Levene’s tests using SPSS V.21 (IBM Corporation 2012). ANCOVAs were performed for mean NDSI (24 hours) and all vegetation and landscape attributes, with mean NDSI (24 hours) serving as the dependent variable, attributes as the covariate and regional ecosystem as the fixed factor. Because ANCOVAs are robust to non-normal data (Barrett 2011), they were used where normality was violated. Where there was a significant interaction between the fixed factor and the covariate, linear regressions were performed on each regional ecosystem separately. Scatterplots of mean NDSI values (24 hours) and significant BioCondition attributes were generated using SPSS V.21 (IBM Corporation 2012).

133

5.3 Results

5.3.1 Descriptive statistics

Bird species richness ranged from 15 to 48 species for RE 12.11.5e and 11 to 36 for RE 12.5.3. Mean species richness was higher for RE 12.11.5e (30) compared to 26 species for RE 12.5.3. NDSI values ranged from -0.13 to 0.65 for RE 12.11.5e and from -0.37 to 0.60 for RE 12.5.3. Average NDSI was higher for RE 12.11.5e (0.35) than for RE 12.5.3 (0.27), although median NDSI for RE 12.11.5e (0.41) was lower than that for RE 12.5.3 (0.49). Results are summarised in Supplementary Table 5.5.

5.3.2 NDSI and bird species richness

5.3.2.1 Normality and homogeneity of variance tests NDSI was normally distributed for RE 12.11.5e, but not for RE 12.5.3. Bird species richness was normally distributed for both regional ecosystems. NDSI and bird species richness exhibited homogeneity of variance for RE 12.11.5e and RE 12.5.3.

Table 5.1: Shapiro-Wilk tests for mean NDSI and bird species richness for RE 12.11.5e and RE 12.5.3.

Variable RE Statistic df Significance

NDSI 12.11.5e 0.878 10 0.122

12.5.3 0.817 9 0.032

Bird species 12.11.5e 0.945 10 0.607 richness 12.5.3 0.935 9 0.533

134

Table 5.2: Levene’s tests for mean NDSI and bird species richness for RE 12.11.5e and RE 12.5.3.

Variable Levene statistic df 1 df 2 Significance

NDSI 0.610 1 17 0.445

Bird species 0.788 1 17 0.387 richness

5.3.2.2 ANCOVAs ANCOVA homogeneity of regression slopes was met, F(1, 15)=1.190, p=0.293. A Levene’s test indicated that the data exhibited homogeneity of variance, F(1, 17)=1.272, p=0.275. There was no significant effect of regional ecosystem on mean NDSI (24 hours) after controlling for the effect of bird species richness, F(1, 16)=0.030, p=0.865. Bird species richness was significantly related to mean NDSI (24 hours), F(1, 16)=6.951, p=0.018. The model had an adjusted R²=0.239.

5.3.2.3 Scatterplot There was a positive relationship between mean NDSI (24 hours) and bird species richness for RE 12.11.5e (R²=0.311, N=10) and RE 12.5.3 (R²=0.377, N=9) (Figure 5.3).

135

Figure 5.3: Scatterplot of Mean NDSI (24 hours) against bird species richness for RE 12.11.5e (blue) and RE 12.5.3 (green).

5.3.2.4 UPGMA analysis of bird species composition UPGMA analysis of bird species composition (Figure 5.4 and Table 5.3) indicated that sites broadly clustered based on regional ecosystem and landscape attributes. RE 12.11.5e and RE 12.5.3 sites separated due to the absence of several urban- adapted species in medium-sized scribbly gum sites. Small, isolated remnants of both regional ecosystems were defined by a suite of urban-adapted birds including the Noisy Miner, Noisy Friarbird, Grey Shrike-thrush, Grey Butcherbird and Blue- faced Honeyeater, and an absence of bushland-dependent species such as the Scarlet Honeyeater, Grey Fantail, White-throated Treecreeper and Rufous Whistler.

136

Figure 5.4: UPGMA of bird species composition for all sites across RE 12.11.5e and RE 12.5.3.

137

Table 5.3: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of bird species presence/absence for RE 12.11.5e and RE 12.5.3.

Cluster Description Sites Important species

1 Large and BRIS1, BRIS4,  Presence of : small/medium BRIS6, BRIS3, . Noisy Miner sized RE 12.11.5e BRIS17, BRIS18 . Noisy Friarbird patches and BRIS19 . Grey Shrike-thrush . Scarlet Honeyeater . Red-browed Finch . Grey Fantail . Black-faced Cuckoo-shrike . Pale-headed Rosella . Galah (large patches) . Leaden Flycatcher (large patches) . Lewin’s Honeyeater (large patches) . White-throated Treecreeper (large patches) . Rufous Whistler (large patches)

2 Small, isolated 1253_20  Presence of : site with an avian . Noisy Miner assemblage . Noisy Friarbird consisting of . Scarlet Honeyeater many ubiquitous . Red-browed Finch species and birds . Lewin’s Honeyeater . White-throated Treecreeper associated with . Blue-faced Honeyeater urban and peri- urban areas

3 Predominantly BRIS2, 1253_14,  Presence of : large and 1253_13, . Grey Shrike-thrush medium sized RE 1253_3, 1253_4, . Galah 12.5.3 patches 1253_5 and . Leaden Flycatcher 1253_12 . Lewin’s Honeyeater . White-throated Treecreeper . Rufous Whistler

 Absence of : . Noisy Miner . Noisy Friarbird . Black-faced Cuckoo-shrike . Pale-headed Rosella

138

Cluster Description Sites Important species

4 Small, isolated RE BRIS20 and  Presence of : 12.11.5e sites BRIS21 . Noisy Miner . Noisy Friarbird . Grey Shrike-thrush . Grey Butcherbird . Blue-faced Honeyeater

 Absence of : . Scarlet Honeyeater . Red-browed Finch . Grey Fantail . Galah . Leaden Flycatcher . Lewin’s Honeyeater . White-throated Treecreeper . Rufous Whistler

5 Small, isolated RE BRIS18 and  Presence of : 12.5.3 sites BRIS19 . Blue-faced Honeyeater . Magpie-lark . Willie-wagtail . Masked Lapwing . Buff-banded Rail . Wandering Whistling-Duck . Channel-billed Cuckoo . Superb Fairy-wren

 Absence of : . Scarlet Honeyeater . Red-browed Finch . Grey Fantail . Galah . Leaden Flycatcher . Lewin’s Honeyeater . White-throated Treecreeper . Rufous Whistler

139

5.3.3 BioCondition and NDSI

5.3.3.1Normality and homogeneity of variance tests BioCondition scores were normally distributed and exhibited homogeneity of variance for RE 12.11.5e and RE 12.5.3.

Table 5.4: Shapiro-Wilk tests for BioCondition scores for RE 12.11.5e and RE 12.5.3.

Variable RE Statistic df Significance

BioCondition score 12.11.5e 0.907 10 0.259

12.5.3 0.892 9 0.210

Table 5.5: Levene’s test for BioCondition scores for RE 12.11.5e and RE 12.5.3.

Variable Levene statistic df 1 df 2 Significance

BioCondition score 0.027 1 17 0.873

5.3.3.2 ANCOVAs ANCOVA homogeneity of regression slopes was met, F(1, 15)=1.194, p=0.292. Levene’s test indicated that the data exhibited homogeneity of variance, F(1, 17)=0.433, p=0.519. There was no significant effect of regional ecosystem on mean NDSI (24 hours) after controlling for the effect of BioCondition score, F(1, 16)=2.557, p=0.129. BioCondition score was significantly related to mean NDSI (24 hours), F(1, 16)=19.86, p<0.001. The model had an adjusted R²=0.513.

140

5.3.3.3 Scatterplot There was a strong positive relationships between mean NDSI (24 hours) and BioCondition scores for RE 12.11.5e (R²=0.506, N=10) and RE 12.5.3 (R²=0.631, N=9) (Figure 5.5).

Figure 5.5: Scatterplot of Mean NDSI (24 hours) against BioCondition scores for RE 12.11.5e (blue) and RE 12.5.3 (green).

5.3.4 BioCondition vegetation and landscape attributes and NDSI

5.3.4.1 Hierarchical partitioning of mean NDSI (24 hours) Hierarchical partitioning values, including Z-scores, p-values and percent variation for mean NDSI (24 hours), based on BioCondition landscape and vegetation attributes for RE 12.11.5e and RE 12.5.3 are provided in Table 5.6 and Table 5.7. Separate datasets composed of landscape attributes, tree- and shrub-related structural attributes, ground-related structural attributes and species richness- related attributes were used for the hierarchical partitioning of NDSI (Supplementary Table 5.6 for RE 12.11.5e and Supplementary Table 5.7 for RE 12.5.3) due to the large number of predictor variables (attributes). Significant

141 attributes identified through this process were combined and used for the final hierarchical partitioning of NDSI.

Table 5.6: Hierarchical partitioning of mean NDSI (24 hours) based on BioCondition landscape and vegetation attributes for RE 12.11.5.e.

Patch size Patch Patch Shrub cover Tree connectivity context richness

Mean NDSI Z=2.452 Z=2.452 Z=2.382 Z= - 2.455 Z= - 2.398 (24 hours) p=0.006 p=0.024 p=0.009 p=0.006 p=0.006

%v=28.87 %v=16.56 %v=15.82 %v=21.02 %v=17.73

Table 5.7: Hierarchical partitioning of mean NDSI (24 hours) based on BioCondition landscape and vegetation attributes for RE 12.5.3.

Patch size Patch Patch CWD Litter Weed connectivity context cover cover

Mean Z=2.404 Z=2.216 Z=2.640 Z=2.498 Z= - 1.980 Z= - 1.954 NDSI (24 p=0.007 p=0.018 p=0.001 p=0.003 p=0.042 p=0.048 hours) %v=28.99 %v=21.15 %v=12.64 %v=18.11 %v=6.47 %v=12.64

5.3.4.2 Normality and homogeneity of variance tests The results of Shapiro-Wilk tests and Levene’s tests for BioCondition attribute data and transformed BioCondition attribute data identified through hierarchical partitioning are provided in Supplementary Tables 5.8 to 5.11.

5.3.4.3 ANCOVAs, linear regressions and scatterplots ANCOVAs for mean NDSI (24 hours) and BioCondition landscape and vegetation attributes identified through hierarchical partitioning are provided in Table 5.8. For

142 attributes and variables where the homogeneity of regression slope was violated, linear regressions are provided in Table 5.9. A flow chart summarising these steps is provided in Supplementary Figure 4.1. Of the attributes identified by hierarchical partitioning, four were identified by ANCOVAs as having a significant relationship with NDSI (patch connectivity, patch context, coarse woody debris and tree species richness) and two were not significantly related (weed cover and litter cover). All six attributes showed no effect of regional ecosystem. Linear regressions indicated that native shrub cover was significant but differed across regional ecosystems. Although patch size was not normally distributed and violated homogeneity of variance, it was still considered an important attribute for NDSI. The ratio of variances between RE 12.11.5e and RE 12.5.3 differed by a factor only slightly larger than 3:1 (3.56 for RE 12.11.5e and 1.08 for RE 12.5.3). Patch size was identified by hierarchical partitioning as the principal predictor variable for NDSI variation for both regional ecosystems. The failure of patch size to meet the requirements of Shapiro-Wilk and Levene’s tests was considered a product of small sample size and constraints associated with availability of various patch sizes. It may be assumed that given a large enough sample size a representative cross section of patch sizes would be present in the study area.

Landscape attributes (scatterplots)

A positive relationship between mean NDSI and log10 patch size was evident for RE 12.11.5e (R²=0.804, N=10) and RE 12.5.3 (R²=0.790, N=9) (Figure 5.9). There were positive relationships between mean NDSI and patch connectivity for RE 12.11.5e (R²=0.458, N=10) and RE 12.5.3 (R²=0.711, N=9), and patch context for RE 12.11.5e (R²=0.732, N=10) and RE 12.5.3 (R²=0.631, N=9) (Figure 5.6)

143

Figure 5.6: Scatterplots of Mean NDSI (24 hours) against log10 patch size, patch connectivity, patch context for RE 12.11.5e (blue) and RE 12.5.3 (green).

144

Tree- and shrub-related structural attributes (scatterplot)

A negative relationship was found between mean NDSI and native shrub cover for RE 12.11.5e (R²=0.712, N=10) while the inverse was identified for RE 12.5.3 (R²=0.470, N=9) (Figure 5.7).

Figure 5.7: Scatterplot of Mean NDSI (24 hours) against native shrub cover for RE 12.11.5e (blue) and RE 12.5.3 (green).

Ground-related structural attributes (scatterplots)

There was a strong positive relationship between mean NDSI and coarse woody debris for RE 12.5.3 (R²=0.656, N=9). There was a negative relationship between mean NDSI and litter cover for RE 12.5.3 (R²=0.306, N=9). There was a negative relationship between mean NDSI and weed cover for RE 12.5.3 (R²=0.470, N=9) (Figure 5.8).

145

Figure 5.8: Scatterplots of Mean NDSI (24 hours) against coarse woody debris, litter cover and weed cover for RE 12.5.3 (green).

146

Species richness-related attributes (scatterplot)

There was a negative relationship between mean NDSI and native tree species richness for RE 12.11.5e (R²=0.251, N=10) (Figure 5.9).

Figure 5.9: Scatterplot of Mean NDSI (24 hours) against tree species richness for RE 12.11.5e (blue).

147

Attribute Homogeneity of Levene’s test Fixed factor (RE) Covariate (NDSI) Adjusted R² regression

Patch size (log) Attribute violates homogeneity of variances (See Supplementary Table 5.11)

Connectivity (arcsin) F=1.703, p=0.212 F=1.577, p=0.226 F=1.433, p=0249 F=15.546, p=0.001 0.446

Context F=3.541, p=0.079 F=2.245, p=0.152 F=0.054, p=0.819 F=22.784, p<0.001 0.550

CWD F=4.317, p=0.055 F=0.194, p=0.665 F=0.054, p=0.819 F=6.770, p=0.019 0.233

Litter cover (arcsin) F=1.698, p=0.212 F=1.023, p=0.326 F=0.617, p=0444 F=2.083, p=0.168 Not significant

Weed cover (arcsin) F=2.785, p=0.116 F=0.385, p=0.543 F=1.414, p=0.252 F=3.977, p=0.063 Not significant

Tree richness F=0.209, p=0.604 F=1.979, p=0.178 F=2.952, p=0.105 F=6.640, p=0.020 0.229

Native shrub cover Linear regressions

148

Attribute Homogeneity of Regional β value t-test Significance Adjusted R² regression ecosystem

12.11.5e β = -0.844 t(8) = -4.442 p = 0.002 R² = 0.675, F(1, 8) = 19.731 Native shrub cover F=18.588, p=0.001 12.5.3 β = 0.686 t(7) = 2.493 p = 0.041 R² = 0.395, F(1, 7) = 6.217

149

5.4 Discussion

BioCondition and other multimetric condition approaches are designed to measure the ecological health of a system by quantifying fundamental habitat components essential for a broad range of plant and animal species (Parkes et al. 2003; Gibbons et al. 2009; Eyre et al. 2011b). Although BioCondition has been established as the primary indicator of terrestrial ecological condition in Queensland, its surrogacy potential has rarely been assessed through comparisons with biodiversity estimates or metrics of similar purpose. In this component of the study, the capacity for BioCondition to reflect target biodiversity values was examined by measuring the level of agreement between condition scores, attributes and the NDSI acoustic metric of ecological health.

Acoustic technologies have the potential to serve as powerful ecological assessment tools because multiple sensors can be deployed over large spatial scales and left to record for long periods of time. Despite their rise in popularity, research has focused on examining the temporal and spatial dynamics of the soundscape (Matsinos et al. 2008; Mazaris et al. 2009; Krause et al. 2011; Pijanowski et al. 2011a), rather than examining the link between the soundscape and ecological condition. However, Joo et al. (2011) and Kuehne et al. (2013) have demonstrated that land use intensification is related to increased human-generated sound, and can have detrimental effects on biophony. The relationship between urban development and reduced soundscape quality suggests that the ecological condition of impacted habitat may also be adversely affected.

5.4.1 NDSI and bird species richness and composition

Normalised Difference Soundscape Index (NDSI) was used as a comparative ecological health metric for BioCondition because it has the capacity to discriminate between anthrophony and biophony by partitioning the audible spectrum. NDSI

150 was significantly related to bird species richness among RE 12.11.5e and RE 12.5.3 sites suggesting that the reduced contribution of biophony to the soundscape may be directly related to fewer bird species calling. Furthermore, multivariate results indicated that regional ecosystem type and patch-level characteristics were the primary drivers of avian assemblages.

Similar conclusions have been drawn by Farina et al. (2011a, b), suggesting that the soundscape is likely to vary across different habitats. In addition, Bormpoudakis et al. (2013) have identified habitat-specific acoustic signatures and linked this to inferred structural and morphological differences in vegetation. Habitat, here described by regional ecosystem vegetation type, can modify bird feeding, breeding and nesting behaviours through variation in structural complexity, plant taxonomic composition and variable niche diversity (Rotenberry 1985; Skowno and Bond 2003; Diaz 2006; Tassicker et al. 2006; Hasui et al. 2007; Khanaposhtani et al. 2012; Pomara et al. 2012; Polyakov et al. 2013).

UPGMA analysis indicated that regional ecosystem type was pivotal for bird communities, with landscape factors such as patch size playing a secondary role. The split between RE 12.11.5e and RE 12.5.3 was produced by the absence of several ubiquitous, urban-adapted species from large- and medium-sized scribbly gum sites, including the Noisy Miner, Noisy Friarbird, Black-faced Cuckoo-shrike and Pale-headed Rosella. Their exclusion suggests that patches of scribbly gum woodland above 25 hectares prohibit the expansion of urban-adapted birds. However, medium-sized RE 12.5.3 patches ranged from 25 to 400 hectares compared to 25 to 200 hectares for RE 12.11.5e due to the scarcity of moderately sized scribbly gum forest remnants. Two of three RE 12.5.3 medium-sized patches within this cluster (Cluster 3) were greater than 200 hectares, suggesting that larger patch area may have contributed to the absence of urban-adapted species. Additionally, the relative effects of landscape heterogeneity may have also contributed to their absence, as large and moderately-sized RE 12.5.3 patches were

151 often associated with reduced land use intensification compared to equivalent RE 12.11.5e sites.

The impact of landscape spatial arrangement on bird species composition values was demonstrated by the significant split between UPGMA clusters representing many small patches and other patch sizes. Large remnants have been shown to have increased species diversity, a wider variety of habitats, more food resources, adequate area for many species to form territories, and increased resilience to anthropogenic and natural disturbance (Mörtberg 2001; Seddon et al. 2003). Furthermore, large patches are often well-connected in the landscape, facilitating dispersal and gene flow and effectively increasing vegetation area and potential habitat from which colonising species can arrive (Mörtberg 2001; Lampila et al. 2005; Lancaster et al. 2011). The results suggest processes related to species area relationships (Arrhenius 1921) and island biogeography theory (MacArthur 1967) were important drivers of avian assemblages across the study area, with habitat fragmentation having a major effect on broader biodiversity.

Small RE 12.11.5e and RE 12.5.3 patches were defined by a suite of urban-adapted bird species. Several, including the Noisy Miner, Noisy Friarbird and Blue-faced Honeyeater are aggressive, large nectarivores, and are often associated with reduced avian diversity (Sewell and Catterall 1998; Chan et al. 2004; Clarke and Oldland 2007; Maron et al. 2013). In particular, the Noisy Miner is noted for its despotic behaviour, forming colonies that actively exclude other passerine birds (Ashley et al. 2009; Maron et al. 2013). Habitat simplification, compounded by pressures related to the activities of aggressive species resulted in the absence of a suite of bushland-dependent avifauna including the Scarlet Honeyeater, Red- browed Finch, Grey Fantail, Leaden Flycatcher, Lewin’s Honeyeater, White-throated Treecreeper and Rufous Whistler from small forest remnants.

152

Associated sites were located near hard edges abutting residential development and surrounded by high road cover, promoting further disturbance and pressures related to wind, light and temperature exposure (Seddon et al. 2003; Malt and Lank 2007; Zharikov et al. 2007; Malt and Lank 2009). Small urban and peri-urban patches also provide favourable habitat for birds which move easily between native vegetation and the urban matrix (Palmer et al. 2008), including large nectarivores and species adapted to open areas. The Magpie-lark, Willie-wagtail, Masked Lapwing and Grey Butcherbird are often found in vegetation with low tree cover, and have expanded their range into developed areas exhibiting similar structural characteristics (Sewell and Catterall 1998).

5.4.2 NDSI and BioCondition

The relationship between NDSI and bird species richness established that changes in the acoustic index reflected variation in patterns of avian diversity. Based on this outcome, and the demonstrated potential for birds to act as biodiversity surrogates for unrelated organisms (Sauberer et al. 2004; Eglington et al. 2012; Larsen et al. 2012) and ecological condition (Jansen and Robertson 2005; Weinberg et al. 2008; Oliver et al. 2014), it may be argued that NDSI can serve as an effective ecological health metric.

NDSI was significantly related to BioCondition scores, confirming that BioCondition can reflect biodiversity values as measured by an independent soundscape metric for ecological health. NDSI was also related to BioCondition landscape attributes, with patch size, connectivity and context explaining much of the variation for NDSI. Similarly, results from Chapter 4 indicated that landscape factors had a significant effect on vegetation attributes related to understorey strata. Dominant fragmentation processes suggested that the spatial arrangement of remnant patches was fundamental to the ecological condition of natural systems within the study area. However, it was demonstrated that individual vegetation attributes

153 were also important for soundscape variation. Faunal and floristic species diversity have been linked to a range of habitat attributes, including fallen timber (Sullivan et al. 2012; Bowman and Facelli 2013) and vegetation structure (Gil-Tena et al. 2007; Pekin et al. 2012; Zhang et al. 2013; Lam et al. 2014), as well as land use intensification and the degradation of natural areas (Suarez-Rubio et al. 2009; Flaspohler et al. 2010; Murphy et al. 2014). The significant relationships between NDSI and coarse woody debris, weed cover, litter cover, native shrub cover and tree species richness underscore the importance of fine-scale ecological patterns and processes to biodiversity, particularly with reference to audible species.

In terms of acoustic measures of ecological health, coarse woody debris serves primarily as habitat for audible fauna such as frogs (Blomquist and Hunter 2010), and organisms that may be eaten by passerine birds (Kappes et al. 2006; Horn and Hanula 2008; Rosenvald et al. 2011; McGregor and Burnett 2014). Consequently, high levels of coarse woody debris may contribute to increased biological sound. Vegetation type and productivity and patterns of disturbance have a substantial influence on coarse woody debris turnover (Woldendorp et al. 2002), and may partly explain the significant relationship between NDSI and coarse woody debris for RE 12.5.3. The dominant tree species for this largely nutrient-deficient ecosystem is Eucalyptus racemosa or scribbly gum. Established trees of 150 to 200 years of age produce hollows (Ross 1998) and may generate large quantities of coarse woody debris. The exclusion of sporadic and intense wildfires from Eucalyptus racemosa woodlands may ensure the persistence of large trees, the continued production of coarse woody debris and the retention of fallen timber in the ground layer. However, strategic fire management is often limited to large patches. Small RE 12.5.3 remnants may lose much of their coarse woody debris, either through high intensity fires, or restricted production where Eucalyptus racemosa is replaced by Allocasuarina littoralis as the dominant canopy species. The relationship between NDSI, levels of coarse woody debris and altered fire regimes suggests that decreased biophony and increased anthrophony, although

154 associated with fine-scale ecological variation, may also be a response to landscape processes.

Similar conclusions may be drawn from the negative response of NDSI to high litter cover. As shown in Chapters 3 and 4, small RE 12.5.3 patches dominated by Allocasuarina littoralis had increased shading and litter accumulation, with a consequent decrease in native shrub cover and reduced shrub species richness (Specht and Specht 1989b; Lunt 1998; Close 2009). Low NDSI values may be attributed to habitat fragmentation and associated changes in management and disturbance regimes rather than the direct effects of altered ground strata.

Although analogous processes may underpin the negative relationship between NDSI and weed cover, invasive plant species can have a direct effect on faunal composition and diversity, including audible taxa, through altered vegetation structure and changes in the relative abundance of food, shelter and nesting resources (Neave et al. 1996). Interactions between exotic plants and other biota may be complex. In many cases, species are adversely affected by invasive vegetation. Maron and Lill (2005) observed that small insectivorous birds avoided areas of high exotic grass cover. Structural changes to the ground stratum made prey detection more difficult, resulting in an increased reliance on energy intensive aerial movements while feeding. Similar impacts have been recorded for ground- feeding granivorous birds excluded from areas dominated by invasive buffel grass, with subsequent reductions in bare soil cover (Smyth et al. 2009). Other taxa, including reptiles, may also be constrained by invasive plant species, with lower numbers attributed to suboptimal thermal conditions, reduced prey levels and increased detection by predators (Valentine et al. 2007; Hacking et al. 2014).

However, exotic vegetation may also benefit species. Gosper and Vivian-Smith (2006) found that the fruit of plants introduced to South-East Queensland had

155 smaller seeds, higher sugar levels and more variable nitrogen than their indigenous counterparts, making them highly attractive to frugivorous avifauna. One weed in particular, Lantana camara, has been shown to support small passerine birds through the provision of food resources and dense shrub cover, although at the expense of many native plant species (Sharma et al. 2005; Gosper and Vivian-Smith 2006; Kath et al. 2009). In this study, it was unlikely that the positive effects of increased weed cover offset the detrimental impacts on broader biodiversity. Weeds were associated with other significant indicators of habitat decline, including fragmentation. Furthermore, increased shrub cover was not a proxy for improved ecological condition or high NDSI values for RE 12.11.5e, suggesting that the positive contribution of shrubby, invasive species such as Lantana camara to avian diversity may be negated by other unfavourable factors.

The inverse relationship of NDSI with native shrub cover paralleled those observed between shrub cover and patch size for RE 12.11.5e and RE 12.5.3, and suggested that fragmentation processes were a definitive driver of variation among NDSI values and individual vegetation attributes. Once again, this was linked to reduced management activity in small forest remnants, with the shrub layer of each vegetation community altered through fire suppression or exclusion. Small RE 12.11.5e patches were defined by the infiltration and subsequent high cover of dry rainforest and shade tolerant species in the tree and shrub strata, while RE 12.5.3 transitioned from a canopy dominated by Eucalyptus racemosa to Allocasuarina littoralis with depauperate and structurally simple understoreys (Specht and Specht 1989b; Lunt 1998; Close 2009).

Similarly, low soundscape values corresponded with increased tree species numbers, representing ecosystem state changes driven by reduced burning. The pattern was consistent across regional ecosystems, although high tree species richness for RE 12.5.3 may also be attributed to edaphic factors. RE 12.5.3 was represented by two distinct communities. One community typically found on

156 shallow, less productive soils had canopy and subcanopy layers with relatively few tree species and understoreys dominated by heathland vegetation. The other community was associated with deeper soils of moderate to high fertility, increased tree species richness and woodland understoreys (Specht and Specht 1989b). It may be argued that RE 12.5.3 forests on deep, fertile substrates were historically targeted for agricultural production, with past and current anthropogenic impacts having a pronounced effect on the associated levels of ecological condition and biodiversity.

5.5 Conclusion

Ecological condition assessment tools are important for managing and protecting natural systems under increasing anthropogenic pressures. However, to be an effective surrogate for biodiversity values it is imperative that multimetric condition approaches, here represented by BioCondition, are assessed against the distributions and diversity of target species and independent indices of similar purpose.

The acoustic index NDSI represented bird species richness and composition for both regional ecosystems. Regional ecosystem type and patch configuration were identified as significant drivers of avian assemblages. The results suggested that NDSI reflects ecological processes underpinning variation in bird diversity, and confirmed that the acoustic index was a suitable metric against which BioCondition could be compared. NDSI values were significantly related to BioCondition scores, suggesting that the BioCondition approach has the capacity to meaningfully measure changes in biodiversity values and ecological condition. In addition, these changes were also exhibited by a number of constituent condition attributes, with consistent, strong relationships identified between NDSI and landscape-level attributes. Although there were several relationships between NDSI and BioCondition vegetation attributes, it was found that much of their impact could be

157 linked to the effects of forest fragmentation. These results indicate that broad landscape processes are fundamental to the measurement of ecological condition.

158

Chapter 6: General discussion and conclusion

6.1 General discussion

As human populations expand and the need for space and resources grow, there is a subsequent increase in the impacts on natural systems. In Australia, and for many other parts of the world, these impacts include habitat fragmentation, the encroachment of invasive species, overgrazing, altered fire regimes and climate change (Martin et al. 2014). In response to these threats, a variety of biodiversity management and conservation planning tools have been developed.

Vegetation mapping and classification systems have been established to describe vegetation extent and associated abiotic factors as discrete ecological units. These units may be used as a foundation for measuring and monitoring biodiversity (Ferrier 2002), and to inform the implementation of other planning approaches. For example, conservation reserve networks are often identified based on their capacity to represent species distributions and ecological communities (Margules and Pressey 2000; Sarkar et al. 2006). Therefore, their design can be directed by current vegetation extent and the modelling and mapping of habitat for flora and fauna (Brotons et al. 2004; Wintle et al. 2005). Vegetation mapping may guide ecological restoration, where a detailed understanding of the composition, structure and function of intact ecosystems is used for comparative assessment (Ruiz-Jaen et al. 2005). Additionally, current vegetation mapping can underpin measurements of ecological condition. Multimetric condition indices can quantify ecosystem health by assessing and scoring sites against benchmarks derived from reference sites belonging to specific vegetation units.

Multimetric assessments of ecological condition have become important biodiversity management tools that are applied across a range of ecosystems under a variety of anthropogenic pressures (Andreasen et al. 2001; Parkes et al. 2003; Bleby et al. 2008; Gibbons et al. 2009; Eyre et al. 2011b). Over the past fifteen years, terrestrial multimetric condition approaches have become established in

161

Australia and used by government agencies to measure the impacts of development (Australian and New Zealand Environment and Conservation Council 1999), regulate vegetation clearing approvals and determine ecological offsets (Gibbons et al. 2008; Eyre et al. 2011b). For Queensland, the multimetric BioCondition approach (Eyre et al. 2011b), in conjunction with the regional ecosystem vegetation mapping framework (Sattler and Williams 1999), has become an integral component of conservation planning and policy for the state.

6.1.1 Vegetation patterns and variability

Climatic and environmental gradients are fundamental drivers of ecosystem assembly and change, and underpin the concept of vegetation communities as uniform entities (Reed et al. 1993; Battaglia and Williams 1996; Van der Maarel 2005). It was demonstrated in this study that the effects of variability on vegetation composition, structure and abundance, and on ecological condition attributes were often restricted to each regional ecosystem, describing a landscape of biotic and abiotic heterogeneity, and conforming to a Gleasonian model of community organisation and structure (Gleason 1926).

It was also shown that latitudinal effects were universally important for vegetation communities within the study area. However, these effects were linked to climatic gradients across two distinct subregions for RE 12.11.5e and to the geographical disjunction between vegetation communities and disturbance processes for RE 12.5.3. Climatic gradients primarily influenced RE 12.11.5e tree layers, with changes in understorey strata driven by disturbance processes and fine-scale environmental conditions. These patterns may be based on plant growth forms responding differentially to variability. The deep roots of mature trees may mitigate the effects of episodic, fine-scale environmental change, except where abiotic factors, such as high soil moisture, become pervasive constraints. In addition, the large size and thick bark of dominant Myrtacaeae species, and their ability to resprout following

162 fire, can protect plants from damage and ensure persistence (Attiwill 1994b; Burrows 2002; Lawes et al. 2011; Wesolowski et al. 2014). Conversely, disturbance processes have sudden and dramatic effects on understorey species. Short-lived and shallow-rooted plants are susceptible to a range of impacts including grazing, low-intensity fires and competition from weeds (Tremont and McIntyre 1994; Best 2008; Dorrough and Scroggie 2008; Price et al. 2011). Disturbance impacts on large canopy species may only become evident following significant lag times, although immediate changes in canopy recruitment can indicate future vegetation patterns (Brown and Wu 2005). In this study, similar effects were also linked to climatic variability, with changes in rainfall and temperature principal drivers of large tree species composition, structure and abundance. It has been suggested that mature trees can persist for centuries in the landscape following climatic change that may be detrimental to seedling growth, effectively retaining vegetation boundaries under markedly different environmental conditions (Brubaker 1986).

Although broad relationships between vegetation and variability for RE 12.11.5e subregions were characterised by the distinct responses of different growth forms to climatic gradients and disturbance regimes, it was shown that climatic effects were not only restricted to canopy layers, and disturbance processes were not confined to shrub and ground strata. Tree species composition and structure can be shaped by a range of disturbance processes, including fire (Jurskis and Walmsley 2012) and logging activity (Horner et al. 2012). In this study it was demonstrated that fire management can have a significant impact on dominant tree species. Fire suppression induced vegetation communities to transition from open eucalypt forest or woodland to closed, dry rainforest for RE 12.11.5e, and to Allocasuarina dominated systems for RE 12.5.3. Conversely, changes among understorey plants can often be linked to precipitation and temperature gradients (Pugnaire and Lazaro 2000; Egan and Williams 1996; Tang et al. 2012). In the current study it was shown that the presence and abundance of some shrub (e.g. Indigofera australis, Persoonia sericea and Xanthorrhoea johnsonii) and ground species (e.g. Drynaria rigidula, Smilax australis, Hardenbergia violacea, Imperata cylindrica and Pteridium

163 esculentum), and forb and grass species richness, were related to temperature and rainfall variability due to geographic location.

6.1.2 Implications for benchmarking

The significance of variability in vegetation patterns caused by ecological factors must be considered in the design of ecological condition assessment frameworks if these metrics are to function effectively. It is a requirement that the measured attributes of multimetric condition frameworks represent the composition, structure and function of an ecosystem, and are sensitive to change. The tension between operating efficiency and scientific rigour dictates that condition metrics must work as effective surrogates without the need for large numbers of attributes (Oliver et al. 2007). Prior studies have found that the vegetation and landscape attributes of BioCondition and similar multimetric approaches can represent ecological variation and detect disturbance (Oliver et al. 2007; Eyre et al. 2011b). Therefore, benchmarks must be identified that represent pre-disturbance states against which attribute values can be compared and quantified (Landsberg and Crowley 2004; McElhinny et al. 2005; Gibbons et al. 2008; Eyre et al. 2011b).

Several mechanisms allow benchmarking to account for climatic (latitude) and environmental (edaphic and landform) variability. Firstly, variability linked to climatic conditions may be addressed by establishing geographically-restricted reference sites. Subregional reference sites are already an important component of the BioCondition framework (Eyre et al. 2011a), with location placement informed by regional ecosystem mapping. In this study it was necessary to establish two benchmark values for the Brisbane and Gympie RE 12.11.5e subregions. Although both areas were described by the same bioregion, land zone and dominant vegetation type, distinct differences based on climatic variability were detected. Brisbane was characterised by more variable climatic conditions and Gympie by increased rainfall and higher temperatures. Establishing subregional benchmarks

164 may be simple where there are geographical disjunctions. However, where the distinction is less clear, then current and historical climatic datasets may provide a reasonable basis for reference site location.

Variability attributed to environmental gradients, including changes in soil characteristics and landform, may also be captured through strategic benchmarks underpinned by the mapping of uniform ecological units. BioCondition vegetation attributes are based on biotic-abiotic associations described by the regional ecosystem framework. However, without a dedicated mapping and conservation planning tool, environmental variability may need to be identified using existing analogue and digital maps and ground-truthing surveys. For regional ecosystems, land zones represent soil, landform and geology (Sattler and Williams 1999), and provide some measure of variation in vegetation, the physical and chemical properties of soils and landscape patterns. However, they are mapped at 1:250000 scale and may be unable to detect fine-scale variability. In this study, differences in vegetation composition and structure between RE 12.5.3 woodland and heathland sites demonstrated the potential for broad-scale mapping to overlook environmental gradients. Although woodland areas with relatively productive, deep soil profiles, and heathland with low nutrient substrates may be considered as within the natural range of variation for this specific regional ecosystem, it may also represent a distinct change in vegetation, soil and topography. If so, it suggests that the two communities may need to be mapped as separate entities with their own reference sites and benchmark values.

Climatic and environmental variability can affect the distribution, composition and structure of vegetation communities, and may be linked to significant human- induced impacts (Williams et al. 2000; Lepetz et al. 2009; Yates et al. 2010; Wiens et al. 2011). Changes in rainfall and temperature, and their relationship with gradients such as soil moisture, can operate at much longer time scales than those of other disturbance processes (Botta and Foley 2002). Whereas changes in edaphic

165 conditions are relatively simple for condition metrics to accommodate, climatic variability, particularly in the context of anthropogenic climate change, may be more difficult. New vegetation patterns driven by changes in rainfall and temperature suggest that condition frameworks would need to establish revised benchmarks over time.

An ideal reference condition metric should be robust to natural variability due to strategic reference site selection, yet sensitive to human impacts (Hawkins 2006; Alahuhta and Aroviita 2016). In this study it was shown that changes in ecological condition attributes were related to disturbance processes. These relationships were largely underpinned by the inclusion of attributes such as patch size. Other elements, including changes in weed cover, shrub cover, and levels of coarse woody debris, were also indicative of disturbance. However, these attributes often co- occurred with broad-scale fragmentation patterns, and their variation may be attributed to landscape-level impacts. This suggested that anthropogenic disturbance in the form of habitat fragmentation was the dominant driver of reduced ecological condition in the study system.

Multimetric ecological condition approaches may account for the considerable variability attributed to fragmentation by heavily weighting related attributes. The BioCondition framework (Eyre et al. 2011b), and other Australian multimetric indices such as Habitat Hectares (Parkes et al. 2003; Department of Sustainability and Environment 2004), have integrated the effects of broad-scale landscape processes on biodiversity and their contribution to measurements of ecological condition by assigning twenty to twenty-five percent respectively to landscape attributes as part of overall condition scores. Similarly, other measurements of ecological condition and habitat change have focused on quantifying landscape attributes in order to determine the effects of urbanisation on fauna and flora (Kim and Pauleit 2005), assess land use impacts on ecosystems over time (Olsen et al. 2007) and examine landscape drivers of species diversity (Lopez-Martinez et al.

166

2013). Landscape attributes are particularly useful and attractive components for biodiversity and condition assessment because of their demonstrated relationships with species distributions and diversity (Alados et al. 2009; Kumar et al. 2009; Uezu and Metzger 2011), and their ease of measurement based on remotely sensed and mapped data (Yang and Liu 2005; Dikou et al. 2011). Due to the inordinate effect that patch dimensions and arrangement have on ecological condition, it is imperative that reference sites are located in large, well connected remnants in close proximity to other areas of intact vegetation.

As discussed above, disturbances that are a consequence of fragmentation are well- accounted for in condition metrics. However, in this study condition measurements were in some instances unreliable in detecting other types of extrinsic anthropogenic disturbance. For example, current and historical logging activity in the Gympie subregion was shown to be strongly related to tree species abundance for RE 12.11.5e, with relative numbers of Corymbia citriodora, Eucalyptus crebra and Eucalyptus siderophloia lower than those for Brisbane sites. However, logging did not have a detectable effect on the number of large trees in the BioCondition metric, and may indicate that ecological condition measures may not be as sensitive as detailed vegetation surveys for certain types of disturbance.

Many disturbance regimes can cause rapid and persistent changes to native vegetation. However, changes may not always be detrimental, and are within the natural range of variation for relatively intact ecosystems. This is pertinent for fire- prone systems such as those examined in this study. Whereas logging and ungulate grazing are mediated by humans, fires can either be ignited by natural events such as lightning strikes, take the form of deliberately or accidentally lit wildfires, or serve as management tools (Bowman et al. 2004; Kilinc and Beringer 2007; Taylor et al. 2013). Alongside increasing aridity, fire has played an important role in shaping the Australian landscape (Black et al. 2007; Mooney et al. 2007). In the current study, fire-excluded sites were significantly altered when compared to

167 those that had been actively managed. The removal of fire was consistently associated with dramatic changes to canopy and understorey vegetation, the production and accumulation of leaf litter, and reduced levels of ecological condition.

The effects of fire have important implications for benchmarking. Determining the natural range of variation for fire regimes is considered essential for reference site selection. BioCondition guidelines stipulate that reference sites should have minimal modification through fire (Eyre et al. 2011a), and that modification is not anthropogenically driven. However, many intact Australian ecosystems have a history of indigenous and European fire management, and continue to be burnt intermittently. State and transition models of vegetation change provide a useful framework for conceptualising fire regimes for reference conditions, but also raise questions regarding the nature of benchmarks and their relationship with disturbance. If state and transition models suggest that the same ecosystem can occur in a range of alternative stable states (Westoby et al. 1989), and that condition metrics require reference sites to represent states that are relatively unmodified by human activity, then how do you determine what serves as a reference site for systems that were historically and are actively managed with fire?

Under these conditions, benchmarking must be informed through a sound ecological understanding of the drivers of ecosystem change and the states that may result. Knowledge of historical processes, including fire, that act on vegetation composition, structure and abundance may be developed using archival material, including survey data, aerial photography, journals, historical images and other records (Lunt 2002; Zhang et al. 2007; Whipp et al. 2009; Whipp et al. 2012; Jurskis and Underwood 2013). Some have attempted to identify and recreate earlier vegetation patterns, often artefacts of indigenous fire practices, in order to manage forest fires (Roccaforte et al. 2015), understand the historical role of fire variability (Benson and Redpath 1997; Lunt 1998; Korb et al. 2013), and predict future

168 environmental change (Nonaka and Spies 2005; Thompson et al. 2009). However, it may be difficult to reveal earlier ecological states in the absence of historical data, suggesting that the use of available reference conditions, without necessarily accounting for long-term fire variability, may be adequate for benchmarking purposes as long as they do not detract too much from a relatively intact vegetation state (Bestelmeyer et al. 2009). If reference sites are reasonably unmodified by human activity, and are used as a point of comparison for sites located in the same ecosystem type, under the same climatic conditions, and exposed to the same environmental gradients and disturbance regimes, then their capacity to function effectively may be preserved. In this study, altered fire regimes was also closely aligned with habitat fragmentation. Decreasing patch size and connectivity were consistently associated with fire suppression. The use of current datasets, including recent fire mapping, changes in land use and expanding infrastructure networks, may resolve patterns of landscape change (Close et al. 2008; Regan et al. 2010; Piquer-Rodriguez et al. 2012), guide reference site selection and indicate potential changes to ecological condition (Kline et al. 2001).

6.1.3 Comparative index measurements

The significant relationship between BioCondition and the Normalised Difference Soundscape Index (NDSI) revealed that multimetric ecological condition indices have the capacity to meaningfully measure changes in biodiversity values and ecological condition. Furthermore, landscape spatial arrangement was as significant for the soundscape as it was for traditional condition assessment methods. In this study, patch size, connectivity and context were fundamental drivers of NDSI and bird species richness and composition for RE 12.11.5e and RE 12.5.3. Soundscape metrics have also shown that changes in species diversity and ecological health are directly linked to changes in land use, with gradients of increasing intensity related to high levels of anthropogenic sound and an associated reduction in audible biological signals (Joo et al. 2011; Kuehne et al. 2013) and vegetation complexity (Joo et al. 2011).

169

The relationships between NDSI, avian assemblages and BioCondition scores suggested that landscape fragmentation was the predominant driver of change in plant and animal populations for the study area. Site-level condition attributes did account for some of the variation in NDSI, but attributes differed across regional ecosystems. However, this may also be due to landscape configuration. NDSI increased with coarse woody debris, as did patch size. Although coarse woody debris is important as habitat for audible species and other fauna, the amount of debris is often a product of fragmentation, with decreased coarse woody debris associated with altered fire regimes, timber collection and edge effects (Woldendorp et al. 2002; Gould et al. 2008). Similarly, the negative relationship between NDSI and weed cover was paralleled by that between weed cover and patch size, indicating that the establishment of invasive plant species is often dependent upon the spatial arrangement of vegetation (Saunders et al. 1991; Fahrig and Merriam 1994; Bradshaw 2012). This was also true for certain structural characteristics of native vegetation. The inverse relationships of NDSI and native shrub cover for RE 12.11.5e and RE 12.5.3 were reflected by those for native shrub cover and patch size. It was determined that these patterns were a product of two different vegetation communities responding to disturbance processes, particularly changes in fragmentation-related fire activity.

6.1.4 Implications for management

The findings of this study have management implications for multimetric condition assessments. In many cases, practical responses to variability have already been incorporated into the BioCondition framework, but remain challenging in a practical sense. Strategic reference site location for benchmarking is fundamental to the operational efficacy of reference condition models. It is therefore necessary for multimetric condition approaches to locate appropriate reference sites based on ecosystem type, climatic associations, environmental gradients and a natural range of disturbance. Dedicated mapping frameworks, such as the regional ecosystem vegetation mapping system for Queensland, are essential for accurate data

170 collection and interpretation, and potentially the first step for establishing a surrogate system for measuring ecological condition. There are many analogous systems based on vegetation groups and ecological associations (Neuhausl 1990; Sun et al. 1997; Bickford and Mackey 2004; Hong et al. 2004; Thackway and Lesslie 2008). Even where they are not employed for environmental surrogacy approaches, they may be easily modified to accommodate their use.

Although ecological boundaries and associations may be determined and mapped for the purposes of strategic benchmarking, incorporating disturbance regimes into the natural range of variation for reference sites can be more complex. Many disturbance processes have shaped what are now considered natural landscapes. In the Australian context, and in other societies with a pre-colonial indigenous history, disturbance is often linked to long-term fire management. Management practices often aim for a mosaic approach to burning, creating a range of alternative stable states for vegetation communities (Perry and Enright 2002; Batllori et al. 2015). For other regions, disturbance may also be related to grazing or other human-induced perturbations. Regardless, thresholds of disturbance, beyond which a landscape is considered modified and unsuitable for reference sites, must be identified. A practical response may be to select reference sites that are ideally informed by historical records and current understandings of disturbance. For example, in Queensland, fire management practices are guided by quantitative and qualitative data and based on expert observation, and are linked to regional ecosystems considered to be in good ecological condition (Department of Environment and Heritage Protection 2014c; Queensland Herbarium 2014). Reference sites must be located in vegetation communities that have not changed to any considerable extent from what is perceived as an ‘unmodified’ state, and as long as they are compared to the same ecosystems under similar environmental conditions, then they may be considered representative. However, finding appropriate benchmarks in fragmented, urbanised regions may not always be possible. For the South-East Queensland bioregion, extensive habitat loss can limit the range and number of appropriate reference sites. A ‘best on offer’ approach is a practical response to

171 these limitations, where benchmarks based on the closest approximation to unmodified ecosystems are used (Eyre et al. 2011b). Even so, where vegetation communities are endangered and highly degraded, finding representative ‘best on offer’ sites may also be difficult.

Whereas management issues related to mapping frameworks and reference site selection are primarily concerned with surrogate application, it may be possible to modify surrogate design to improve utility without compromising the capacity to measure ecological condition. As indicated in the current study, landscape attributes are a significant and easily measured source of variability. Although vegetation attributes contribute to the ability to discriminate between sites, their inclusion may also provide little additional information. Plant species richness is a common component of multimetric condition approaches. However, some have argued that they are an unnecessary feature. Multimetric condition approaches serve as surrogates for species diversity, and using richness measures of dominant life forms as an attribute may be a case of inbuilt redundancy (Oliver et al. 2007; Mandelik et al. 2010; Oliver et al. 2014). In addition, collecting species data for sites can be time consuming and technically difficult for non-specialists, particularly when identifying seasonal grass species (Cook et al. 2010). Therefore, it may be appropriate to remove tree, shrub, forb and grass species richness from BioCondition attributes, and equivalent components from other multimetric condition approaches, if scores remain unaffected.

The proposal for removing richness estimates from multimetric indices was tested using BioCondition scores for the 62 survey sites. Species richness attributes were removed from the BioCondition score assessment for each site and the resulting values were readjusted (/0.8) to give a score out of 1. Adjusted scores were compared with original BioCondition scores for each site using a two-tailed Wilcoxon Signed Ranks test. There was no significant difference between the scores (Z=-1.383, N=62, p=0.167), with an adjusted mean score of 0.8057 (+/- standard

172 deviation 0.1244), and an original mean score of 0.8002 (+/- standard deviation 0.1063). These results suggest that multimetric condition approaches with a combination of structural vegetation and landscape attributes alone can quantify ecological condition without the need for vegetation richness estimates.

A final consideration for multimetric condition assessments is the potential to augment existing field methods with novel metrics of ecological health. In the current study it was shown that acoustic indices can be effective tools for measuring ecological change. Remote techniques such as soundscape assessments (Gage et al. 2001; Kasten et al. 2012) can add to ground-truthed surveys, or replace them where they are physically or temporally constrained.

6.2 Conclusion and further studies

Multimetric condition assessments have become important tools for biodiversity management. They are designed to be repeatable and transparent measurements, while reducing reliance on subjective judgement (Parkes et al. 2003; McCarthy et al. 2004, Eyre et al. 2011b). This study is the first to examine the reliability of multimetric condition assessments across variable environments, and the implications this variability may have for surrogate efficacy. It was shown that strategic attribute selection and relative weighting, and appropriate reference site location can mitigate the effects of geography, climate, environmental gradients and disturbance regimes as long as decisions are based on a sound understanding of the ecological processes shaping vegetation communities.

To further assess the robustness of multimetric indices, it is necessary to replicate the approach used in this study in a variety of other ecosystems (e.g. grasslands, rainforests) associated with a disparate range of gradients and disturbance patterns. Furthermore, comparisons of multimetric condition approaches with

173 other indices of ecological health or integrity, and independent assessments of species and functional diversity can help to determine whether terrestrial ecological condition can accurately represent biodiversity values and ecological processes.

The current study has shown that multimetric condition approaches are robust to ecological variability. However, there is scope for further, small adjustments in order to improve metric efficiency without significant loss of data. The limitations of multimetric condition assessments need to be recognised. Rare and threatened species may be overlooked by condition assessments (Wintle 2008). In addition, small remnant patches with high vegetation condition attribute values will always score poorly in terms of overall ecological condition, which has important implications for vegetation offsetting where ecological quality may be replaced with quantity elsewhere. Although the application of ecological condition indices may be restricted by their design and the underlying assumptions on which they are based, if used in conjunction with other assessment approaches and datasets (e.g. species distribution models), they have the potential to serve as practical and powerful tools for conservation planning.

174

References

Abensperg-Traun, M., Atkins, L., Hobbs, R., & Steven, D. (1998). Exotic plant invasion and understorey species richness: A comparison of two types of eucalypt woodland in agricultural Western Australia. Pacific Conservation Biology, 4(1), 21-32.

Adler, P. B., HilleRisLambers, J., Kyriakidis, P. C., Guan, Q., & Levine, J. M. (2006). Climate variability has a stabilizing effect on the coexistence of prairie grasses. Proceedings of the National Academy of Sciences, 103(34), 12793-12798.

Alados, C. L., Navarro, T., Komac, B., Pascual, V., Martinez, F., Cabezudo, B., & Pueyo, Y. (2009). Do vegetation patch spatial patterns disrupt the spatial organization of plant species? Ecological Complexity, 6(2), 197-207.

Alahuhta, J., & Aroviita, J. (2016). Quantifying the relative importance of natural variables, human disturbance and spatial processes in ecological status indicators of boreal lakes. Ecological indicators, 63, 240-248. Alho, C. J. R. (2008). The value of biodiversity. Brazilian Journal of Biology, 68(4), 1115-1118.

Alofs, K. M., & Fowler, N. L. (2013). Loss of native herbaceous species due to woody plant encroachment facilitates the establishment of an invasive grass. Ecology, 94(3), 751-760.

Alvarez-Cansino, L., Zunzunegui, M., Barradas, M. C. D., Correia, O., & Esquivias, M. P. (2013). Effects of temperature and rainfall variation on population structure and sexual dimorphism across the geographical range of a dioecious species. Population Ecology, 55(1), 135-146.

Anand, M. (2000). The fundamentals of vegetation change-complexity rules. Acta Biotheoretica, 48(1), 1-14.

Andersen, A. N., & Majer, J. D. (2004). Ants show the way Down Under: Invertebrates as bioindicators in land management. Frontiers in Ecology and the Environment, 2(6), 291-298.

Andreasen, J. K., O’Neill, R. V., Noss, R., & Slosser, N. C. (2001). Considerations for the development of a terrestrial index of ecological integrity. Ecological Indicators, 1(1), 21-35.

177

Arnaiz, O. L., Wilson, A. L., Watts, R. J., & Stevens, M. M. (2011). Influence of riparian condition on aquatic macroinvertebrate communities in an agricultural catchment in south-eastern Australia. Ecological Research, 26(1), 123-131.

Arrhenius, O. (1921). Species and area. Journal of Ecology, 9(1), 95-99.

Ashley, L. C., Major, R. E., & Taylor, C. E. (2009). Does the presence of grevilleas and eucalypts in urban gardens influence the distribution and foraging ecology of Noisy Miners? Emu, 109(2), 135-142.

Attiwill, P. M. (1994a). The disturbance of forest ecosystems—the ecological basis for conservative management. Forest Ecology and Management, 63(2-3), 247-300.

Attiwill, P. M. (1994b). Ecological disturbance and the conservative management of eucalypt forests in Australia. Forest Ecology and Management, 63(2-3), 301-346.

Austin, M. P. (1985). Continuum concept, ordination methods, and niche theory. Annual Review of Ecology and Systematics, 39-61.

Austin, M. P., & Smith, T. M. (1989). A new model for the continuum concept. Vegetatio, 83(1-2), 35-47.

Austin, M., Nicholls, A., & Margules, C. R. (1990). Measurement of the realized qualitative niche: Environmental niches of five Eucalyptus species. Ecological Monographs, 60(2), 161-177.

Austin, M., Pausas, J. G., & Nicholls, A. (1996). Patterns of tree species richness in relation to environment in southeastern New South Wales, Australia. Australian Journal of Ecology, 21(2), 154-164.

Australian and New Zealand Environment and Conservation Council (1999). The National Framework for the Management and Monitoring of Australia's Native Vegetation. Canberra: Australian Government.

Avidano, L., Gamalero, E., Cossa, G. P., & Carraro, E. (2005). Characterization of soil health in an Italian polluted site by using microorganisms as bioindicators. Applied Soil Ecology, 30(1), 21-33.

178

Baker, W. L., & Weisberg, P. J. (1995). Landscape Analysis of the Forest-Tundra Ecotone in Rocky Mountain National Park, Colorado∗. The Professional Geographer, 47(4), 361-375.

Barrett, T. J. (2011). Computations using analysis of covariance. Wiley Interdisciplinary Reviews: Computational Statistics, 3(3), 260-268.

Barton, P. S., Westgate, M. J., Lane, P. W., MacGregor, C., & Lindenmayer, D. B. (2014). Robustness of habitat-based surrogates of animal diversity: A multitaxa comparison over time. Journal of Applied Ecology, 51(5), 1434-1443.

Batllori, E., Ackerly, D. D., & Moritz, M. A. (2015). A minimal model of fire-vegetation feedbacks and disturbance stochasticity generates alternative stable states in grassland-shrubland-woodland systems. Environmental Research Letters, 10(3).

Battaglia, M., & Williams, K. J. (1996). Mixed species stands of eucalypts as ecotones on a water supply gradient. Oecologia, 108(3), 518-528.

Bauerfeind, S. S., Theisen, A., & Fischer, K. (2009). Patch occupancy in the endangered butterfly Lycaena helle in a fragmented landscape: Effects of habitat quality, patch size and isolation. Journal of Insect Conservation, 13(3), 271-277.

Bazelet, C. S., & Samways, M. J. (2012). Grasshopper and butterfly local congruency in grassland remnants. Journal of Insect Conservation, 16(1), 71-85.

Beketov, M. A., & Liess, M. (2008). Acute and delayed effects of the neonicotinoid insecticide thiacloprid on seven freshwater arthropods. Environmental Toxicology and Chemistry, 27(2), 461-470.

Belbin, L. (2004). PATN (Version 3): Blatant Fabrications Pty Ltd. Retrieved from http://www.patn.com.au/

Benson, D., & McDougall, L. (1998). Ecology of Sydney plant species Part 6: Dicotyledon family Myrtaceae in Cunninghamia. 5 (4): 808-983. National Herbarium of New South Wales, Royal Botanic Gardens, Sydney.

179

Benson, J., & Redpath, P. (1997). The Nature of Pre-European Native Vegetation in South-eastern Australia: A Critique of Ryan, DG, Ryan JR and Starr, BJ (1995), The Australian Landscape: Observations of Explorers and Early Settlers: Royal Botanic Gardens, National Herbarium of New South Wales.

Bergeron, J. A. C., Blanchet, F. G., Spence, J. R., & Volney, W. J. A. (2012). Ecosystem classification and inventory maps as surrogates for ground beetle assemblages in boreal forest. Journal of Plant Ecology, 5(1), 97-108.

Best, R. J. (2008). Exotic grasses and feces deposition by an exotic herbivore combine to reduce the relative abundance of native forbs. Oecologia, 158(2), 319-327.

Bestelmeyer, B. T., Tugel, A. J., Peacock, G. L., Jr., Robinett, D. G., Sbaver, P. L., Brown, J. R., Herrick, J. E., Sanchez, H., & Havstad, K. M. (2009). State-and-transition models for heterogeneous landscapes: A strategy for development and application. Rangeland Ecology & Management, 62(1), 1-15.

Bickford, S., & Mackey, B. (2004). Reconstructing pre-impact vegetation cover in modified landscapes using environmental modelling, historical surveys and remnant vegetation data: A case study in the Fleurieu Peninsula, South Australia. Journal of Biogeography, 31(5), 787-805.

Bitterlich, W. (1948). Die Winkelzählprobe. Allgemeine Forstund Holzwirtschaftliche Zeitung, 59, 4–5 cited in Fastie, C. L. (2010). Estimating stand basal area from forest panoramas.

Black, M. P., Mooney, S. D., & Haberle, S. G. (2007). The fire, human and climate nexus in the Sydney Basin, eastern Australia. Holocene, 17(4), 469-480.

Blair, R. B. (1999). Birds and butterflies along an urban gradient: Surrogate taxa for assessing biodiversity? Ecological Applications, 9(1), 164-170.

Bleby, K., Harvey, J., Garkaklis, M., & Martin, L. (2008). Resource Condition Monitoring—Native Vegetation Integrity Project Literature Review: Vegetation Condition Assessment, Monitoring & Evaluation. (Version 4). Western Australia: The Government of Western Australia.

180

Blomquist, S. M., & Hunter Jr, M. L. (2010). A multi-scale assessment of amphibian habitat selection: wood frog response to timber harvesting. Ecoscience, 17(3), 251- 264.

Blumstein, D. T., Mennill, D. J., Clemins, P., Girod, L., Yao, K., Patricelli, G., Deppe, J. L., Krakauer, A. H., Clark, C., Cortopassi, K. A., Hanser, S. F., McCowan, B., Ali, A. M., & Kirschel, A. N. G. (2011). Acoustic monitoring in terrestrial environments using microphone arrays: Applications, technological considerations and prospectus. Journal of Applied Ecology, 48(3), 758-767.

Bonn, A., Rodrigues, A. S., & Gaston, K. J. (2002). Threatened and endemic species: Are they good indicators of patterns of biodiversity on a national scale? Ecology Letters, 5(6), 733-741.

Borja, A., & Dauer, D. M. (2008). Assessing the environmental quality status in estuarine and coastal systems: Comparing methodologies and indices. Ecological Indicators, 8(4), 331-337.

Bormpoudakis, D., Sueur, J., & Pantis, J. D. (2013). Spatial heterogeneity of ambient sound at the habitat type level: Ecological implications and applications. Landscape Ecology, 28(3), 495-506.

Botta, A., & Foley, J. A. (2002). Effects of climate variability and disturbances on the Amazonian terrestrial ecosystems dynamics. Global Biogeochemical Cycles, 16(4).

Bowman, A. S., & Facelli, J. M. (2013). Fallen logs as sources of patchiness in chenopod shrublands of South Australia. Journal of Arid Environments, 97, 66-72.

Bowman, D., Walsh, A., & Prior, L. D. (2004). Landscape analysis of Aboriginal fire management in Central Arnhem Land, north Australia. Journal of Biogeography, 31(2), 207-223.

Bowman, D., Wilson, B. A., & McDonough, L. (1991). Monsoon forests in Northwestern Australia—vegetation classification and the environmental control of tree species. Journal of Biogeography, 18(6), 679-686.

181

Boyce, M. S. (1992). Population viability analysis. Annual Review of Ecology and Systematics, 481-506.

Bradshaw, C. J. (2012). Little left to lose: Deforestation and forest degradation in Australia since European colonization. Journal of Plant Ecology, 5(1), 109-120.

Bray, J. R., & Curtis, J. T. (1957). An ordination of the upland forest communities of southern Wisconsin. Ecological Monographs, 27(4), 325-349.

Briske, D. D., Fuhlendorf, S. D., & Smeins, F. (2005). State-and-transition models, thresholds, and rangeland health: A synthesis of ecological concepts and perspectives. Rangeland Ecology & Management, 58(1), 1-10.

Brotons, L., Thuiller, W., Araújo, M. B., & Hirzel, A. H. (2004). Presence‐absence versus presence‐only modelling methods for predicting bird habitat suitability. Ecography, 27(4), 437-448.

Brown, P. M., & Wu, R. (2005). Climate and disturbance forcing of episodic tree recruitment in a southwestern ponderosa pine landscape. Ecology, 86(11), 3030- 3038.

Brubaker, L. B. (1986). Responses of tree populations to climatic change. Vegetatio, 67(2), 119-130.

Bureau of Meteorology, Australian Government (2013). Climate Data Online [Excel files]. Retrieved from: http://www.bom.gov.au/climate/data/

Burrows, G. E. (2002). Epicormic strand structure in Angophora, Eucalyptus and Lophostemon (Myrtaceae)—implications for fire resistance and recovery. New Phytologist, 153(1), 111-131.

Buxbaum, C., & Vanderbilt, K. (2007). Soil heterogeneity and the distribution of desert and steppe plant species across a desert-grassland ecotone. Journal of Arid Environments, 69(4), 617-632.

Cabrini, I., Grella, M. D., Andrade, C. F., & Thyssen, P. J. (2013). Richness and composition of Calliphoridae in an Atlantic Forest fragment: Implication for the use

182

of dipteran species as bioindicators. Biodiversity and Conservation, 22(11), 2635- 2643.

Callaghan, T. V., Bjorn, L. O., Chernov, Y., Chapin, T., Christensen, T. R., Huntley, B., Ims, R. A., Johansson, M., Jolly, D., Jonasson, S., Matveyeva, N., Panikov, N., Oechel, W., & Shaver, G. (2004). Effects on the function of arctic ecosystems in the short- and long-term perspectives. Ambio, 33(7), 448-458.

Capotorti, G., Guida, D., Siervo, V., Smiraglia, D., & Blasi, C. (2012). Ecological classification of land and conservation of biodiversity at the national level: The case of Italy. Biological Conservation, 147(1), 174-183.

Carmel, Y., & Stoller-Cavari, L. (2006). Comparing environmental and biological surrogates for biodiversity at a local scale. Israel Journal of Ecology & Evolution, 52(1), 11-27.

Caro, T. M., & O'Doherty, G. (1999). On the use of surrogate species in conservation biology. Conservation Biology, 13(4), 805-814.

Carrascal, L. M., Seoane, J., Palomino, D., & Polo, V. (2008). Explanations for bird species range size: Ecological correlates and phylogenetic effects in the Canary Islands. Journal of Biogeography, 35(11), 2061-2073.

Catford, J. A., Vesk, P. A., White, M. D., & Wintle, B. A. (2011). Hotspots of plant invasion predicted by propagule pressure and ecosystem characteristics. Diversity and Distributions, 17(6), 1099-1110.

Catterall, C. P., Kingston, M. B., Park, K., & Sewell, S. (1998). Deforestation, urbanisation and seasonality: Interacting effects on a regional bird assemblage. Biological Conservation, 84(1), 65-81.

Celis‐Murillo, A., Deppe, J. L., & Allen, M. F. (2009). Using soundscape recordings to estimate bird species abundance, richness, and composition. Journal of Field Ornithology, 80(1), 64-78.

183

CGIAR Consortium for Spatial Information (2009). Global Aridity and PET Database [ESRI raster map file]. Retrieved from: http://www.cgiar-csi.org/data/global-aridity- and-pet-database

Chace, J. F., & Walsh, J. J. (2006). Urban effects on native avifauna: A review. Landscape and Urban Planning, 74(1), 46-69.

Chan, K. (2004). Effect of patch size and bird aggression on bird species richness: A community-based project in tropical/subtropical eucalypt woodland. Royal Society of Queensland.

Chase, M. K., Kristan, W. B., Lynam, A. J., Price, M. V., & Rotenberry, J. T. (2000). Single species as indicators of species richness and composition in California coastal sage scrub birds and small mammals. Conservation Biology, 14(2), 474-487.

Chesson, P. (2000). Mechanisms of maintenance of species diversity. Annual Review of Ecology and Systematics, 31, 343-366.

Chesson, P., & Huntly, N. (1989). Short-term instabilities and long-term community dynamics. Trends in Ecology & Evolution, 4(10), 293-298.

Chevan, A., & Sutherland, M. (1991). Hierarchical partitioning. The American Statistician, 45(2), 90-96.

Clarke, M. F., & Oldland, J. M. (2007). Penetration of remnant edges by Noisy Miners (Manorina melanocephala) and implications for habitat restoration. Wildlife Research, 34(4), 253-261.

Clements, F. E. (1916). Plant Succession: An Analysis of the Development of Vegetation. Washington D.C.: Carnegie Institution of Washington cited in van der Maarel, E. (2005). Vegetation Ecology. United Kingdom: Blackwell Publishing.

Close, D., Davidson, N., Johnson, D., Abrams, M., Hart, S., Lunt, I., Archibald, R., Horton, B., & Adams, M. (2009). Premature decline of Eucalyptus and altered ecosystem processes in the absence of fire in some Australian forests. The Botanical Review, 75(2), 191-202.

184

Close, D. C., Davidson, N. J., & Watson, T. (2008). Health of remnant woodlands in fragments under distinct grazing regimes. Biological Conservation, 141(9), 2395- 2402.

Cochrane, M. A., & Schulze, M. D. (1999). Fire as a recurrent event in tropical forests of the Eastern Amazon: Effects on forest structure, biomass, and species composition. Biotropica, 31(1), 2-16.

Collins, L., Bradstock, R. A., Tasker, E. M., & Whelan, R. J. (2012). Impact of fire regimes, logging and topography on hollows in fallen logs in eucalypt forest of south eastern Australia. Biological Conservation, 149(1), 23-31.

Connell, J. H., & Slatyer, R. O. (1977). Mechanisms of succession in natural communities and their role in community stability and organization. American Naturalist, 1119-1144.

Cook, C. N., Wardell-Johnson, G., Keatley, M., Gowans, S. A., Gibson, M. S., Westbrooke, M. E., & Marshall, D. J. (2010). Is what you see what you get? Visual vs. measured assessments of vegetation condition. Journal of Applied Ecology, 47(3), 650-661.

Corenblit, D., Steiger, J., Gurnell, A. M., Tabacchi, E., & Roques, L. (2009). Control of sediment dynamics by vegetation as a key function driving biogeomorphic succession within fluvial corridors. Earth Surface Processes and Landforms, 34(13), 1790-1810.

Costello, D. A., Lunt, I. D., & Williams, J. E. (2000). Effects of invasion by the indigenous shrub Acacia sophorae on plant composition of coastal grasslands in south-eastern Australia. Biological Conservation, 96(1), 113-121.

Cousins, S. A. O., Lavorel, S., & Davies, I. (2003). Modelling the effects of landscape pattern and grazing regimes on the persistence of plant species with high conservation value in grasslands in south-eastern Sweden. Landscape Ecology, 18(3), 315-332.

185

Cowling, R. M., & Heijnis, C. E. (2001). The identification of broad habitat units as biodiversity entities for systematic conservation planning in the Cape Floristic Region. South African Journal of Botany, 67(1), 15-38.

CSIRO Land & Water (2011a). Soil description [ESRI raster map file]. Retrieved from: http://www.asris.csiro.au/

CSIRO Land & Water (2011b). Plant available water [ESRI raster map file]. Retrieved from: http://www.asris.csiro.au/

CSIRO Land & Water. (2011c). Bulk density [ESRI raster map file]. Retrieved from: http://www.asris.csiro.au/

Currie, D. J. (1991). Energy and large-scale patterns of animal-and plant-species richness. American Naturalist, 27-49.

Curtis, J. T. (1959). The vegetation of Wisconsin: An ordination of plant communities: University of Wisconsin Press.

Davis, M. A., Grime, J. P., & Thompson, K. (2000). Fluctuating resources in plant communities: A general theory of invasibility. Journal of Ecology, 88(3), 528-534. de Lima, R. F., Dallimer, M., Atkinson, P. W., & Barlow, J. (2013). Biodiversity and land- use change: Understanding the complex responses of an endemic-rich bird assemblage. Diversity and Distributions, 19(4), 411-422.

Department of Environment, Australian Government (2009). Australian Natural Resources Atlas. Retrieved 4 June 2013, from http://www.anra.gov.au/

Department of Environment, Australian Government (2014). Australia's bioregions (IBRA). Retrieved 14 May 2014, from http://www.environment.gov.au/land/nrs/science/ibra

Department of Environment and Heritage Protection, Queensland Government (2010a). Landcover map for Queensland, pre-clearing vegetation extent [ESRI shapefile]. Pre-clearing vegetation extent Queensland.

186

Department of Environment and Heritage Protection, Queensland Government (2010b). Landcover map for Queensland, current vegetation extent (2009) [ESRI shapefile]. Current vegetation extent Queensland (2009).

Department of Environment and Heritage Protection, Queensland Government (2012a). Current land use (2012) [ESRI shapefile]. Retrieved from: http://qldspatial.information.qld.gov.au/catalogue/custom/index.page

Department of Environment and Heritage Protection, Queensland Government (2012b). Current road extent (2012) [ESRI shapefile]. Retrieved from: http://qldspatial.information.qld.gov.au/catalogue/custom/index.page

Department of Environment and Heritage Protection, Queensland Government (2014a). Regional Ecosystem Descriptions, Regional Ecosystem Details 12.11.5e. Retrieved 20 July 2014, from https://environment.ehp.qld.gov.au/regional- ecosystems/details/?re=12.11.5

Department of Environment and Heritage Protection, Queensland Government (2014b). Regional Ecosystem Descriptions, Regional Ecosystem Details 12.5.3. Retrieved 20 July 2014, from https://environment.ehp.qld.gov.au/regional- ecosystems/details/?re=12.5.3

Department of Environment and Heritage Protection, Queensland Government (2014c). Fire Management Guidelines. Retrieved 30 May, 2015, from http://www.qld.gov.au/environment/plants-animals/plants/ecosystems/fire- management/

Department of Environment and Heritage Protection, Queensland Government (2015). Regional Ecosystem Descriptions. Retrieved 7 November 2012, from https://environment.ehp.qld.gov.au/regional-ecosystems/

Department of Sustainability and Environment, Victorian Government (2004). Vegetation Quality Assessment Manual—Guidelines for applying the habitat hectares scoring method. Melbourne.

187

Depraetere, M., Pavoine, S., Jiguet, F., Gasc, A., Duvail, S., & Sueur, J. (2012). Monitoring animal diversity using acoustic indices: Implementation in temperate woodland. Ecological Indicators, 13(1), 46-54.

Di Minin, E., & Moilanen, A. (2014). Improving the surrogacy effectiveness of charismatic megafauna with well-surveyed taxonomic groups and habitat types. Journal of Applied Ecology, 51(2), 281-288.

Diaz, L. (2006). Influences of forest type and forest structure on bird communities in oak and pine woodlands in Spain. Forest Ecology and Management, 223(1-3), 54-65.

Diaz, S., Hodgson, J. G., Thompson, K., Cabido, M., Cornelissen, J. H. C., Jalili, A., Montserrat-Marti, G., Grime, J. P., Zarrinkamar, F., Asri, Y., Band, S. R., Basconcelo, S., Castro-Diez, P., Funes, G., Hamzehee, B., Khoshnevi, M., Perez-Harguindeguy, N., Perez-Rontome, M. C., Shirvany, F. A., Vendramini, F., Yazdani, S., Abbas-Azimi, R., Bogaard, A., Boustani, S., Charles, M., Dehghan, M., de Torres-Espuny, L., Falczuk, V., Guerrero-Campo, J., Hynd, A., Jones, G., Kowsary, E., Kazemi-Saeed, F., Maestro- Martinez, M., Romo-Diez, A., Shaw, S., Siavash, B., Villar-Salvador, P., & Zak, M. R. (2004). The plant traits that drive ecosystems: Evidence from three continents. Journal of Vegetation Science, 15(3), 295-304.

Dikou, A., Papapanagiotou, E., & Troumbis, A. (2011). Integrating Landscape Ecology and Geoinformatics to Decipher Landscape Dynamics for Regional Planning. Environmental Management, 48(3), 523-538.

Dixon, I., Douglas, M., Dowe, J., & Burrows, D. (2006). Tropical Rapid Appraisal of Riparian Condition: Land & Water Australia.

Dorrough, J., & Scroggie, M. P. (2008). Plant responses to agricultural intensification. Journal of Applied Ecology, 45(4), 1274-1283.

Dray, S., Dufour, A., & Thioulouse, J. (2014). ade4: Analysis of Ecological Data: Exploratory and Euclidean Methods in Environmental Sciences (Version 1.6-2): R package.

188

Drever, M. C., Aitken, K. E., Norris, A. R., & Martin, K. (2008). Woodpeckers as reliable indicators of bird richness, forest health and harvest. Biological Conservation, 141(3), 624-634.

Driscoll, D. A., Lindenmayer, D. B., Bennett, A. F., Bode, M., Bradstock, R. A., Cary, G. J., Clarke, M. F., Dexter, N., Fensham, R., Friend, G., Gill, M., James, S., Kay, G., Keith, D. A., MacGregor, C., Russell-Smith, J., Salt, D., Watson, J. E. M., Williams, R. J., & York, A. (2010). Fire management for biodiversity conservation: Key research questions and our capacity to answer them. Biological Conservation, 143(9), 1928- 1939.

Driscoll, D. A., & Weir, T. (2005). Beetle responses to habitat fragmentation depend on ecological traits, habitat condition, and remnant size. Conservation Biology, 19(1), 182-194.

Dunning, J. B., Danielson, B. J., & Pulliam, H. R. (1992). Ecological processes that affect populations in complex landscapes. Oikos, 65(1), 169-175.

Dures, S. G., & Cumming, G. S. (2010). The confounding influence of homogenising invasive species in a globally endangered and largely urban biome: Does habitat quality dominate avian biodiversity? Biological Conservation, 143(3), 768-777.

Eberbach, P. L., & Burrows, G. E. (2006). The transpiration response by four topographically distributed Eucalyptus species, to rainfall occurring during drought in south eastern Australia. Physiologia Plantarum, 127(3), 483-493.

Egan, J., & Williams, R. (1996). Lifeform distributions of woodland plant species along a moisture availability gradient in Australia's monsoonal tropics. Australian Systematic Botany, 9(2), 205-217.

Eglington, S. M., Noble, D. G., & Fuller, R. J. (2012). A meta-analysis of spatial relationships in species richness across taxa: Birds as indicators of wider biodiversity in temperate regions. Journal for Nature Conservation, 20(5), 301-309.

189

Ehrenfeld, J. G., Han, X. G., Parsons, W. F. J., & Zhu, W. X. (1997). On the nature of environmental gradients: Temporal and spatial variability of soils and vegetation in the New Jersey Pinelands. Journal of Ecology, 85(6), 785-798.

ESRI (2013). ArcGIS Desktop (Version 10.1). Redlands, California: Environmental Systems Research Institute.

EUCLID Eucalypts of Australia (2014a). Angophora woodsiana. Retrieved 17 September, 2014, from https://www.anbg.gov.au/cpbr/cd- keys/euclid3/euclidsample/html/Angophora_woodsiana.htm

EUCLID Eucalypts of Australia (2014b). Corymbia citriodora. Retrieved October 1, 2014, from https://www.anbg.gov.au/cpbr/cd- keys/euclid3/euclidsample/html/Corymbia_citriodora.htm

Eyre, T., Kelly, A., & Neldner, V. (2011a). Method for the Establishment and Survey of Reference Sites for BioCondition. (Vol. 2.0). Brisbane: Department of Environment and Resource Management (DERM), Biodiversity and Ecological Sciences Unit.

Eyre, T., Kelly, A., Neldner, V., Wilson, B., Ferguson, D., Laidlaw, M., & Franks, A. (2011b). BioCondition: A Condition Assessment Framework for Terrestrial Biodiversity in Queensland. Assessment Manual. Version 2.1. Queensland Herbarium, Department of Science. Information Technology, Innovation and Arts, Brisbane.

Fahrig, L. (2003). Effects of habitat fragmentation on biodiversity. Annual review of ecology, evolution, and systematics, 487-515.

Fahrig, L., & Merriam, G. (1994). Conservation of fragmented populations. Conservation Biology, 8(1), 50-59.

Faith, D. P., Ferrier, S., & Walker, P. (2004). The ED strategy: How species‐level surrogates indicate general biodiversity patterns through an ‘environmental diversity’ perspective. Journal of Biogeography, 31(8), 1207-1217.

190

Faith, D. P., Margules, C. R., Walker, P. A., Stein, J., & Natera, G. (2001). Practical application of biodiversity surrogates and percentage targets for conservation in Papua New Guinea. Pacific Conservation Biology, 6(4), 289-303.

Falkiner, R., Nambiar, E., Polglase, P., Theiveyanathan, S., & Stewart, L. (2006). Root distribution of Eucalyptus grandis and Corymbia maculata in degraded saline soils of south-eastern Australia. Agroforestry systems, 67(3), 279-291.

Fang, J., Shen, Z., Tang, Z., Wang, X., Wang, Z., Feng, J., Liu, Y., Qiao, X., Wu, X., & Zheng, C. (2012). Forest community survey and the structural characteristics of forests in China. Ecography, 35(12), 1059-1071.

Farina, A., Lattanzi, E., Malavasi, R., Pieretti, N., & Piccioli, L. (2011a). Avian soundscapes and cognitive landscapes: Theory, application and ecological perspectives. Landscape Ecology, 26(9), 1257-1267.

Farina, A., Pieretti, N., & Piccioli, L. (2011b). The soundscape methodology for long- term bird monitoring: A Mediterranean Europe case-study. Ecological Informatics, 6(6), 354-363.

Favero, S., Souza, H., & Oliveira, A. (2011). Coleoptera (Insecta) as forest fragmentation indicators in the Rio Negro sub-region of the Pantanal, Mato Grosso do Sul, Brazil. Brazilian Journal of Biology, 71(1), 291-295.

Feld, C. K., Sousa, J. P., da Silva, P. M., & Dawson, T. P. (2010). Indicators for biodiversity and ecosystem services: Towards an improved framework for ecosystems assessment. Biodiversity and Conservation, 19(10), 2895-2919.

Ferdinands, K., Beggs, K., & Whitehead, P. (2005). Biodiversity and invasive grass species: Multiple-use or monoculture? Wildlife Research, 32(5), 447-457.

Ferrier, S. (2002). Mapping spatial pattern in biodiversity for regional conservation planning: Where to from here? Systematic biology, 51(2), 331-363.

Ferrier, S. (2005). An integrated approach to addressing terrestrial biodiversity in NRM investment planning and evaluation across state, catchment and property scales.

191

Department of Environment and Conservation, NSW. Available at: http://www.nrc.nsw.gov.au/content/documents/Submission

Ferrier, S., & Watson, G. (1997). An evaluation of the effectiveness of environmental surrogates and modelling techniques in predicting the distribution of biological diversity: Environment Australia.

Ferris, R., & Humphrey, J. W. (1999). A review of potential biodiversity indicators for application in British forests. Forestry, 72(4), 313-328.

Field, G. J., Burns, G. L., & Dale, P. E. (2012). Managing vegetation clearing in the South East Queensland urban footprint. Local Government Law Journal, 17, 215-230.

Fischer, J., Sherren, K., Stott, J., Zerger, A., Warren, G., & Stein, J. (2010). Toward landscape-wide conservation outcomes in Australia's temperate grazing region. Frontiers in Ecology and the Environment, 8(2), 69-74.

Flaspohler, D. J., Giardina, C. P., Asner, G. P., Hart, P., Price, J., Lyons, C. K. a., & Castaneda, X. (2010). Long-term effects of fragmentation and fragment properties on bird species richness in Hawaiian forests. Biological Conservation, 143(2), 280- 288.

Florabank. (2014). Corymbia citriodora. Retrieved October 1, 2014, 2014, from http://www.florabank.org.au/lucid/key/species%20navigator/media/html/Corymbi a_citriodora.htm

Flory, S. L., & Clay, K. (2009). Effects of roads and forest successional age on experimental plant invasions. Biological Conservation, 142(11), 2531-2537.

Forman, R. T., & Alexander, L. E. (1998). Roads and their major ecological effects. Annual Review of Ecology and Systematics, 207-231.

Francis, C. D., Paritsis, J., Ortega, C. P., & Cruz, A. (2011). Landscape patterns of avian habitat use and nest success are affected by chronic gas well compressor noise. Landscape Ecology, 26(9), 1269-1280.

192

Fraser, S. E. M., Beresford, A. E., Peters, J., Redhead, J. W., Welch, A. J., Mayhew, P. J., & Dytham, C. (2009). Effectiveness of vegetation surrogates for parasitoid wasps in reserve selection. Conservation Biology, 23(1), 142-150.

Gage, S.H. (2008). MATLAB biophony code: MATLAB.

Gage, S. H., Napoletano, B. M., & Cooper, M. C. (2001). Assessment of ecosystem biodiversity by acoustic diversity indices. Journal of the Acoustical Society of America, 109(5), 2430-2430.

Gardner, T. A. (2006). Tree-grass coexistence in the Brazilian cerrado: Demographic consequences of environmental instability. Journal of Biogeography, 33(3), 448- 463.

Ghilarov, A. M. (2000). Ecosystem functioning and intrinsic value of biodiversity. Oikos, 90(2), 408-412.

Gibbons, P., Briggs, S., Ayers, D., Seddon, J., Doyle, S., Cosier, P., McElhinny, C., Pelly, V., & Roberts, K. (2009). An operational method to assess impacts of land clearing on terrestrial biodiversity. Ecological Indicators, 9(1), 26-40.

Gibbons, P., Briggs, S. V., Ayers, D. A., Doyle, S., Seddon, J., McElhinny, C., Jones, N., Sims, R., & Doody, J. S. (2008). Rapidly quantifying reference conditions in modified landscapes. Biological Conservation, 141(10), 2483-2493.

Gibbons, P., & Freudenberger, D. (2006). An overview of methods used to assess vegetation condition at the scale of the site. Ecological Management & Restoration, 7(Suppl. 1), S10-S17.

Giesecke, T., Miller, P. A., Sykes, M. T., Ojala, A. E. K., Seppa, H., & Bradshaw, R. H. W. (2010). The effect of past changes in inter-annual temperature variability on tree distribution limits. Journal of Biogeography, 37(7), 1394-1405.

Gilfedder, L., & Kirkpatrick, J. B. (1998). Factors influencing the integrity of remnant bushland in subhumid Tasmania. Biological Conservation, 84(1), 89-96.

193

Gill, A. M., & Williams, J. E. (1996). Fire regimes and biodiversity: The effects of fragmentation of southeastern Australian eucalypt forests by urbanisation, agriculture and pine plantations. Forest Ecology and Management, 85(1–3), 261- 278.

Gil-Tena, A., Saura, S., & Brotons, L. (2007). Effects of forest composition and structure on bird species richness in a Mediterranean context: Implications for forest ecosystem management. Forest Ecology and Management, 242(2–3), 470-476.

Glass, G. V., Peckham, P. D., & Sanders, J. R. (1972). Consequences of failure to meet assumptions underlying the fixed effects analyses of variance and covariance. Review of educational research, 237-288.

Gleason, H. A. (1926). The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club, 53, 1-20 cited in van der Maarel, E. (2005). Vegetation Ecology. United Kingdom: Blackwell Publishing.

Gonzalez, A., & De Feo, O. (2007). Environmental variability modulates the insurance effects of diversity in non-equilibrium communities. The Impact of Environmental Variability on Ecological Systems (pp. 159-177): Springer.

Gonzalez, A., & Loreau, M. (2009). The causes and consequences of compensatory dynamics in ecological communities. Annual Review of Ecology, Evolution and Systematics, 40, 393-414.

Gooden, B., French, K., Turner, P. J., & Downey, P. O. (2009). Impact threshold for an alien plant invader, Lantana camara L., on native plant communities. Biological Conservation, 142(11), 2631-2641.

Google Earth (2013). South-East Queensland, Australia (Version 7.1.2.2041): Google. Retrieved from http://www.earth.google.com

Gosper, C. R., & Vivian-Smith, G. (2006). Selecting replacements for invasive plants to support frugivores in highly modified sites: A case study focusing on Lantana camara. Ecological Management & Restoration, 7(3), 197-203.

194

Gould, W. A., Gonzalez, G., Hudak, A. T., Hollingsworth, T. N., & Hollingsworth, J. (2008). Forest structure and downed woody debris in boreal, temperate, and tropical forest fragments. Ambio, 37(7-8), 577-587.

Grime, J. (1973). Control of species density in herbaceous vegetation. Journal of Environmental Management, 1, 151-167.

Gude, P. H., Hansen, A. J., Rasker, R., & Maxwell, B. (2006). Rates and drivers of rural residential development in the Greater Yellowstone. Landscape and Urban Planning, 77(1), 131-151.

Guo, Z. W., Gan, Y. L., & Li, Y. M. (2003). Spatial pattern of ecosystem function and ecosystem conservation. Environmental Management, 32(6), 682-692.

Hacking, J., Abom, R., & Schwarzkopf, L. (2014). Why do lizards avoid weeds? Biological Invasions, 16(4), 935-947.

Haines-Young, R. (2009). Land use and biodiversity relationships. Land Use Policy, 26, S178-S186.

Halford, J. J., & Fensham, R. J. (2014). Vegetation and environmental relations of ephemeral subtropical wetlands in central Queensland, Australia. Australian Journal of Botany, 62(6), 499-510.

Hammer, O. (2013). PAST (Version 3): Retrieved from http://folk.uio.no/ohammer/past/

Hampel, M., Moreno-Garrido, I., Gonzalez-Mazo, E., & Blasco, J. (2009). Suitability of the marine prosobranch snail Hydrobia ulvae for sediment toxicity assessment: A case study with the anionic surfactant linear alkylbenzene sulphonate (LAS). Ecotoxicology and Environmental Safety, 72(4), 1303-1308.

Hansen, A. J., Knight, R. L., Marzluff, J. M., Powell, S., Brown, K., Gude, P. H., & Jones, A. (2005). Effects of exurban development on biodiversity: Patterns, mechanisms, and research needs. Ecological Applications, 15(6), 1893-1905.

Hanski, I. (1998). Metapopulation dynamics. Nature, 396(6706), 41-49.

195

Harborne, A. R., Mumby, P. J., Kappel, C. V., Dahlgren, C. P., Micheli, F., Holmes, K. E., & Brumbaugh, D. R. (2008). Tropical coastal habitats as surrogates of fish community structure, grazing, and fisheries value. Ecological Applications, 18(7), 1689-1701.

Harmon, M. E., Franklin, J. F., Swanson, F. J., Sollins, P., Gregory, S., Lattin, J., Anderson, N., Cline, S., Aumen, N., & Sedell, J. (1986). Ecology of coarse woody debris in temperate ecosystems. Advances in ecological research, 15(133), 302.

Hasui, E., Gomes, V. S. D., & Silva, W. R. (2007). Effects of vegetation traits on habitat preferences of frugivorous birds in Atlantic rain forest. Biotropica, 39(4), 502-509.

Hawkins, C. (2006). Quantifying biological integrity by taxonomic completeness: evaluation of a potential indicator for use in regional-and global-scale assessments. Ecological Applications, 16, 1251.

Hawkins, C. P., Olson, J. R., & Hill, R. A. (2010). The reference condition: Predicting benchmarks for ecological and water-quality assessments. Journal of the North American Benthological Society, 29(1), 312-343.

Hawksworth, D. L. (2011). Biodiversity and conservation of insects and other invertebrates. Biodiversity and Conservation, 20(13), 2863-2866.

Heino, J., Tolonen, K. T., Kotanen, J., & Paasivirta, L. (2009). Indicator groups and congruence of assemblage similarity, species richness and environmental relationships in littoral macroinvertebrates. Biodiversity and Conservation, 18(12), 3085-3098.

Helson, J. E., & Williams, D. D. (2013). Development of a macroinvertebrate multimetric index for the assessment of low-land streams in the neotropics. Ecological Indicators, 29, 167-178.

Herrick, J., Bestelmeyer, B., Archer, S., Tugel, A., & Brown, J. (2006). An integrated framework for science-based arid land management. Journal of Arid Environments, 65(2), 319-335.

196

Higgins, S. I., Pickett, S. T. A., & Bond, W. J. (2000). Predicting extinction risks for plants: Environmental stochasticity can save declining populations. Trends in Ecology & Evolution, 15(12), 516-520.

Hill, S. J., Tung, P. J., & Leishman, M. R. (2005). Relationships between anthropogenic disturbance, soil properties and plant invasion in endangered Cumberland Plain Woodland, Australia. Austral Ecology, 30(7), 775-788.

Hoffmann, W. A., & Jackson, R. B. (2000). Vegetation-climate feedbacks in the conversion of tropical savanna to grassland. Journal of Climate, 13(9), 1593-1602.

Holdaway, R. J., Richardson, S. J., Dickie, I. A., Peltzer, D. A., & Coomes, D. A. (2011). Species‐and community‐level patterns in fine root traits along a 120 000‐year soil chronosequence in temperate rain forest. Journal of Ecology, 99(4), 954-963.

Holmgren, M., Hirota, M., van Nes, E. H., & Scheffer, M. (2013). Effects of interannual climate variability on tropical tree cover. Nature Climate Change, 3(8), 755-758.

Hong, S.-K., Kim, S., Cho, K.-H., Kim, J.-E., Kang, S., & Lee, D. (2004). Ecotope mapping for landscape ecological assessment of habitat and ecosystem. Ecological Research, 19(1), 131-139.

Hooper, D. U., Chapin, F. S., Ewel, J. J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J. H., Lodge, D. M., Loreau, M., Naeem, S., Schmid, B., Setala, H., Symstad, A. J., Vandermeer, J., & Wardle, D. A. (2005). Effects of biodiversity on ecosystem functioning: A consensus of current knowledge. Ecological Monographs, 75(1), 3-35.

Horn, S., & Hanula, J. L. (2008). Relationship of coarse woody debris to arthropod availability for red-cockaded woodpeckers and other bark-foraging birds on loblolly pine boles. Journal of Entomological Science, 43(2), 153-168.

Horner, G. J., Cunningham, S. C., Thomson, J. R., Baker, P. J., & Mac Nally, R. (2012). Forest structure, flooding and grazing predict understorey composition of floodplain forests in southeastern Australia. Forest Ecology and Management, 286, 148-158.

197

Hothorn, T., Hornik, K., van de Wiel, M. A., & Zeileis, A. (2014). coin: Conditional Inference Procedures in a Permutation Test Framework (Version 1.0-24): R package.

Hughes, L. (2011). Climate change and Australia: Key vulnerable regions. Regional Environmental Change, 11, S189-S195.

Hutley, L. B., Beringer, J., Isaac, P. R., Hacker, J. M., & Cernusak, L. A. (2011). A sub- continental scale living laboratory: Spatial patterns of savanna vegetation over a rainfall gradient in northern Australia. Agricultural and Forest Meteorology, 151(11), 1417-1428.

IBM Corporation (2012). IBM SPSS Statistics for Windows (Version 21). Armonk, NY: IBM Corp.

Ikin, K., Barton, P. S., Stirnemann, I. A., Stein, J. R., Michael, D., Crane, M., Okada, S., & Lindenmayer, D. B. (2014). Multi-scale associations between vegetation cover and woodland bird communities across a large agricultural region. Plos One, 9(5).

Imai, N., Tanaka, A., Samejima, H., Sugau, J. B., Pereira, J. T., Titin, J., Kurniawan, Y., & Kitayama, K. (2014). Tree community composition as an indicator in biodiversity monitoring of REDD. Forest Ecology and Management, 313, 169-179.

Ives, A. R., & Carpenter, S. R. (2007). Stability and diversity of ecosystems. Science, 317(5834), 58-62.

Jackson, R. D., & Bartolome, J. W. (2002). A state-transition approach to understanding nonequilibrium plant community dynamics in Californian grasslands. Plant Ecology, 162(1), 49-65.

Jansen, A., & Robertson, A. I. (2005). Grazing, ecological condition and biodiversity in riparian river red gum forests in south-eastern Australia. Proceedings of the Royal Society of Victoria, 117(1), 85-95.

Jiang, J., Gao, D., & DeAngelis, D. L. (2012). Towards a theory of ecotone resilience: coastal vegetation on a salinity gradient. Theoretical Population Biology, 82(1), 29- 37.

198

Joo, W., Gage, S. H., & Kasten, E. P. (2011). Analysis and interpretation of variability in soundscapes along an urban-rural gradient. Landscape and Urban Planning, 103(3- 4), 259-276.

Jurskis, V., & Underwood, R. (2013). Human fires and wildfires on Sydney sandstones: History informs management. Fire Ecology, 9(3), 8-24.

Jurskis, V., & Walmsley, T. (2012). Eucalypt ecosystems predisposed to chronic decline: Estimated distribution in coastal New South Wales NSW, Australia: Bushfire Cooperative Research Centre.

Juutinen, A., & Mönkkönen, M. (2004). Testing alternative indicators for biodiversity conservation in old-growth boreal forests: Ecology and economics. Ecological Economics, 50(1), 35-48.

Juutinen, A., Mönkkönen, M., & Ollikainen, M. (2008). Do environmental diversity approaches lead to improved site selection? A comparison with the multi-species approach. Forest Ecology and Management, 255(11), 3750-3757.

Kappes, H., Topp, W., Zach, P., & Kulfan, J. (2006). Coarse woody debris, soil properties and snails (Mollusca : Gastropoda) in European primeval forests of different environmental conditions. European Journal of Soil Biology, 42(3), 139-146.

Karr, J. R. (1981). Assessment of biotic integrity using fish communities. Fisheries, 6(6), 21-27.

Kasten, E. P., Gage, S. H., Fox, J., & Joo, W. (2012). The remote environmental assessment laboratory's acoustic library: An archive for studying soundscape ecology. Ecological Informatics, 12(0), 50-67.

Kath, J., Maron, M., & Dunn, P. K. (2009). Interspecific competition and small bird diversity in an urbanizing landscape. Landscape and Urban Planning, 92(2), 72-79.

Kaufmann, P. R., Peck, D. V., Paulsen, S. G., Seeliger, C. W., Hughes, R. M., Whittier, T. R., & Kamman, N. C. (2014). Lakeshore and littoral physical habitat structure in a national lakes assessment. Lake and Reservoir Management, 30(2), 192-215.

199

Kelley, J., & Kelly, A. (2012). Calculating BioCondition v 2.1 Landscape attributes (patch size, connectivity and context) using ArcGIS 10 (software implementation as ArcGIS toolbox): A technical manual. V1.0. Unpublished: Queensland Herbarium.

Kelly, A. L., Franks, A. J., & Eyre, T. J. (2011). Assessing the assessors: Quantifying observer variation in vegetation and habitat assessment. Ecological Management & Restoration, 12(2), 144-148.

Kent, M., Gill, W. J., Weaver, R. E., & Armitage, R. P. (1997). Landscape and plant community boundaries in biogeography. Progress in Physical Geography, 21(3), 315-353.

Kerans, B. L., Karr, J. R., & Ahlstedt, S. A. (1992). Aquatic invertebrate assemblages— spatial and temporal differences among sampling protocols. Journal of the North American Benthological Society, 11(4), 377-390.

Khanaposhtani, M. G., Kaboli, M., Karami, M., & Etemad, V. (2012). Effect of habitat complexity on richness, abundance and distributional pattern of forest birds. Environmental Management, 50(2), 296-303.

Kilinc, M., & Beringer, J. (2007). The spatial and temporal distribution of lightning strikes and their relationship with vegetation type, elevation, and fire sears in the Northern Territory. Journal of Climate, 20(7), 1161-1173.

Kim, K. H., & Pauleit, S. (2005). Landscape metrics to assess the ecological conditions of city regions: Application to Kwangju City, South Korea. International Journal of Sustainable Development and World Ecology, 12(3), 227-244.

Kline, J. D., Moses, A., & Alig, R. J. (2001). Integrating urbanization into landscape-level ecological assessments. Ecosystems, 4(1), 3-18.

Knox, K., & Clarke, P. (2012). Fire severity, feedback effects and resilience to alternative community states in forest assemblages. Forest Ecology and Management, 265, 47-54.

200

Korb, J. E., Fule, P. Z., & Wu, R. (2013). Variability of warm/dry mixed conifer forests in southwestern Colorado, USA: Implications for ecological restoration. Forest Ecology and Management, 304, 182-191.

Krause, B., Gage, S. H., & Joo, W. (2011). Measuring and interpreting the temporal variability in the soundscape at four places in Sequoia National Park. Landscape Ecology, 26(9), 1247-1256.

Krauss, J., Bommarco, R., Guardiola, M., Heikkinen, R. K., Helm, A., Kuussaari, M., Lindborg, R., Ockinger, E., Partel, M., Pino, J., Poyry, J., Raatikainen, K. M., Sang, A., Stefanescu, C., Teder, T., Zobel, M., & Steffan-Dewenter, I. (2010). Habitat fragmentation causes immediate and time-delayed biodiversity loss at different trophic levels. Ecology Letters, 13(5), 597-605.

Kuehne, L. M., Padgham, B. L., & Olden, J. D. (2013). The Soundscapes of lakes across an urbanization gradient. Plos One, 8(2).

Kumar, S., Simonson, S. E., & Stohlgren, T. J. (2009). Effects of spatial heterogeneity on butterfly species richness in Rocky Mountain National Park, CO, USA. Biodiversity and Conservation, 18(3), 739-763.

Kusumoto, B., Shiono, T., Miyoshi, M., Maeshiro, R., Fujii, S.-j., Kuuluvainen, T., & Kubota, Y. (2015). Functional response of plant communities to clearcutting: Management impacts differ between forest vegetation zones. Journal of Applied Ecology, 52(1), 171-180.

Kyle, G. P., Beard, K. H., & Kulmatiski, A. (2007). Reduced soil compaction enhances establishment of non-native plant species. Plant Ecology, 193(2), 223-232.

Lafortezza, R., Coomes, D. A., Kapos, V., & Ewers, R. M. (2010). Assessing the impacts of fragmentation on plant communities in New Zealand: Scaling from survey plots to landscapes. Global Ecology and Biogeography, 19(5), 741-754.

Lam, T. Y., Fletcher, C., Ramage, B. S., Doll, H. M., Joann, C. L., Nur‐Zati, A. M., Butod, E., Kassim, A. R., Harrison, R. D., & Potts, M. D. (2014). Using habitat characteristics to predict faunal diversity in tropical production forests. Biotropica, 46(1), 50-57.

201

Lampila, P., Mönkkönen, M., & Desrochers, A. (2005). Demographic responses by birds to forest fragmentation. Conservation Biology, 19(5), 1537-1546.

Lancaster, M. L., Taylor, A. C., Cooper, S. J., & Carthew, S. M. (2011). Limited ecological connectivity of an arboreal marsupial across a forest/plantation landscape despite apparent resilience to fragmentation. Molecular ecology, 20(11), 2258-2271.

Landres, P. B., Verner, J., & Thomas, J. W. (1988). Ecological uses of vertebrate indicator species—a critique. Conservation Biology, 2(4), 316-328.

Landsberg, J., & Crowley, G. (2004). Monitoring rangeland biodiversity: Plants as indicators. Austral Ecology, 29(1), 59-77.

Larsen, F. W., Bladt, J., Balmford, A., & Rahbek, C. (2012). Birds as biodiversity surrogates: Will supplementing birds with other taxa improve effectiveness? Journal of Applied Ecology, 49(2), 349-356.

Laiolo P. (2010). The emerging significance of bioacoustics in animal species conservation. Biol Conserv 143:1635-1645.

Laughlin, D. C., Bakker, J. D., Stoddard, M. T., Daniels, M. L., Springer, J. D., Gildar, C. N., Green, A. M., & Covington, W. W. (2004). Toward reference conditions: Wildfire effects on flora in an old-growth ponderosa pine forest. Forest Ecology and Management, 199(1), 137-152.

Lawes, M. J., Richards, A., Dathe, J., & Midgley, J. J. (2011). Bark thickness determines fire resistance of selected tree species from fire-prone tropical savanna in north Australia. Plant Ecology, 212(12), 2057-2069.

Lawler, J., & White, D. (2008). Assessing the mechanisms behind successful surrogates for biodiversity in conservation planning. Animal Conservation, 11(4), 270-280.

Lawler, J. J., White, D., Sifneos, J. C., & Master, L. L. (2003). Rare species and the use of indicator groups for conservation planning. Conservation Biology, 17(3), 875-882.

Legendre, P., & Legendre, L. (1998). Numerical Ecology, Volume 24, (Developments in Environmental Modelling): Elsevier.

202

Leiper, G., Glazebrook, J., Cox, D., & Rathie, K. (2008). Mangroves to Mountains Revised Edition: A field guide to the native plants of South-east Queensland. Brisbane: Logan River Branch of the Society for Growing Australian Plants (QLD Region).

Lepetz, V., Massot, M., Schmeller, D. S., & Clobert, J. (2009). Biodiversity monitoring: Some proposals to adequately study species' responses to climate change. Biodiversity and Conservation, 18(12), 3185-3203.

Letten, A. D., Ashcroft, M. B., Keith, D. A., Gollan, J. R., & Ramp, D. (2013). The importance of temporal climate variability for spatial patterns in plant diversity. Ecography, 36(12), 1341-1349.

Levine, J. M., McEachern, A. K., & Cowan, C. (2008). Rainfall effects on rare annual plants. Journal of Ecology, 96(4), 795-806.

Lewis, T., & Debuse, V. J. (2012). Resilience of a eucalypt forest woody understorey to long-term (34–55 years) repeated burning in subtropical Australia. International Journal of Wildland Fire, 21(8), 980-991.

Lindenmayer, D. B., Barton, P. S., Lane, P. W., Westgate, M. J., McBurney, L., Blair, D., Gibbons, P., & Likens, G. E. (2014). An empirical assessment and comparison of species-based and habitat-based surrogates: A case study of forest vertebrates and large old trees. Plos One, 9(2).

Lindenmayer, D. B., Cunningham, R. B., Donnelly, C. F., & Lesslie, R. (2002). On the use of landscape surrogates as ecological indicators in fragmented forests. Forest Ecology and Management, 159(3), 203-216.

Lindenmayer, D. B., & Likens, G. E. (2010). The science and application of ecological monitoring. Biological Conservation, 143(6), 1317-1328.

Lindenmayer, D. B., Margules, C. R., & Botkin, D. B. (2000). Indicators of biodiversity for ecologically sustainable forest management. Conservation Biology, 14(4), 941- 950.

203

Lindsay, E. A., & French, K. (2005). Litterfall and nitrogen cycling following invasion by Chrysanthemoides monilifera ssp rotundata in coastal Australia. Journal of Applied Ecology, 42(3), 556-566.

Lombard, A. T., Cowling, R. M., Pressey, R. L., & Rebelo, A. G. (2003). Effectiveness of land classes as surrogates for species in conservation planning for the Cape Floristic Region. Biological Conservation, 112(1-2), 45-62.

Lopez, R. D., & Siobhan Fennessy, M. (2002). Testing the floristic quality assessment index as an indicator of wetland condition. Ecological Applications, 12(2), 487-497.

Lopez-Martinez, J. O., Hernandez-Stefanoni, J. L., Dupuy, J. M., & Meave, J. A. (2013). Partitioning the variation of woody plant beta-diversity in a landscape of secondary tropical dry forests across spatial scales. Journal of Vegetation Science, 24(1), 33-45.

Lunt, I. D. (1997). Effects of long-term vegetation management on remnant grassy forests and anthropogenic native grasslands in south-eastern Australia. Biological Conservation, 81(3), 287-297.

Lunt, I. D. (1998). Allocasuarina (Casuarinaceae) invasion of an unburnt coastal woodland at Ocean Grove, Victoria: Structural changes 1971–1996. Australian Journal of Botany, 46(6), 649-656.

Lunt, I. D. (2002). Grazed, burnt and, cleared: How ecologists have studied century- scale vegetation changes in Australia. Australian Journal of Botany, 50(4), 391-407.

Lunt, I. D., Eldridge, D. J., Morgan, J. W., & Witt, G. B. (2007). A framework to predict the effects of livestock grazing and grazing exclusion on conservation values in natural ecosystems in Australia. Australian Journal of Botany, 55(4), 401-415.

MacNally, R. (2002b). Multiple regression and inference in ecology and conservation biology: Further comments on identifying important predictor variables. Biodiversity & Conservation, 11(8), 1397-1401.

MacNally, R., Bennett, A. F., Brown, G. W., Lumsden, L. F., Yen, A., Hinkley, S., Lillywhite, P., & Ward, D. A. (2002a). How well do ecosystem-based planning units

204

represent different components of biodiversity? Ecological Applications, 12(3), 900- 912.

MacArthur, R. H. (1967). The theory of island biogeography (Vol. 1): Princeton University Press.

Malt, J., & Lank, D. (2007). Temporal dynamics of edge effects on nest predation risk for the marbled murrelet. Biological Conservation, 140(1-2), 160-173.

Malt, J. M., & Lank, D. B. (2009). Marbled Murrelet nest predation risk in managed forest landscapes: Dynamic fragmentation effects at multiple scales. Ecological Applications, 19(5), 1274-1287.

Mandelik, Y., Dayan, T., Chikatunov, V., & Kravchenko, V. (2012). The relative performance of taxonomic vs. environmental indicators for local biodiversity assessment: A comparative study. Ecological Indicators, 15(1), 171-180.

Mandelik, Y., Roll, U., & Fleischer, A. (2010). Cost‐efficiency of biodiversity indicators for Mediterranean ecosystems and the effects of socio‐economic factors. Journal of Applied Ecology, 47(6), 1179-1188.

Margules, C. R., Nicholls, A., & Austin, M. (1987). Diversity of Eucalyptus species predicted by a multi-variable environmental gradient. Oecologia, 71(2), 229-232.

Margules, C. R., & Pressey, R. L. (2000). Systematic conservation planning. Nature, 405(6783), 243-253.

Maron, M., Grey, M. J., Catterall, C. P., Major, R. E., Oliver, D. L., Clarke, M. F., Loyn, R. H., Mac Nally, R., Davidson, I., & Thomson, J. R. (2013). Avifaunal disarray due to a single despotic species. Diversity and Distributions, 19(12), 1468-1479.

Maron, M., & Lill, A. (2005). The influence of livestock grazing and weed invasion on habitat use by birds in grassy woodland remnants. Biological Conservation, 124(4), 439-450.

205

Martin, T. G., Carwardine, J., Broadhurst, L., Ferrier, S., James, C., Sheppard, A., Whitten, S., & Chades, I. (2014). Tools for managing and restoring biodiversity. Australia: CSIRO Publishing.

Mathews, S., & Bonser, S. P. (2005). Life histories, ecological tolerance limits, and the evolution of geographic range size in Eucalyptus (Myrtaceae). Australian Journal of Botany, 53(6), 501-508.

MATLAB (2012). MATLAB and statistics toolbox (version release 2012b). Natick, Massachusetts, United States: The MathWorks, Inc.

Matsinos, Y. G., Mazaris, A. D., Papadimitriou, K. D., Mniestris, A., Hatzigiannidis, G., Maioglou, D., & Pantis, J. D. (2008). Spatio-temporal variability in human and natural sounds in a rural landscape. Landscape Ecology, 23(8), 945-959.

Mazaris, A. D., Kallimanis, A. S., Chatzigianidis, G., Papadimitriou, K., & Pantis, J. D. (2009). Spatiotemporal analysis of an acoustic environment: Interactions between landscape features and sounds. Landscape Ecology, 24(6), 817-831.

McAlpine, C. A., Fensham, R. J., & Temple-Smith, D. E. (2002). Biodiversity conservation and vegetation clearing in Queensland: Principles and thresholds. Rangeland Journal, 24(1), 36-55.

McAlpine, C. A., Syktus, J., Ryan, J. G., Deo, R. C., McKeon, G. M., McGowan, H. A., & Phinn, S. R. (2009). A continent under stress: Interactions, feedbacks and risks associated with impact of modified land cover on Australia's climate. Global Change Biology, 15(9), 2206-2223.

McCarthy, M. A., Parris, K. M., Van Der Ree, R., McDonnell, M. J., Burgman, M. A., Williams, N. S., McLean, N., Harper, M. J., Meyer, R., & Hahs, A. (2004). The habitat hectares approach to vegetation assessment: An evaluation and suggestions for improvement. Ecological Management & Restoration, 5(1), 24-27.

McElhinny, C., Gibbons, P., Brack, C., & Bauhus, J. (2005). Forest and woodland stand structural complexity: Its definition and measurement. Forest Ecology and Management, 218(1-3), 1-24.

206

McGregor, M., & Burnett, S. (2014). Proximity to coarse woody debris increases reptile presence in an Australian subtropical rainforest remnant. Australian Zoologist, 37(2), 267-274.

Medina, N. G., Albertos, B., Lara, F., Mazimpaka, V., Garilleti, R., Draper, D., & Hortal, J. (2014). Species richness of epiphytic bryophytes: Drivers across scales on the edge of the Mediterranean. Ecography, 37(1), 80-93.

Menges, E. S. (2000). Population viability analyses in plants: Challenges and opportunities. Trends in Ecology & Evolution, 15(2), 51-56.

Mitchell, S. W., & Csillag, F. (2001). Assessing the stability and uncertainty of predicted vegetation growth under climatic variability: Northern mixed grass prairie. Ecological Modelling, 139(2-3), 101-121.

Mitchley, J., & Xofis, P. (2005). Landscape structure and management regime as indicators of calcareous grassland habitat condition and species diversity. Journal for Nature Conservation (Jena), 13(2-3), 171-183.

Mooney, S. D., Webb, M., & Attenbrow, V. (2007). A comparison of charcoal and archaeological information to address the influences on Holocene fire activity in the Sydney Basin. Australian Geographer, 38(2), 177-194.

Moreno-Rueda, G., & Pizarro, M. (2009). Relative influence of habitat heterogeneity, climate, human disturbance, and spatial structure on vertebrate species richness in Spain. Ecological Research, 24(2), 335-344.

Mörtberg, U. M. (2001). Resident bird species in urban forest remnants; landscape and habitat perspectives. Landscape Ecology, 16(3), 193-203.

Murphy, G. E., & Romanuk, T. N. (2014). A meta‐analysis of declines in local species richness from human disturbances. Ecology and evolution, 4(1), 91-103.

Nangula, S., & Oba, G. (2004). Effects of artificial water points on the Oshana ecosystem in Namibia. Environmental Conservation, 31(1), 47-54.

207

National Herbarium of New South Wales (2014a). PlantNET New South Wales flora online: Arundinella nepalensis Trin. Retrieved 29 August 2014, from http://plantnet.rbgsyd.nsw.gov.au/cgi- bin/NSWfl.pl?page=nswfl&lvl=sp&name=Arundinella~nepalensis

National Herbarium of New South Wales (2014b). PlantNET New South Wales flora online: Eucalyptus acmenoides Schauer. Retrieved 15 September, 2014, from http://plantnet.rbgsyd.nsw.gov.au/cgi- bin/NSWfl.pl?page=nswfl&lvl=sp&name=Eucalyptus~acmenoides

National Herbarium of New South Wales (2014c). PlantNET New South Wales flora online: Angophora woodsiana F. M. Bailey. Retrieved 17 September, 2014, from http://plantnet.rbgsyd.nsw.gov.au/cgi- bin/NSWfl.pl?page=nswfl&lvl=sp&name=Angophora~woodsiana

Neave, H. M., Cunningham, R. B., Norton, T. W., & Nix, H. A. (1996). Biological inventory for conservation evaluation—III. Relationships between birds, vegetation and environmental attributes in southern Australia. Forest Ecology and Management, 85(1-3), 197-218.

Neave, H. M., & Norton, T. W. (1998). Biological inventory for conservation evaluation—IV. Composition, distribution and spatial prediction of vegetation assemblages in southern Australia. Forest Ecology and Management, 106(2-3), 259- 281.

Neldner, V. J., Wilson, B. A., Thompson, E. J., & H.A., D. (2005). Methodology for survey and mapping of regional ecosystems and vegetation communities in Queensland. Brisbane: Queensland Government.

Neuhausl, R. (1990). Unified classification of European natural forests—the approach of the vegetation map of Europe. Vegetatio, 89(2), 173-181.

Ni, J., & Zhang, X. S. (2000). Climate variability, ecological gradient and the northeast China transect (NECT). Journal of Arid Environments, 46(3), 313-325.

208

Niemi, G. J., & McDonald, M. E. (2004). Application of ecological indicators. Annual review of ecology, evolution, and systematics, 89-111.

Nonaka, E., & Spies, T. A. (2005). Historical range of variability in landscape structure: A simulation study in Oregon, USA. Ecological Applications, 15(5), 1727-1746.

Norton, B. G. (2000). Biodiversity and environmental values: In search of a universal earth ethic. Biodiversity & Conservation, 9(8), 1029-1044.

Noss, R. F. (1999). Assessing and monitoring forest biodiversity: A suggested framework and indicators. Forest Ecology and Management, 115(2-3), 135-146.

Odgers, B. (1999). Seasonal variation in buried germinable seed banks of grass species in an urban eucalypt forest reserve. Australian Journal of Botany, 47(4), 623-638.

O'Donnell, J., Gallagher, R. V., Wilson, P. D., Downey, P. O., Hughes, L., & Leishman, M. R. (2012). Invasion hotspots for non-native plants in Australia under current and future climates. Global Change Biology, 18(2), 617-629.

Oksanen, M. (1997). The moral value of biodiversity. Ambio, 26(8), 541-545.

Oliver, I., Eldridge, D. J., Nadolny, C., & Martin, W. K. (2014). What do site condition multi-metrics tell us about species biodiversity? Ecological Indicators, 38, 262-271.

Oliver, I., Holmes, A., Dangerfield, J. M., Gillings, M., Pik, A. J., Britton, D. R., Holley, M., Montgomery, M. E., Raison, M., Logan, V., Pressey, R. L., & Beattie, A. J. (2004). Land systems as surrogates for biodiversity in conservation planning. Ecological Applications, 14(2), 485-503.

Oliver, I., Jones, H., & Schmoldt, D. L. (2007). Expert panel assessment of attributes for natural variability benchmarks for biodiversity. Austral Ecology, 32(4), 453-475.

Olsen, L. M., Dale, V. H., & Foster, T. (2007). Landscape patterns as indicators of ecological change at Fort Benning, Georgia, USA. Landscape and Urban Planning, 79(2), 137-149.

Opdam, P. (1991). Metapopulation theory and habitat fragmentation: A review of holarctic breeding bird studies. Landscape Ecology, 5(2), 93-106.

209

Osmond, C., Austin, M., Berry, J., Billings, W., Boyer, J., Dacey, J., Nobel, P., Smith, S., & Winner, W. (1987). Stress physiology and the distribution of plants. BioScience, 38- 48.

Palmer, G. C., Fitzsimons, J. A., Antos, M. J., & White, J. G. (2008). Determinants of native avian richness in suburban remnant vegetation: Implications for conservation planning. Biological Conservation, 141(9), 2329-2341.

Parkes, D., Newell, G., & Cheal, D. (2003). Assessing the quality of native vegetation: The ‘habitat hectares’ approach. Ecological Management & Restoration, 4(s1), S29- S38.

Pausas, J. G., & Austin, M. P. (2001). Patterns of plant species richness in relation to different environments: An appraisal. Journal of Vegetation Science, 12(2), 153-166.

Pearson, D. L. (1994). Selecting indicator taxa for the quantitative assessment of biodiversity. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 345(1311), 75-79.

Pearson, D. L., & Cassola, F. (1992). World‐wide species richness patterns of tiger beetles (Coleoptera: Cicindelidae): Indicator taxon for biodiversity and conservation studies. Conservation Biology, 6(3), 376-391.

Peet, R. K. (1978). Forest vegetation of the Colorado Front Range: patterns of species diversity. Vegetatio, 37(2), 65-78.

Pekin, B. K., Jung, J., Villanueva-Rivera, L. J., Pijanowski, B. C., & Ahumada, J. A. (2012). Modeling acoustic diversity using soundscape recordings and LIDAR-derived metrics of vertical forest structure in a neotropical rainforest. Landscape Ecology, 27(10), 1513-1522.

Perry, G. L. W., & Enright, N. J. (2002). Humans, fire and landscape pattern: Understanding a maquis-forest complex, Mont Do, New Caledonia, using a spatial 'state-and-transition' model. Journal of Biogeography, 29(9), 1143-1158.

210

Perry, N. (2013). The precautionary principle, uncertainty and the Noah’s Ark problem. Wildlife Research, 40(2), 117-125.

Pert, P. L., Butler, J. R. A., Bruce, C., & Metcalfe, D. (2012). A composite threat indicator approach to monitor vegetation condition in the Wet Tropics, Queensland, Australia. Ecological Indicators, 18, 191-199.

Pickett, S., Cadenasso, M., & Meiners, S. (2009). Ever since Clements: From succession to vegetation dynamics and understanding to intervention. Applied Vegetation Science, 12(1), 9-21.

Pieretti, N., Farina, A., & Morri, D. (2011). A new methodology to infer the singing activity of an avian community: The Acoustic Complexity Index (ACI). Ecological Indicators, 11(3), 868-873.

Pierson, J. C., Barton, P. S., Lane, P. W., & Lindenmayer, D. B. (2015). Can habitat surrogates predict the response of target species to landscape change? Biological Conservation, 184, 1-10.

Pijanowski, B. C., Farina, A., Gage, S. H., Dumyahn, S. L., & Krause, B. L. (2011a). What is soundscape ecology? An introduction and overview of an emerging new science. Landscape Ecology, 26(9), 1213-1232.

Pijanowski, B. C., Villanueva-Rivera, L. J., Dumyahn, S. L., Farina, A., Krause, B. L., Napoletano, B. M., Gage, S. H., & Pieretti, N. (2011b). Soundscape Ecology: The Science of Sound in the Landscape. BioScience, 61(3), 203-216.

Pineda, E., Moreno, C., Escobar, F., & Halffter, G. (2005). Frog, bat, and dung beetle diversity in the cloud forest and coffee agroecosystems of Veracruz, Mexico. Conservation Biology, 19(2), 400-410.

Piquer-Rodriguez, M., Kuemmerle, T., Alcaraz-Segura, D., Zurita-Milla, R., & Cabello, J. (2012). Future land use effects on the connectivity of protected area networks in southeastern Spain. Journal for Nature Conservation, 20(6), 326-336.

211

Polyakov, M., Rowles, A. D., Radford, J. Q., Bennett, A. F., Park, G., Roberts, A., & Pannell, D. (2013). Using habitat extent and composition to predict the occurrence of woodland birds in fragmented landscapes. Landscape Ecology, 28(2), 329-341.

Pomara, L. Y., Ruokolainen, K., Tuomisto, H., & Young, K. R. (2012). Avian composition co‐varies with floristic composition and soil nutrient concentration in Amazonian upland forests. Biotropica, 44(4), 545-553.

Potter, K. M., & Woodall, C. W. (2012). Trends over time in tree and seedling phylogenetic diversity indicate regional differences in forest biodiversity change. Ecological Applications, 22(2), 517-531.

Powell, K. I., Chase, J. M., & Knight, T. M. (2011). A synthesis of plant invasion effects on biodiversity across spatial scales. American Journal of Botany, 98(3), 539-548.

Price, J. N., Berney, P. J., Ryder, D., Whalley, R. D. B., & Gross, C. L. (2011). Disturbance governs dominance of an invasive forb in a temporary wetland. Oecologia, 167(3), 759-769.

Price, S. J., Howe, R. W., Hanowski, J. M., Regal, R. R., Niemi, G. J., & Smith, C. R. (2007). Are anurans of Great Lakes coastal wetlands reliable indicators of ecological condition? Journal of Great Lakes Research, 33, 211-223.

Prober, S. M., & Wiehl, G. (2012). Relationships among soil fertility, native plant diversity and exotic plant abundance inform restoration of forb-rich eucalypt woodlands. Diversity and Distributions, 18(8), 795-807.

Proppe, D. S., Sturdy, C. B., & St Clair, C. C. (2013). Anthropogenic noise decreases urban songbird diversity and may contribute to homogenization. Global Change Biology, 19(4), 1075-1084.

Pugnaire, F. I., & Lazaro, R. (2000). Seed bank and understorey species composition in a semi-arid environment: The effect of shrub age and rainfall. Annals of Botany, 86(4), 807-813.

212

Qi, J., Gage, S., Joo, W., Napoletano, B., & Biswas, S. (2008). Soundscape characteristics of an environment: A new ecological indicator of ecosystem health. Wetland and water resource modeling and assessment. CRC Press, New York, 201-211.

Queensland Herbarium (2014). Regional ecosystem fire guidelines. Queensland Department of Science, Information Technology, Innovation and the Arts. Brisbane.

Quétier, F., Thébault, A., & Lavorel, S. (2007). Plant traits in a state and transition framework as markers of ecosystem response to land-use change. Ecological Monographs, 77(1), 33-52.

R Core Team (2013). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.R- project.org/

Ranta, E., Kaitala, V., Fowler, M. S., Laakso, J., Ruokolainen, L., & O'Hara, R. (2008). Detecting compensatory dynamics in competitive communities under environmental forcing. Oikos, 117(12), 1907-1911.

Reed, R. A., Peet, R. K., Palmer, M. W., & White, P. S. (1993). Scale dependence of vegetation‐environment correlations: A case study of a North Carolina piedmont woodland. Journal of Vegetation Science, 4(3), 329-340.

Regan, H. M., Crookston, J. B., Swab, R., Franklin, J., & Lawson, D. M. (2010). Habitat fragmentation and altered fire regime create trade-offs for an obligate seeding shrub. Ecology, 91(4), 1114-1123.

Rempel, R. S., Hobson, K. A., Holborn, G., Van Wilgenburg, S. L., & Elliott, J. (2005). Bioacoustic monitoring of forest songbirds: Interpreter variability and effects of configuration and digital processing methods in the laboratory. Journal of Field Ornithology, 76(1), 1-11.

Ricketts, T. H., Daily, G. C., & Ehrlich, P. R. (2002). Does butterfly diversity predict moth diversity? Testing a popular indicator taxon at local scales. Biological Conservation, 103(3), 361-370.

213

Roccaforte, J. P., Huffman, D. W., Fule, P. Z., Covington, W. W., Chancellor, W. W., Stoddard, M. T., & Crouse, J. E. (2015). Forest structure and fuels dynamics following ponderosa pine restoration treatments, White Mountains, Arizona, USA. Forest Ecology and Management, 337, 174-185.

Rodrigues, A. S., & Brooks, T. M. (2007). Shortcuts for biodiversity conservation planning: The effectiveness of surrogates. Annual review of ecology, evolution, and systematics, 713-737.

Rodriguez, J. P., Pearson, D. L., & Barrera, R. R. (1998). A test for the adequacy of bioindicator taxa: Are tiger beetles (Coleoptera: Cicindelidae) appropriate indicators for monitoring the degradation of tropical forests in Venezuela? Biological Conservation, 83(1), 69-76.

Rohde, K. (1992). Latitudinal gradients in species diversity: The search for the primary cause. Oikos, 514-527.

Rooney, R. C., Bayley, S. E., Creed, I. F., & Wilson, M. J. (2012). The accuracy of land cover-based wetland assessments is influenced by landscape extent. Landscape Ecology, 27(9), 1321-1335.

Rosenvald, R., Lohmus, A., Kraut, A., & Remm, L. (2011). Bird communities in hemiboreal old-growth forests: The roles of food supply, stand structure, and site type. Forest Ecology and Management, 262(8), 1541-1550.

Ross, K. A., Fox, B. J., & Fox, M. D. (2002). Changes to plant species richness in forest fragments: Fragment age, disturbance and fire history may be as important as area. Journal of Biogeography, 29(5-6), 749-765.

Ross, Y. (1998). Habitat trees and hollow dependent fauna. Brisbane: Queensland Government.

Rotenberry, J. T. (1985). The role of habitat in avian community composition: Physiognomy or floristics? Oecologia, 67(2), 213-217.

214

Rothrock, P. E., Simon, T. P., & Stewart, P. M. (2008). Development, calibration, and validation of a littoral zone plant index of biotic integrity (PIBI) for lacustrine wetlands. Ecological Indicators, 8(1), 79-88.

Ruiz-Jaen, M. C., & Aide, T. M. (2005). Vegetation structure, species diversity, and ecosystem processes as measures of restoration success. Forest Ecology and Management, 218(1), 159-173.

Russell-Smith, J., Ryan, P. G., Klessa, D., Waight, G., & Harwood, R. (1998). Fire regimes, fire-sensitive vegetation and fire management of the sandstone Arnhem Plateau, monsoonal northern Australia. Journal of Applied Ecology, 35(6), 829-846.

Saetersdal, M., Gjerde, I., Blom, H. H., Ihlen, P. G., Myrseth, E. W., Pommeresche, R., Skartveit, J., Solhøy, T., & Aas, O. (2004). Vascular plants as a surrogate species group in complementary site selection for bryophytes, macrolichens, spiders, carabids, staphylinids, snails, and wood living polypore fungi in a northern forest. Biological Conservation, 115(1), 21-31.

Sarkar, S., Justus, J., Fuller, T., Kelley, C., Garson, J., & Mayfield, M. (2005). Effectiveness of environmental surrogates for the selection of conservation area networks. Conservation Biology, 19(3), 815-825.

Sarkar, S., Pressey, R. L., Faith, D. P., Margules, C. R., Fuller, T., Stoms, D. M., Moffett, A., Wilson, K. A., Williams, K. J., & Williams, P. H. (2006). Biodiversity conservation planning tools: Present status and challenges for the future. Annu. Rev. Environ. Resour., 31, 123-159.

Sattler, P. S., & Williams, R. D. (1999). The Conservation Status of Queensland’s Bioregional Ecosystems. Brisbane: The Queensland Government.

Sauberer, N., Zulka, K. P., Abensperg-Traun, M., Berg, H. M., Bieringer, G., Milasowszky, N., Moser, D., Plutzar, C., Pollheimer, M., Storch, C., Trostl, R., Zechmeister, H., & Grabherr, G. (2004). Surrogate taxa for biodiversity in agricultural landscapes of eastern Austria. Biological Conservation, 117(2), 181-190.

215

Saunders, D. A., Hobbs, R. J., & Margules, C. R. (1991). Biological consequences of ecosystem fragmentation: A review. Conservation Biology, 5(1), 18-32.

Saunders, K. M., McMinn, A., Roberts, D., Hodgson, D. A., & Heijnis, H. (2007). Recent human-induced salinity changes in Ramsar-listed Orielton Lagoon, south-east Tasmania, Australia: A new approach for coastal lagoon conservation and management. Aquatic Conservation-Marine and Freshwater Ecosystems, 17(1), 51- 70.

Schmit, J. P., Mueller, G. M., Leacock, P. R., Mata, J. L., & Huang, Y. (2005). Assessment of tree species richness as a surrogate for macrofungal species richness. Biological Conservation, 121(1), 99-110.

Schneiders, A., Van Daele, T., Van Landuyt, W., & Van Reeth, W. (2012). Biodiversity and ecosystem services: Complementary approaches for ecosystem management? Ecological Indicators, 21, 123-133.

Schuster, R., & Arcese, P. (2013). Using bird species community occurrence to prioritize forests for old growth restoration. Ecography, 36(4), 499-507.

Seddon, J., Briggs, S., & Doyle, S. (2003). Relationships between bird species and characteristics of woodland remnants in central New South Wales. Pacific Conservation Biology, 9(2), 95.

Sewell, S. R., & Catterall, C. P. (1998). Bushland modification and styles of urban development: Their effects on birds in south-east Queensland. Wildlife Research, 25(1), 41-63.

Shanley, C. S., Pyare, S., & Smith, W. P. (2013). Response of an ecological indicator to landscape composition and structure: Implications for functional units of temperate rainforest ecosystems. Ecological Indicators, 24, 68-74.

Sharma, G. R., Raghubanshi, A. S., & Singh, J. S. (2005). Lantana invasion: An overview. Weed Biology and Management, 5(4), 157-165.

216

Shive, K. L., Sieg, C. H., & Fule, P. Z. (2013). Pre-wildfire management treatments interact with fire severity to have lasting effects on post-wildfire vegetation response. Forest Ecology and Management, 297, 75-83.

Simberloff, D. (1998). Flagships, umbrellas, and keystones: Is single-species management passé in the landscape era? Biological Conservation, 83(3), 247-257.

Singh, J., Roy, P., Murthy, M., & Jha, C. (2010). Application of landscape ecology and remote sensing for assessment, monitoring and conservation of biodiversity. Journal of the Indian Society of Remote Sensing, 38(3), 365-385.

Skowno, A. L., & Bond, W. J. (2003). Bird community composition in an actively managed savanna reserve, importance of vegetation structure and vegetation composition. Biodiversity and Conservation, 12(11), 2279-2294.

Skroblin, A., & Legge, S. (2012). Influence of fine-scale habitat requirements and riparian degradation on the distribution of the Purple-crowned Fairy-wren (Malurus coronatus coronatus) in northern Australia. Austral Ecology, 37(8), 874-884.

Skull, S., & Adams, F. (1996). Fire management and the vegetation communities of the Townsville and Thuringowa Shires: Some ecological considerations. Queensland, Australia: Australian Centre for Tropical Freshwater Research, James Cook University of North Queensland.

Smyth, A., Friedel, M., & O'Malley, C. (2009). The influence of buffel grass (Cenchrus ciliaris) on biodiversity in an arid Australian landscape. Rangeland Journal, 31(3), 307-320.

Soons, M., Messelink, J., Jongejans, E., & Heil, G. (2005). Habitat fragmentation reduces grassland connectivity for both short‐distance and long‐distance wind‐ dispersed forbs. Journal of Ecology, 93(6), 1214-1225.

Specht, A., & Specht, R. L. (1993). Species richness and canopy productivity of Australian plant communities. Biodiversity & Conservation, 2(2), 152-167.

217

Specht, R., & Specht, A. (1989a). Species richness of overstorey strata in Australian plant communities— the influence of overstorey growth rates. Australian Journal of Botany, 37(4), 321-336.

Specht, R., & Specht, A. (1989b). Species richness of sclerophyll (heathy) plant communities in Australia—the influence of overstorey cover. Australian Journal of Botany, 37(4), 337-350.

Speight, J. G. (1990). Australian soil and land survey field handbook. Melbourne: Inkata Press.

Sperber, T. D., Wraith, J. M., & Olson, B. E. (2003). Soil physical properties associated with the invasive spotted knapweed and native grasses are similar. Plant and Soil, 252(2), 241-249.

Spies, T. A., Lindenmayer, D. B., Gill, A. M., Stephens, S. L., & Agee, J. K. (2012). Challenges and a checklist for biodiversity conservation in fire-prone forests: Perspectives from the Pacific Northwest of USA and Southeastern Australia. Biological Conservation, 145(1), 5-14.

Standovar, T., Odor, P., Aszalos, R., & Galhidy, L. (2006). Sensitivity of ground layer vegetation diversity descriptors in indicating forest naturalness. Community Ecology, 7(2), 199-209.

Stephenson, N. L. (1990). Climatic control of vegetation distribution: The role of the water balance. American Naturalist, 649-670.

Stephenson, N. L. (1999). Reference conditions for giant sequoia forest restoration: structure, process, and precision. Ecological Applications, 9(4), 1253-1265.

Stringham, T. K., Krueger, W. C., & Shaver, P. L. (2003). State and transition modeling: An ecological process approach. Journal of Range Management, 106-113.

Suarez-Rubio, M., & Thomlinson, J. R. (2009). Landscape and patch-level factors influence bird communities in an urbanized tropical island. Biological Conservation, 142(7), 1311-1321.

218

Sueur, J., Pavoine, S., Hamerlynck, O., & Duvail, S. (2008). Rapid acoustic survey for biodiversity appraisal. Plos One, 3(12).

Sullivan, T. P., Sullivan, D. S., Lindgren, P. M., & Ransome, D. B. (2012). If we build habitat, will they come? Woody debris structures and conservation of forest mammals. Journal of Mammalogy, 93(6), 1456-1468.

Sun, D., Hnatiuk, R., & Neldner, V. (1997). Review of vegetation classification and mapping systems undertaken by major forested land management agencies in Australia. Australian Journal of Botany, 45(6), 929-948.

Sverdrup-Thygeson, A., & Lindenmayer, D. (2003). Ecological continuity and assumed indicator fungi in boreal forest: the importance of the landscape matrix. Forest Ecology and Management, 174(1), 353-363.

Tang, C. Q., He, L. Y., Gao, Z. R., Zhao, X. F., Sun, W. B., & Ohsawa, M. (2011). Habitat fragmentation, degradation, and population status of endangered Michelia coriacea in southeastern Yunnan, China. Mountain Research and Development, 31(4), 343- 350.

Tang, Z., Fang, J., Chi, X., Feng, J., Liu, Y., Shen, Z., Wang, X., Wang, Z., Wu, X., Zheng, C., & Gaston, K. J. (2012). Patterns of plant beta-diversity along elevational and latitudinal gradients in mountain forests of China. Ecography, 35(12), 1083-1091.

Tassicker, A. L., Kutt, A. S., Vanderduys, E., & Mangru, S. (2006). The effects of vegetation structure on the birds in a tropical savanna woodland in north-eastern Australia. Rangeland Journal, 28(2), 139-152.

Taylor, A. H. (2004). Identifying forest reference conditions on early cut-over lands, Lake Tahoe Basin, USA. Ecological Applications, 14(6), 1903-1920.

Taylor, P. D., Fahrig, L., Henein, K., & Merriam, G. (1993). Connectivity is a vital element of landscape structure. Oikos, 68(3), 571-573.

Taylor, R. S., Watson, S. J., Bennett, A. F., & Clarke, M. F. (2013). Which fire management strategies benefit biodiversity? A landscape-perspective case study

219

using birds in mallee ecosystems of south-eastern Australia. Biological Conservation, 159, 248-256.

Thackway, R., & Lesslie, R. (2008). Describing and mapping human-induced vegetation change in the Australian landscape. Environmental Management, 42(4), 572-590.

Theoharides, K. A., & Dukes, J. S. (2007). Plant invasion across space and time: factors affecting nonindigenous species success during four stages of invasion. New Phytologist, 176(2), 256-273.

Thomas, J. A., Bourn, N. A. D., Clarke, R. T., Stewart, K. E., Simcox, D. J., Pearman, G. S., Curtis, R., & Goodger, B. (2001). The quality and isolation of habitat patches both determine where butterflies persist in fragmented landscapes. Proceedings of the Royal Society B-Biological Sciences, 268(1478), 1791-1796.

Thompson, J. R., Duncan, S. L., & Johnson, K. N. (2009). Is there potential for the historical range of variability to guide conservation given the social range of variability? Ecology and Society, 14(1).

Tognelli, M. F. (2005). Assessing the utility of indicator groups for the conservation of South American terrestrial mammals. Biological Conservation, 121(3), 409-417.

Torras, O., Gil-Tena, A., & Saura, S. (2008). How does forest landscape structure explain tree species richness in a Mediterranean context? Biodiversity and Conservation, 17(5), 1227-1240.

Travers, S. K., & Eldridge, D. J. (2013). Increased rainfall frequency triggers an increase in litter fall rates of reproductive structures in an arid eucalypt woodland. Austral Ecology, 38(7), 820-830.

Tremont, R. M., & McIntyre, S. (1994). Natural grassy vegetation and native forbs in temperate Australia—structure, dynamics and life-histories. Australian Journal of Botany, 42(6), 641-658.

220

Trisurat, Y., Alkemade, R., & Verburg, P. H. (2010). Projecting land-use change and its consequences for biodiversity in northern Thailand. Environmental Management, 45(3), 626-639.

Uezu, A., & Metzger, J. P. (2011). Vanishing bird species in the Atlantic Forest: Relative importance of landscape configuration, forest structure and species characteristics. Biodiversity and Conservation, 20(14), 3627-3643.

Valentine, L. E., Roberts, B., & Schwarzkopf, L. (2007). Mechanisms driving avoidance of non-native plants by lizards. Journal of Applied Ecology, 44(1), 228-237.

Van der Maarel, E. (1990). Ecotones and ecoclines are different. Journal of Vegetation Science, 135-138.

Van der Maarel, E. (2005). Vegetation ecology. United Kingdom: Blackwell Publishing.

Van Leeuwen, C. (1966). A relation theoretical approach to pattern and process in vegetation. Wentia, 15(S1), 25-46.

Walker, B., & Salt, D. (2012). Resilience thinking: Sustaining ecosystems and people in a changing world: Island Press.

Walker, S., Wilson, J. B., Steel, J. B., Rapson, G., Smith, B., King, W. M., & Cottam, Y. H. (2003). Properties of ecotones: evidence from five ecotones objectively determined from a coastal vegetation gradient. Journal of Vegetation Science, 14(4), 579-590.

Walsh, C., & MacNally, R. (2013). hier.part: Hierarchical Partitioning (Version 1.0-4): R package.

Warman, L. D., Forsyth, D. M., Sinclair, A., Freemark, K., Moore, H. D., Barrett, T. W., Pressey, R. L., & White, D. (2004). Species distributions, surrogacy, and important conservation regions in Canada. Ecology Letters, 7(5), 374-379.

Weeds Australia (2014). Silver-leaf Desmodium (Desmodium uncinatum). Retrieved 25 September, 2014, from http://www.weeds.org.au/cgi- bin/weedident.cgi?tpl=plant.tpl&ibra=all&card=V31

221

Wehenkel, C., Corral-Rivas, J. J., Castellanos-Bocaz, H. A., & Pinedo-Alvarez, A. (2009). Is there a positive relationship between naturalness and genetic diversity in forest tree communities? Investigacion Agraria-Sistemas Y Recursos Forestales, 18(1), 20- 27.

Weinberg, A. Z., Kavanagh, R. P., Law, B. S., & Penman, T. D. (2008). Testing biodiversity toolkits: How well do they predict vertebrate species richness. Sydney: Science and Research Division, New South Wales Department of Primary Industries.

Welch, P.D. (1967). The use of fast fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms. Trans Audio Electro 15:70-73.

Welsh, H. H., & Ollivier, L. M. (1998). Stream amphibians as indicators of ecosystem stress: A case study from California's redwoods. Ecological Applications, 8(4), 1118- 1132.

Wesolowski, A., Adams, M. A., & Pfautsch, S. (2014). Insulation capacity of three bark types of temperate Eucalyptus species. Forest Ecology and Management, 313, 224- 232.

Westman, W. E. (1978). Measuring the inertia and resilience of ecosystems. BioScience, 28(11), 705-710.

Westoby, M., Walker, B., & Noy-Meir, I. (1989). Opportunistic management for rangelands not at equilibrium. Journal of Range Management, 266-274.

Whipp, R. K., Lunt, I. D., Deane, A., & Spooner, P. G. (2009). Historical forest survey data from Eucalyptus-Callitris forests: A valuable resource for long-term vegetation studies. Australian Journal of Botany, 57(7), 541-555.

Whipp, R. K., Lunt, I. D., Spooner, P. G., & Bradstock, R. A. (2012). Changes in forest structure over 60 years: Tree densities continue to increase in the Pilliga forests, New South Wales, Australia. Australian Journal of Botany, 60(1), 1-8.

222

Whittaker, R. H. (1956). Vegetation of the great smoky mountains. Ecological Monographs, 26(1), 1-80.

Wiens, J. A., Seavy, N. E., & Jongsomjit, D. (2011). Protected areas in climate space: What will the future bring? Biological Conservation, 144(8), 2119-2125.

Wildlife Acoustics (2013). Wildlife Acoustic Bioacoustic Monitoring Systems. Retrieved April 2013, from http://www.wildlifeacoustics.com/

Williams, J. E. (2000). The biodiversity crisis and adaptation to climate change: A case study from Australia's forests. Environmental Monitoring and Assessment, 61(1), 65-74.

Williams, M. C., & Wardle, G. M. (2005). The invasion of two native eucalypt forests by Pinus radiata in the Blue Mountains, New South Wales, Australia. Biological Conservation, 125(1), 55-64.

Williams, R., Duff, G., Bowman, D., & Cook, G. (1996). Variation in the composition and structure of tropical savannas as a function of rainfall and soil texture along a large‐ scale climatic gradient in the Northern Territory, Australia. Journal of Biogeography, 23(6), 747-756.

Wilson, B. A., Neldner, V. J., & Accad, A. (2002). The extent and status of remnant vegetation in Queensland and its implications for statewide vegetation management and legislation. Rangeland Journal, 24(1), 6-35.

Wimmer, J., Towsey, M., Planitz, B., Williamson, I., & Roe, P. (2013b). Analysing environmental acoustic data through collaboration and automation. Future Generation Computer Systems, 29(2), 560-568.

Wimmer, J., Towsey, M., Roe, P., & Williamson, I. (2013a). Sampling environmental acoustic recordings to determine bird species richness. Ecological Applications, 23(6), 1419-1428.

223

Winter, S., Fischer, H. S., & Fischer, A. (2010). Relative quantitative reference approach for naturalness assessments of forests. Forest Ecology and Management, 259(8), 1624-1632.

Wintle, B. A. (2008). A review of biodiversity investment prioritization tools. A report to the Biodiversity Expert Working Group toward the development of the Investment Framework for Environmental Resources.

Wintle, B. A., Elith, J., & Potts, J. M. (2005). Fauna habitat modelling and mapping: A review and case study in the Lower Hunter Central Coast region of NSW. Austral Ecology, 30(7), 719-738.

Woldendorp, G., & Keenan, R. J. (2005). Coarse woody debris in Australian forest ecosystems: A review. Austral Ecology, 30(8), 834-843.

Woldendorp, G., Keenan, R., & Ryan, M. (2002). Coarse woody debris in Australian forest ecosystems. A report for the national greenhouse strategy, Module, 6.

WorldClim v1.4. (2004a). Mean annual temperature [ESRI raster map file]. Retrieved from: http://www.worldclim.org/current

WorldClim v1.4. (2004b). Temperature variability [ESRI raster map file]. Retrieved from: http://www.worldclim.org/current

Wright, R. G., Maccracken, J. G., & Hall, J. (1994). An ecological evaluation of proposed new conservation areas in Idaho—evaluating proposed Idaho national parks. Conservation Biology, 8(1), 207-216.

Wulf, M., & Kolk, J. (2014). Plant species richness of very small forests related to patch configuration, quality, heterogeneity and history. Journal of Vegetation Science, 25(5), 1267-1277.

Yang, X. J., & Liu, Z. (2005). Quantifying landscape pattern and its change in an estuarine watershed using satellite imagery and landscape metrics. International Journal of Remote Sensing, 26(23), 5297-5323.

224

Yapp, G., Walker, J., & Thackway, R. (2010). Linking vegetation type and condition to ecosystem goods and services. Ecological Complexity, 7(3), 292-301.

Yates, C. J., Elith, J., Latimer, A. M., Le Maitre, D., Midgley, G. F., Schurr, F. M., & West, A. G. (2010). Projecting climate change impacts on species distributions in megadiverse South African Cape and Southwest Australian Floristic Regions: Opportunities and challenges. Austral Ecology, 35(4), 374-391.

Yoccoz, N. G., Nichols, J. D., & Boulinier, T. (2001). Monitoring of biological diversity in space and time. Trends in Ecology & Evolution, 16(8), 446-453.

Zhang, J., Kissling, W. D., & He, F. (2013). Local forest structure, climate and human disturbance determine regional distribution of boreal bird species richness in Alberta, Canada. Journal of Biogeography, 40(6), 1131-1142.

Zharikov, Y., Lank, D. B., & Cooke, F. (2007). Influence of landscape pattern on breeding distribution and success in a threatened Alcid, the marbled murrelet: Model transferability and management implications. Journal of Applied Ecology, 44(4), 748-759.

225

Supplementary Material

Supplementary Figure 2.1: BioCondition survey plot layout.

229

Supplementary Table 2.1: Full site details including site code, site name, regional ecosystem, subregion (RE 12.11.5e), UTM coordinates, patch area, and patch size and connectivity descriptions for RE 12.11.5e and RE 12.5.3. * indicates a reference site used to calculate benchmarks.

Site code Site name Regional Subregion UTM coordinates Patch size Patch size Patch connectivity ecosystem (ha) description description BRIS1 * D'Aguilar 1 12.11.5e Brisbane 56 490319 6962945 44110.12225 Large High BRIS2 D'Aguilar 2 12.11.5e Brisbane 56 488074 6963777 44110.12225 Large High BRIS3 D'Aguilar 3 12.11.5e Brisbane 56 481767 6975887 44110.12225 Large High BRIS4 * Moggill 1 12.11.5e Brisbane 56 486216 6956386 44110.12225 Large High BRIS5 Moggill 2 12.11.5e Brisbane 56 486234 6955963 44110.12225 Large High BRIS6 Moggill 3 12.11.5e Brisbane 56 486804 6956434 44110.12225 Large High BRIS7 Tamborine 1 12.11.5e Brisbane 56 517000 6918504 20684.52 Large High BRIS8 Tamborine 2 12.11.5e Brisbane 56 517192 6919226 20684.52 Large High BRIS9 Samford Lanita 12.11.5e Brisbane 56 490524 6969392 44110.12225 Large High BRIS10 Kimberley 1 12.11.5e Brisbane 56 517999 6941924 16248.78 Large High BRIS11 Kimberley 2 12.11.5e Brisbane 56 518032 6942335 16248.78 Large High BRIS12 Leacroft 12.11.5e Brisbane 56 517402 6949703 12988.77 Large High BRIS13 Plunkett C.P. 12.11.5e Brisbane 56 516374 6922951 14662.81 Large High BRIS14 Clear Mountain 12.11.5e Brisbane 56 490609 6979048 3158.5 Large High BRIS15 Toohey Forest Park 1 12.11.5e Brisbane 56 507055 6953603 128.18 Medium Low BRIS16 Toohey Forest Park 2 12.11.5e Brisbane 56 505424 6954073 391.88 Medium High BRIS17 Whites Hill 1 12.11.5e Brisbane 56 508537 6957257 125.09 Medium Low BRIS18 Whites Hill 2 12.11.5e Brisbane 56 507699 6956503 125.09 Medium Low BRIS19 Mount Warren Park 12.11.5e Brisbane 56 518570 6932829 3.44 Small Low BRIS20 Nursery Road 12.11.5e Brisbane 56 507838 6955073 3.94 Small Low

230

Site code Site name Regional Subregion UTM coordinates Patch size Patch size Patch connectivity ecosystem (ha) description BRIS21 Wellers Hill Reservoir 12.11.5e Brisbane 56 505045 6955343 5.18 Small Low BRIS22 Morrison Road 12.11.5e Brisbane 56 482342 6975841 1.37 Small Low BRIS23 Narracott Street 12.11.5e Brisbane 56 509716 6958581 1.73 Small Low BRIS24 Greenslopes 12.11.5e Brisbane 56 504319 6956866 7.27 Small Low GYM1 Gympie N.P. 1 12.11.5e Gympie 56 472534 7114734 17328.75 Large High GYM2 * Gympie N.P. 2 12.11.5e Gympie 56 471196 7117225 17328.75 Large High GYM3 Meadows Lane 1 12.11.5e Gympie 56 458672 7107040 932.73 Large High GYM4 Meadows Lane 2 12.11.5e Gympie 56 459426 7107170 932.73 Large High GYM5 Woondum S.F. 12.11.5e Gympie 56 471954 7096674 971.45 Large High GYM6 Lynchs Hill 12.11.5e Gympie 56 459558 7100280 4919.46 Large High GYM7 * Goomboorian 12.11.5e Gympie 56 479394 7103975 17328.75 Large High GYM8 Rocky Ridge Road 12.11.5e Gympie 56 468490 7106878 247.32 Medium Low GYM9 Curra 1 12.11.5e Gympie 56 465217 7117445 9107.94 Large High GYM10 Curra 2 12.11.5e Gympie 56 463489 7114590 9107.94 Large High GYM11 Neerdie 1 12.11.5e Gympie 56 472236 7124993 17328.75 Large High GYM12 Neerdie 2 12.11.5e Gympie 56 473487 7124582 17328.75 Large High GYM13 Mothar Mountain 1 12.11.5e Gympie 56 474969 7096790 194.28 Medium Low GYM14 Mothar Mountain 2 12.11.5e Gympie 56 474820 7096716 194.28 Medium Low GYM15 Greenmount Lane 12.11.5e Gympie 56 466471 7105809 29.3 Medium Low GYM16 David Street 12.11.5e Gympie 56 466886 7105610 5.97 Small Low GYM17 Butler Street 12.11.5e Gympie 56 467370 7105122 9.08 Small Low GYM18 Tamaree Road 12.11.5e Gympie 56 467209 7109750 13.24 Small Low 1253_1 Noosa N.P. 12.5.3 NA 56 508240 7070170 13501.07 Large High 1253_2 Weyba 12.5.3 NA 56 504805 7076189 13000.16 Large High 1253_3 * Caves Road 12.5.3 NA 56 493730 7018758 955.72 Large High 1253_4 * Murphys Road 12.5.3 NA 56 491531 7024818 1350.56 Large Low

231

Site code Site name Regional Subregion UTM coordinates Patch size Patch size Patch connectivity ecosystem (ha) description 1253_5 * SA1 (Roys Road) 12.5.3 NA 56 499199 7029070 730.10 Large High 1253_6 Tingalpa Creek 12.5.3 NA 56 499199 7029070 11663.57 Large High 1253_7 Greater Glider 12.5.3 NA 56 521096 6954031 47.98 Medium Low 1253_8 Sheepstation Creek 12.5.3 NA 56 490673 6999161 253.70 Medium High 1253_9 Stern Road 12.5.3 NA 56 488993 7004487 123.17 Medium Low 1253_10 Harper Road 12.5.3 NA 56 489335 7004860 123.17 Medium Low 1253_11 Greening Road 12.5.3 NA 56 491369 7007313 317.94 Medium Low 1253_12 Freshwater 12.5.3 NA 56 498706 6994317 358.60 Medium Low 1253_13 Saddleback 12.5.3 NA 56 492513 7012187 232.85 Medium Low 1253_14 King Road 12.5.3 NA 56 489555 7012191 124.90 Medium High 1253_15 Korawal Street 12.5.3 NA 56 521289 6952426 1.70 Small Low 1253_16 Coolnwynpin N.R. 12.5.3 NA 56 520743 6952645 3.23 Small Low 1253_17 Komraus Court 12.5.3 NA 56 492899 6999280 5.16 Small Low 1253_18 New Settlement Rd 12.5.3 NA 56 497638 6993974 2.67 Small Low 1253_19 Steve Irwin Way 12.5.3 NA 56 496177 7025440 5.33 Small Low 1253_20 Newells Road 12.5.3 NA 56 495605 7030517 13.48 Small Low

232

Supplementary Table 2.2: BioCondition survey sheet used for recording ecological condition data.

GENERAL INFORMATION Date: Collector: Bioregion: Site ID: Time: Property: RE: Photos: Photo Number Direction N S E W 0m 100m Photo Number Description Datum: Zone: 0m AMGE: AMGN: Derivation: Accuracy: 50m AMGE: AMGN: Transect Bearing: Description: Specht Code:

LANDFORM Situation: Element: Erosion Pattern: Pattern:

SLOPE Type: Degrees: Aspect: Altitude:

100x50m AREA VEGETATION STRUCTURE AND COMPOSITION Stratum Median Range Cover (intercept) Dominance Species Individual Cover E T1

T1 T2

T2 T3

S1 S2

S2 S3

S3 G

233

100x50m AREA: NUMBER OF LARGE TREES AND TREE SPECIES RICHNESS Species (E or N) Tally of DBH size classes (cm) <20 20-25 25-30 30-35 35-40

Species (E or N) 40-45 45-50 50-55 55-60 >60

Eucalypts: Average DBH= Non-eucalypts: Average DBH= Total large trees: Tree canopy height Sub-canopy and/or emergent height S: E: Proportion of canopy recruitment Total tree species richness

50x20m AREA: COARSE WOODY DEBRIS CWD Length CWD Length CWD Length CWD Length

50x10m AREA: SHRUB AND GROUND SPECIES RICHNESS Shrub species richness: Grass species richness: Forbs and other species richness: Non-native cover: Hollows <10cm Hollows >10cm

234

1x1m QUADRATS: GROUND COVER Ground Cover 1 2 3 4 5 Mean Native grass cover Native forbs Native shrubs (<1m) Non-native grass Non-native forbs and shrubs Litter Rock Bare ground Cryptogams Total 100 100 100 100 100 100

100m TRANSECT: TREE AND SHRUB CANOPY COVERS Species Native/Exotic Height T1/T2/S1/S2 Intercept Total Range

235

Supplementary Table 2.3: Vegetation composition and abundance survey sheet.

GENERAL INFORMATION Date: Collector: Bioregion: Site ID: Time: Property: RE: Datum: Zone: AMGE: AMGN: Derivation: Accuracy: Transect Bearing: Aspect: Description:

236

COVER (%) STEM COUNT BASAL AREA - FACTOR 1 (NUMBER) (NUMBER) Species Name E T1 T2 T3 S1 S2 G E T1 T2 T3 S1 S2 E T1 T2 T3 S1 Exterior

237

Supplementary Figure 3.1: CCA of all species presence/absence and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons.

238

Supplementary Figure 3.2: CCA of all large tree species abundance and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons.

239

Supplementary Figure 3.3: CCA of all tree species abundance and environmental and disturbance variables across RE 12.11.5e, with Brisbane and Gympie sites highlighted by black polygons.

240

Supplementary Figure 3.4: CCA of all shrub species abundance and environmental and disturbance variables across RE 12.11.5e, with broad associations highlighted by black polygons.

241

Supplementary Figure 3.5: CCA of all ground species cover abundance and environmental and disturbance variables across RE 12.11.5e, with broad associations highlighted by black polygons.

242

Supplementary Figure 3.6: CCA of all species presence/absence and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons.

243

Supplementary Figure 3.7: CCA of all large tree species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons.

244

Supplementary Figure 3.8: CCA of all tree species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons.

245

Supplementary Figure 3.9: CCA of all shrub species abundance and environmental and disturbance variables across RE 12.5.3, with broad associations highlighted by black polygons.

246

Supplementary Table 3.1: Pearson two-tailed correlations testing for collinearity of disturbance variables for RE 12.11.5e (N=42). Highlighted variables indicate those selected for analyses.

Patch Road Road Age Fire Fire Age Fire Log. (L) Cons. (C) For. (F) Res. (R) Rural Grazing Local Size (PS) Prop. (RA) Prop. (FA) Height Res. (RR) (GNE) Road (RP) (FP) (FH) (LR) Patch Size (PS) r 1 -.174 -.294 -.221 -.153 -.202 -.510 .609 -.125 -.384 -.042 -.228 -.502 p value .272 .059 .160 .334 .199 .001 .000 .430 .012 .790 .147 .001 Road Prop. (RP) r -.174 1 .902 -.117 -.203 -.078 .056 -.260 .000 .409 -.076 -.078 .498 p value .272 .000 .461 .198 .625 .724 .096 .999 .007 .634 .626 .001 Road Age (RA) r -.294 .902 1 -.123 -.161 -.042 .080 -.205 .000 .386 -.114 -.081 .473 p value .059 .000 .438 .309 .793 .617 .192 .999 .012 .473 .608 .002 Fire Prop. (FP) r -.221 -.117 -.123 1 .363 .368 .216 -.175 .193 .030 -.093 .043 -.069 p value .160 .461 .438 .018 .016 .169 .268 .221 .853 .559 .788 .662 Fire Age (FA) r -.153 -.203 -.161 .363 1 .303 .306 -.193 .045 -.058 .014 .334 -.072 p value .334 .198 .309 .018 .051 .049 .221 .777 .715 .931 .031 .649 Fire Height (FH) r -.202 -.078 -.042 .368 .303 1 .180 -.152 -.068 .174 -.075 .200 .169 p value .199 .625 .793 .016 .051 .253 .335 .671 .270 .637 .203 .284 Log. (L) r -.510 .056 .080 .216 .306 .180 1 -.584 .332 .045 .107 .280 .170 p value .001 .724 .617 .169 .049 .253 .000 .032 .776 .498 .072 .283 Cons. (C) r .609 -.260 -.205 -.175 -.193 -.152 -.584 1 -.496 -.248 -.224 -.273 -.331 p value .000 .096 .192 .268 .221 .335 .000 .001 .114 .153 .080 .033 For. (F) r -.125 .000 .000 .193 .045 -.068 .332 -.496 1 -.297 -.264 -.157 -.297 p value .430 .999 .999 .221 .777 .671 .032 .001 .056 .091 .320 .056 Res. (R) r -.384 .409 .386 .030 -.058 .174 .045 -.248 -.297 1 -.167 -.228 .925 p value .012 .007 .012 .853 .715 .270 .776 .114 .056 .290 .146 .000

247

Patch Road Road Age Fire Fire Age Fire Logging Cons. (C) Forestry Res. (R) Rural Grazing Local Size (PS) Prop. (RA) Prop. (FA) Height (L) (F) Res. (RR) (GNE) Road (RP) (FP) (FH) (LR) Rural Res. (RR) r -.042 -.076 -.114 -.093 .014 -.075 .107 -.224 -.264 -.167 1 .023 -.154 p value .790 .634 .473 .559 .931 .637 .498 .153 .091 .290 .886 .330 Graz. (GNE) r -.228 -.078 -.081 .043 .334 .200 .280 -.273 -.157 -.228 .023 1 -.043 p value .147 .626 .608 .788 .031 .203 .072 .080 .320 .146 .886 .785 Local Road (LR) r -.502 .498 .473 -.069 -.072 .169 .170 -.331 -.297 .925 -.154 -.043 1 p value .001 .001 .002 .662 .649 .284 .283 .033 .056 .000 .330 .785

248

Supplementary Table 3.2: Pearson two-tailed correlations testing for collinearity of disturbance variables for RE 12.5.3 (N=20). Highlighted variables indicate those selected for analyses.

Patch Road Road Fire Fire Fire Log. Cons. For. Res. Rural Graz. Hort. Main Local Track Size Prop. Age Prop. Age Height (L) (C) (F) (R) Res. (GNE) (H) Road Road (T) (PS) (RP) (RA) (FP) (FA) (FH) (RR) (MR) (LR) Patch Size r 1 -.331 -.317 .185 -.071 .203 -.290 .489 -.171 -.256 -.104 .070 -.282 .021 -.071 .001 (PS) p value .154 .174 .435 .766 .390 .216 .028 .471 .275 .662 .769 .229 .929 .768 .995 Road Prop. r -.331 1 .855 -.498 -.574 -.133 -.223 -.415 -.324 .095 .213 .069 .458 .441 .471 .333 (RP) p value .154 .000 .026 .008 .578 .345 .069 .164 .690 .368 .773 .042 .052 .036 .151 Road Age r -.317 .855 1 -.529 -.376 .000 -.166 -.530 -.306 .361 .153 .028 .323 .404 .700 .051 (RA) p value .174 .000 .016 .103 1.000 .484 .016 .190 .117 .518 .907 .165 .078 .001 .831 Fire Prop. r .185 -.498 -.529 1 .437 .437 .252 .391 .178 -.455 -.080 .166 .047 .068 -.113 .087 (FP) p value .435 .026 .016 .054 .054 .285 .088 .453 .044 .738 .483 .843 .776 .637 .715 Fire Age (FA) r -.071 -.574 -.376 .437 1 .310 .614 .268 .406 -.081 -.361 .022 .005 -.226 -.266 -.083 p value .766 .008 .103 .054 .183 .004 .253 .076 .735 .118 .928 .982 .338 .257 .729 Fire Height r .203 -.133 .000 .437 .310 1 .064 .162 .268 -.145 -.324 .209 .289 .080 -.011 .012 (FH) p value .390 .578 1.000 .054 .183 .788 .494 .253 .542 .164 .376 .217 .738 .965 .959 Log. (L) r -.290 -.223 -.166 .252 .614 .064 1 -.222 .305 .045 -.111 .195 .234 -.009 -.102 .017 p value .216 .345 .484 .285 .004 .788 .347 .191 .852 .641 .410 .321 .970 .669 .942 Cons. (C) r .489 -.415 -.530 .391 .268 .162 -.222 1 .043 -.492 -.592 .163 -.075 -.217 -.539 .271 p value .028 .069 .016 .088 .253 .494 .347 .858 .028 .006 .492 .754 .358 .014 .247 For. (F) r -.171 -.324 -.306 .178 .406 .268 .305 .043 1 -.230 -.233 -.061 -.042 -.236 -.350 .033 p value .471 .164 .190 .453 .076 .253 .191 .858 .330 .324 .798 .860 .316 .130 .890 Res. (R) r -.256 .095 .361 -.455 -.081 -.145 .045 -.492 -.230 1 -.075 -.379 -.339 -.091 .232 -.552 p value .275 .690 .117 .044 .735 .542 .852 .028 .330 .753 .100 .144 .701 .325 .012 Rural Res. r -.104 .213 .153 -.080 -.361 -.324 -.111 -.592 -.233 -.075 1 -.330 -.198 .226 .467 -.151 (RR) p value .662 .368 .518 .738 .118 .164 .641 .006 .324 .753 .156 .402 .338 .038 .526

249

Patch Road Road Fire Fire Fire Log. Cons. For. Res. Rural Graz. Hort. Main Local Track Size Prop. Age Prop. Age Height (L) (C) (F) (R) Res. (GNE) (H) Road Road (T) (PS) (RP) (RA) (FP) (FA) (FH) (RR) (MR) (LR) Graz. (GNE) r .070 .069 .028 .166 .022 .209 .195 .163 -.061 -.379 -.330 1 .499 .146 .018 .360 p value .769 .773 .907 .483 .928 .376 .410 .492 .798 .100 .156 .025 .538 .941 .119 Hort. (H) r -.282 .458 .323 .047 .005 .289 .234 -.075 -.042 -.339 -.198 .499 1 .181 .047 .560 p value .229 .042 .165 .843 .982 .217 .321 .754 .860 .144 .402 .025 .445 .845 .010 Main Road r .021 .441 .404 .068 -.226 .080 -.009 -.217 -.236 -.091 .226 .146 .181 1 .365 .110 (MR) p value .929 .052 .078 .776 .338 .738 .970 .358 .316 .701 .338 .538 .445 .113 .645 Local Road r -.071 .471 .700 -.113 -.266 -.011 -.102 -.539 -.350 .232 .467 .018 .047 .365 1 -.349 (LR) p value .768 .036 .001 .637 .257 .965 .669 .014 .130 .325 .038 .941 .845 .113 .132 Track (T) r .001 .333 .051 .087 -.083 .012 .017 .271 .033 -.552 -.151 .360 .560 .110 -.349 1 p value .995 .151 .831 .715 .729 .959 .942 .247 .890 .012 .526 .119 .010 .645 .132

250

Supplementary Table 3.3: Pearson two-tailed correlations testing for collinearity of environmental variables for RE 12.11.5e (N=42). Highlighted variables indicate those selected for analyses.

North/ Slope Aspect Alt. Dist. Av. Rain. Pot. Av. Temp. Bulk Plant Lat. (N) (S) (ASP) (ALT) from Annual Variab. Evapo- Annual Variab. Dens. Avail. Coast Rain. (RCV) trans. Temp. (TCV) (BD) Water (DC) (AAR) (PET) (AAT) (PAW) North/Lat. (N) r 1 -.032 -.313 .333 .797 .220 -.739 .779 .675 .043 .122 .452 p value .843 .043 .031 .000 .162 .000 .000 .000 .784 .443 .003 Slope (S) r -.032 1 .008 .482 .098 -.118 .038 -.033 -.268 .113 .213 -.182 p value .843 .958 .001 .536 .456 .810 .834 .086 .477 .175 .248 Aspect (ASP) r -.313 .008 1 .008 -.315 .139 .260 -.222 -.321 .144 -.152 .016 p value .043 .958 .959 .042 .380 .096 .157 .038 .362 .337 .918 Alt. (ALT) r .333 .482 .008 1 .365 .247 -.052 .246 -.309 .258 .218 -.049 p value .031 .001 .959 .018 .115 .743 .116 .047 .099 .165 .756 Dist. from Coast (DC) r .797 .098 -.315 .365 1 -.185 -.560 .935 .433 .480 .239 .420 p value .000 .536 .042 .018 .240 .000 .000 .004 .001 .128 .006 Av. Annual Rain. (AAR) r .220 -.118 .139 .247 -.185 1 .096 -.121 -.023 -.271 -.176 -.046 p value .162 .456 .380 .115 .240 .547 .445 .884 .083 .265 .771 Rain. Variab. (RCV) r -.739 .038 .260 -.052 -.560 .096 1 -.565 -.600 .057 -.054 -.548 p value .000 .810 .096 .743 .000 .547 .000 .000 .718 .735 .000 Pot. Evapotrans. (PET) r .779 -.033 -.222 .246 .935 -.121 -.565 1 .452 .568 .058 .486 p value .000 .834 .157 .116 .000 .445 .000 .003 .000 .713 .001 Av. Annual Temp. (AAT) r .675 -.268 -.321 -.309 .433 -.023 -.600 .452 1 -.247 .053 .427 p value .000 .086 .038 .047 .004 .884 .000 .003 .114 .740 .005 Temp. Variab. (TCV) r .043 .113 .144 .258 .480 -.271 .057 .568 -.247 1 .025 .123 p value .784 .477 .362 .099 .001 .083 .718 .000 .114 .873 .437

251

252

Supplementary Table 3.4: Pearson two-tailed correlations testing for collinearity of environmental variables for RE 12.5.3 (N=20). Highlighted variables indicate those selected for analyses.

North/ Slope Aspect Alt. Dist. Av. Rain. Pot. Av. Temp. Bulk Plant Lat. (N) (S) (ASP) (ALT) from Annual Variab. Evapo- Annual Variab. Dens. Avail. Coast Rain. (RCV) trans. Temp. (TCV) (BD) Water (DC) (AAR) (PET) (AAT) (PAW) North/Lat. (N) r 1 .023 -.099 -.089 .007 .759 -.825 -.079 .603 -.469 .023 -.504 p value .924 .677 .708 .977 .000 .000 .742 .005 .037 .924 .023 Slope (S) r .023 1 -.393 .813 .330 .158 .191 .211 -.491 .291 -.007 -.217 p value .924 .086 .000 .156 .505 .419 .372 .028 .213 .975 .357 Aspect (ASP) r -.099 -.393 1 -.207 -.100 -.174 .017 -.052 .038 .007 -.263 -.052 p value .677 .086 .381 .673 .463 .944 .829 .872 .978 .263 .827 Alt. (ALT) r -.089 .813 -.207 1 .422 .209 .302 .265 -.606 .388 -.288 -.100 p value .708 .000 .381 .064 .377 .196 .258 .005 .091 .219 .676 Dist. from Coast (DC) r .007 .330 -.100 .422 1 -.008 .089 .820 -.540 .713 -.099 -.323 p value .977 .156 .673 .064 .972 .708 .000 .014 .000 .677 .165 Av. Annual Rain. (AAR) r .759 .158 -.174 .209 -.008 1 -.777 -.281 .421 -.503 -.093 -.148 p value .000 .505 .463 .377 .972 .000 .230 .064 .024 .697 .534 Rain. Variab. (RCV) r -.825 .191 .017 .302 .089 -.777 1 .301 -.712 .634 -.154 .171 p value .000 .419 .944 .196 .708 .000 .198 .000 .003 .518 .471 Pot. Evapotrans. (PET) r -.079 .211 -.052 .265 .820 -.281 .301 1 -.525 .883 -.116 -.522 p value .742 .372 .829 .258 .000 .230 .198 .017 .000 .627 .018 Av. Annual Temp. (AAT) r .603 -.491 .038 -.606 -.540 .421 -.712 -.525 1 -.789 .277 .014 p value .005 .028 .872 .005 .014 .064 .000 .017 .000 .237 .952 Temp. Variab. (TCV) r -.469 .291 .007 .388 .713 -.503 .634 .883 -.789 1 -.114 -.341 p value .037 .213 .978 .091 .000 .024 .003 .000 .000 .633 .142 Bulk Dens.(BD) r .023 -.007 -.263 -.288 -.099 -.093 -.154 -.116 .277 -.114 1 -.017 p value .924 .975 .263 .219 .677 .697 .518 .627 .237 .633 .943 Plant Avail. Water (PAW) r -.504 -.217 -.052 -.100 -.323 -.148 .171 -.522 .014 -.341 -.017 1 p value .023 .357 .827 .676 .165 .534 .471 .018 .952 .142 .943

253

Supplementary Table 3.5: Botanical species list for RE 12.11.5e. * indicates exotic species.

Scientific name Family Abildgaardia ovata Cyperaceae Acacia amblygona Fabaceae Acacia concurrens Fabaceae Acacia deanei Fabaceae Acacia disparrima Fabaceae Acacia falcata Fabaceae Acacia fimbriata Fabaceae Acacia implexa Fabaceae Acacia leiocalyx Fabaceae Acacia maidenii Fabaceae Acacia oshanesii Fabaceae Acacia penninervis Fabaceae Acacia podalyriifolia Fabaceae Acrotriche aggregata Ericaceae Adiantum hispidulum Pteridaceae Adiantum silvaticum Pteridaceae Ageratum conyzoides * Asteraceae Alchornea ilicifolia Euphorbiaceae Alectryon subdentatus Sapindaceae Alectryon tomentosus Sapindaceae Allocasuarina littoralis Casuarinaceae Allocasuarina torulosa Casuarinaceae Alloteropsis semialata Poaceae Alphitonia excelsa Rhamnaceae Alyxia ruscifolia Apocynaceae Angophora leiocarpa Myrtaceae Angophora woodsiana Myrtaceae Aphananthe philippinensis Ulmaceae Aristida benthamii Poaceae Aristida gracilipes Poaceae Aristida queenslandica Poaceae Aristolochia meridionalis Aristolochiaceae Arundinella nepalensis Poaceae Asclepias curassavica * Apocynaceae Asparagus aethiopicus * Asparagaceae Asparagus africanus * Asparagaceae Asparagus sp. * Asparagaceae Astrotricha latifolia Araliaceae Atractocarpus chartaceus Rubiaceae Babingtonia bidwillii Myrtaceae Bidens pilosa * Asteraceae Boronia polygalifolia Rutaceae Boronia rosmarinifolia Rutaceae Brachychiton populneus Malvaceae Breynia oblongifolia Euphorbiaceae

254

Scientific name Family Brunoniella australis Acanthaceae Bryophyllum delagoense * Crassulaceae Callerya megasperma Fabaceae Calotis dentex Asteraceae Ipomoea indica * Convolvulaceae Capillipedium spicigerum Poaceae Capparis arborea Capparaceae Capparis sarmentosa Capparaceae Capparis velutina Capparaceae Cardiospermum grandiflorum * Sapindaceae Carissa ovata Apocynaceae Cassytha glabella Lauraceae Cassytha pubescens Lauraceae Cayratia clematidea Vitaceae Celtis sinensis * Ulmaceae Cenchrus robustus Poaceae Centella asiatica Apiaceae Cheilanthes distans Pteridaceae Chrysocephalum apiculatum Asteraceae Chrysopogon fallax Poaceae Chrysopogon silvaticus * Poaceae Cinnamomum camphora * Lauraceae Cirsium vulgare * Asteraceae Claoxylon australe Euphorbiaceae Clematicissus opaca Vitaceae Commelina cyanea Commelinaceae Commersonia bartramia Malvaceae Conyza sumatrensis * Asteraceae Correa reflexa Rutaceae Corymbia citriodora Myrtaceae Corymbia intermedia Myrtaceae Corymbia trachyphloia Myrtaceae Crotalaria montana Fabaceae Cupaniopsis parvifolia Sapindaceae Cyanthillium cinereum Asteraceae Cyclophyllum coprosmoides Rubiaceae Cyclophyllum longipetalum Rubiaceae Cymbopogon refractus Poaceae Cyperus haspan Cyperaceae Cyperus polystachyos Cyperaceae Cyperus tetraphyllus Cyperaceae Daviesia ulicifolia Fabaceae Daviesia villifera Fabaceae Deeringia amaranthoides Amaranthaceae Desmodium brachypodum Fabaceae Desmodium gunnii Fabaceae Desmodium heterocarpon Fabaceae Desmodium intortum * Fabaceae Desmodium rhytidophyllum Fabaceae Desmodium uncinatum * Fabaceae

255

Scientific name Family Desmodium varians Fabaceae Dianella caerulea Phormiaceae Dichelachne micrantha Poaceae Dichelachne montana Poaceae Digitaria parviflora Poaceae Dillwynia retorta var. retorta Fabaceae Diospyros fasciculosa Ebenaceae Diospyros geminata Ebenaceae Dipodium variegatum Orchidaceae Dodonaea viscosa Sapindaceae Drynaria rigidula Polypodiaceae Einadia nutans Chenopodiaceae Elymus scaber Poaceae Emilia sonchifolia Asteraceae Entolasia marginata Poaceae Entolasia stricta Poaceae Eragrostis brownii Poaceae Eragrostis mexicana * Poaceae Eragrostis sororia Poaceae Eremochloa bimaculata Poaceae Eremophila debilis Scrophulariaceae Eucalyptus acmenoides Myrtaceae Eucalyptus carnea Myrtaceae Eucalyptus cloeziana Myrtaceae Eucalyptus crebra Myrtaceae Eucalyptus exserta Myrtaceae Eucalyptus fibrosa Myrtaceae Eucalyptus longirostrata Myrtaceae Eucalyptus major Myrtaceae Eucalyptus microcorys Myrtaceae Eucalyptus moluccana Myrtaceae Eucalyptus propinqua Myrtaceae Eucalyptus siderophloia Myrtaceae Eucalyptus tereticornis Myrtaceae Eucalyptus acmenoides Myrtaceae Euroschinus falcatus Anacardiaceae Eustrephus latifolius Luzuriagaceae Ficus virens Moraceae Fimbristylis dichotoma Cyperaceae Fimbristylis brownii Cyperaceae Flemingia parviflora Fabaceae Flindersia australis Rutaceae Gahnia aspera Cyperaceae Geitonoplesium cymosum Luzuriagaceae Glochidion ferdinandi Euphorbiaceae Glossocardia bidens Asteraceae Glycine clandestina Fabaceae Glycine tabacina Fabaceae Gomphocarpus fruticosus * Apocynaceae Gompholobium latifolium Fabaceae

256

Scientific name Family Gompholobium pinnatum Fabaceae Goodenia rotundifolia Goodeniaceae Guioa semiglauca Sapindaceae Hardenbergia violacea Fabaceae Helichrysum collinum Asteraceae Hibbertia aspera Dilleniaceae Hibbertia linearis Dilleniaceae Hibbertia obtusifolia ssp. linearis Dilleniaceae Hibbertia vestita Dilleniaceae Hibiscus heterophyllus Malvaceae Hippocratea barbata Celastraceae Hovea acutifolia Fabaceae Hovea heterophylla Fabaceae Hybanthus stellarioides Violaceae Hypoestes floribunda Acanthaceae Hypoxis pratensis var. pratensis Hypoxidaceae Impatiens walleriana * Balsaminaceae Imperata cylindrica Poaceae Indigofera australis Fabaceae Jacaranda mimosifolia * Bignoniaceae Jacksonia scoparia Fabaceae Jagera pseudorhus Sapindaceae Lagenophora gracilis Asteraceae Lantana camara * Verbenaceae Lantana montevidensis * Verbenaceae Laxmannia gracilis Anthericaceae Lepidosperma laterale Cyperaceae Leptospermum speciosum Myrtaceae Leucopogon juniperinus Dilleniaceae Lobelia purpurascens Lobeliaceae Lomandra confertifolia Lomandraceae Lomandra filiformis Lomandraceae Lomandra longifolia Lomandraceae Lomandra multiflora Lomandraceae Lomandra spicata Lomandraceae Lophostemon confertus Myrtaceae Lophostemon suaveolens Myrtaceae Maclura cochinchinensis Moraceae Macrozamia lucida Zamiaceae Macrozamia pauli-guilielmi Zamiaceae Mallotus discolor Euphorbiaceae Mallotus philippensis Euphorbiaceae Marsdenia fraseri Apocynaceae Marsdenia sp. Apocynaceae Maytenus silvestris Celastraceae Megathyrsus maximus * Poaceae Melaleuca saligna Myrtaceae Melichrus adpressus Ericaceae Melinis repens * Poaceae Mentha diemenica Lamiaceae

257

Scientific name Family Mentha grandiflora Lamiaceae Microlaena stipoides Poaceae Microseris lanceolata Asteraceae Murdannia graminea Commelinaceae Myoporum montanum Scrophulariaceae Myrsine variabilis Myrsinaceae Notelaea longifolia Oleaceae Nyssanthes diffusa Amaranthaceae Ochna serrulata * Ochnaceae Opercularia diphylla Rubiaceae Opercularia hispida Rubiaceae Oplismenus aemulus Poaceae Oplismenus hirtellus var. imbecillis Poaceae Opuntia stricta * Cactaceae Ottochloa gracillima Poaceae Oxalis chnoodes Oxalidaceae Ozothamnus bidwillii Asteraceae Pandorea pandorana Bignoniaceae Panicum effusum Poaceae Panicum simile Poaceae Parsonsia leichhardtii Apocynaceae Parsonsia straminea Apocynaceae Paspalidium distans Poaceae Passiflora suberosa * Passifloraceae Persoonia sericea Proteaceae Petalostigma pubescens Picrodendraceae Petalostigma triloculare Picrodendraceae Phyllanthus fuernrohrii Phyllanthaceae Phyllanthus gunnii Phyllanthaceae Phyllanthus virgatus Phyllanthaceae Pimelea latifolia Thymelaeaceae Pipturus argenteus Urticaceae Pittosporum revolutum Pittosporaceae Pittosporum spinescens Pittosporaceae Pittosporum undulatum Pittosporaceae Pittosporum viscidum Pittosporaceae Plantago lanceolata * Plantaginaceae Plectranthus parviflorus Lamiaceae Podolobium aciculiferum Fabaceae Podolobium ilicifolium Fabaceae Podolobium scandens Fabaceae Polygala japonica Polygalaceae Polymeria calycina Convolvulaceae Polyscias elegans Araliaceae Pomaderris ferruginea Rhamnaceae Pomax umbellata Rubiaceae Pratia concolor Lobeliaceae Praxelis clematidea * Asteraceae Pseuderanthemum variabile Acanthaceae Psychotria daphnoides Rubiaceae

258

Scientific name Family Psychotria loniceroides Rubiaceae Pteridium esculentum Dennstaedtiaceae Pterocaulon redolens Asteraceae Pultenaea spinosa Fabaceae Pultenaea villosa Fabaceae Rapanea variabilis Myrsinaceae Raphanus raphanistrum * Brassicaceae Rhodamnia rubescens Myrtaceae Rhodosphaera rhodanthema Anacardiaceae Rostellularia adscendens Acanthaceae Rostellularia obtusa Acanthaceae Salvia plebeia Lamiaceae Schefflera actinophylla * Araliaceae Schenkia australis Gentianaceae Schoenus sp. Cyperaceae Scleria mackaviensis Cyperaceae Scleria sp. Cyperaceae Senecio tenuiflorus Asteraceae Senecio hispidulus Asteraceae Senecio quadridentatus Asteraceae Senecio spanomerus Asteraceae Senna pendula var. glabrata * Fabaceae Sida filiformis Malvaceae Siegesbeckia orientalis * Asteraceae Smilax australis Smilacaceae Solanum densevestitum Solanaceae Solanum ellipticum Solanaceae Solanum mauritianum * Solanaceae Solanum nemophilum Solanaceae Solanum nigrum * Solanaceae Sonchus oleraceus * Asteraceae Sporobolus sp. Poaceae Stackhousia monogyna Stackhousiaceae Stephania japonica Menispermaceae Streblus brunonianus Moraceae Swainsona galegifolia Fabaceae Symplocos harroldii Symplocaceae Taraxacum officinale * Asteraceae Tecoma stans * Bignoniaceae Tephrosia sp. Fabaceae Themeda triandra Poaceae Trema aspera Ulmaceae Tricoryne elatior Anthericaceae Urochloa decumbens * Poaceae Velleia paradoxa Goodeniaceae Velleia spathulata Goodeniaceae Wedelia spilanthoides Asteraceae Wikstroemia indica Thymelaeaceae Xanthorrhoea johnsonii Xanthorrhoeaceae Xanthorrhoea latifolia Xanthorrhoeaceae

259

Scientific name Family Zieria minutiflora ssp. minutiflora Rutaceae Zieria smithii Rutaceae Zornia muriculata Fabaceae

260

Supplementary Table 3.6: Botanical species list for RE 12.5.3. * indicates exotic species.

Species name Family Abutilon grandifolium * Malvaceae Acacia complanata Fabaceae Acacia concurrens Fabaceae Acacia disparrima Fabaceae Acacia fimbriata Fabaceae Acacia hubbardiana Fabaceae Acacia leiocalyx Fabaceae Acacia maidenii Fabaceae Acrotriche aggregata Ericaceae Ageratum houstonianum * Asteraceae Agiortia pedicellata Ericaceae Alectryon connatus Sapindaceae Allocasuarina littoralis Casuarinaceae Alloteropsis semialata Poaceae Alphitonia excelsa Rhamnaceae Angophora leiocarpa Myrtaceae Angophora woodsiana Myrtaceae Aristida benthamii Poaceae Aristida gracilipes Poaceae Aristida queenslandica Poaceae Arthrochilus irritabilis Orchidaceae Asparagus aethiopicus * Asparagaceae Asparagus plumosus * Asparagaceae Austromyrtus dulcis Myrtaceae Baccharis halimifolia * Asteraceae Baeckea frutescens Myrtaceae Baloskion tetraphyllum Restionaceae Banksia integrifolia Proteaceae Banksia oblongifolia Proteaceae Banksia spinulosa Proteaceae Bidens pilosa * Asteraceae Billardiera scandens Pittosporaceae Boronia polygalifolia Rutaceae Boronia rosmarinifolia Rutaceae Brachyloma daphnoides Ericaceae Breynia oblongifolia Euphorbiaceae Callitris columellaris Cupressaceae Cassytha glabella Lauraceae Cassytha pubescens Lauraceae Centella asiatica Apiaceae Centratherum punctatum ssp. Asteraceae australianum Cheilanthes distans Pteridaceae Chrysanthemoides monilifera* Asteraceae Cinnamomum camphora * Lauraceae Convolvulus erubescens Convolvulaceae

261

Scientific name Family Corymbia gummifera Myrtaceae Corymbia intermedia Myrtaceae Corymbia trachyphloia Myrtaceae Cyanthillium cinereum Asteraceae Cymbopogon refractus Poaceae Dampiera stricta Goodeniaceae Dampiera sylvestris Goodeniaceae Daviesia ulicifolia Fabaceae Daviesia umbellulata Fabaceae Desmodium gunnii Fabaceae Desmodium rhytidophyllum Fabaceae Desmodium uncinatum * Fabaceae Dianella caerulea Phormiaceae Dianella rara Phormiaceae Digitaria parviflora Poaceae Dipodium variegatum Orchidaceae Dodonaea triquetra Sapindaceae Drosera spatulata Droseraceae Emilia sonchifolia Asteraceae Entolasia stricta Poaceae Epacris obtusifolia Ericaceae Epacris pulchella Ericaceae Eragrostis brownii Poaceae Eragrostis interrupta Poaceae Eragrostis spartinoides Poaceae Eremochloa bimaculata Poaceae Eucalyptus crebra Myrtaceae Eucalyptus microcorys Myrtaceae Eucalyptus pilularis Myrtaceae Eucalyptus propinqua Myrtaceae Eucalyptus racemosa Myrtaceae Eucalyptus resinifera Myrtaceae Eucalyptus seeana Myrtaceae Eucalyptus tindaliae Myrtaceae Eustrephus latifolius Luzuriagaceae Fimbristylis cinnamomatorum Cyperaceae Fimbristylis velata Cyperaceae Flemingia parviflora Fabaceae Gahnia aspera Cyperaceae Geitonoplesium cymosum Luzuriagaceae Geodorum densiflorum Orchidaceae Glochidion ferdinandi Euphorbiaceae Glochidion sumatranum Euphorbiaceae Glycine clandestina Fabaceae Glycine tabacina Fabaceae Gomphocarpus fruticosus * Apocynaceae Gompholobium pinnatum Fabaceae Gompholobium virgatum var. virgatum Fabaceae Gonocarpus tetragynus Haloragaceae Goodenia rotundifolia Goodeniaceae

262

Scientific name Family Grevillea leiophylla Proteaceae Guioa semiglauca Sapindaceae Haemodorum austroqueenslandicum Haemodoraceae Hakea actites Proteaceae Hakea florulenta Proteaceae Hardenbergia violacea Fabaceae Hibbertia aspera Dilleniaceae Hibbertia scandens Dilleniaceae Hibbertia vestita Dilleniaceae Hovea acutifolia Fabaceae Imperata cylindrica Poaceae Jacaranda mimosifolia * Bignoniaceae Jacksonia scoparia Fabaceae Jagera pseudorhus Sapindaceae Juncus planifolius Juncaceae Juncus usitatus Juncaceae Lantana camara * Verbenaceae Laxmannia gracilis Anthericaceae Lepidosperma laterale Cyperaceae Leptocarpus tenax Restionaceae Leptospermum juniperinum Myrtaceae Leptospermum liversidgei Myrtaceae Leptospermum polygalifolium Myrtaceae Leptospermum trinervium Myrtaceae Leucopogon juniperinus Dilleniaceae Leucopogon pimeleoides Dilleniaceae Lindsaea incisa Lindsaeaceae Lindsaea linearis Lindsaeaceae Lobelia purpurascens Lobeliaceae Lomandra confertifolia Lomandraceae Lomandra longifolia Lomandraceae Lomandra multiflora Lomandraceae Lomandra spicata Lomandraceae Lomatia silaifolia Proteaceae Lophostemon confertus Myrtaceae Lophostemon suaveolens Myrtaceae Macaranga tanarius Euphorbiaceae Marsdenia fraseri Apocynaceae Megathyrsus maximus * Poaceae Melaleuca decora Myrtaceae Melaleuca linariifolia Myrtaceae Melaleuca quinquenervia Myrtaceae Melaleuca sieberi Myrtaceae Melastoma affine Melastomataceae Mitrasacme polymorpha Loganiaceae Notelaea ovata Oleaceae Ochna serrulata * Ochnaceae Oplismenus aemulus Poaceae Oplismenus hirtellus var. imbecillis Poaceae Ottochloa gracillima Poaceae

263

Scientific name Family Oxalis chnoodes Oxalidaceae Palm sp. * Arecaceae Panicum effusum Poaceae Parsonsia leichhardtii Apocynaceae Parsonsia straminea Apocynaceae Paspalidium distans Poaceae Passiflora suberosa * Passifloraceae Patersonia sericea Iridaceae Persoonia cornifolia Proteaceae Persoonia sericea Proteaceae Persoonia tenuifolia Proteaceae Persoonia virgata Proteaceae Petalostigma pubescens Picrodendraceae Petrophile shirleyae Proteaceae Philotheca queenslandica Rutaceae Phyllanthus virgatus Phyllanthaceae Pimelea linifolia Thymelaeaceae Pinus sp. (elliotii) * Pinaceae Pipturus argenteus Urticaceae Platysace linearifolia Apiaceae Polymeria calycina Convolvulaceae Pseuderanthemum variabile Acanthaceae Pteridium esculentum Dennstaedtiaceae Ptilothrix deusta Cyperaceae Pultenaea myrtoides Fabaceae Pultenaea spinosa Fabaceae Pultenaea villosa Fabaceae Schefflera actinophylla * Araliaceae Schizaea bifida Schizaeaceae Schoenus apogon Cyperaceae Schoenus calostachyus Cyperaceae Senna pendula var. glabrata * Fabaceae Solanum mauritianum* Solanaceae Solanum torvum * Solanaceae Sphagnum cuspidatum Sphagnaceae Stephania japonica Menispermaceae Strangea linearis Proteaceae Swainsona brachycarpa Fabaceae Taraxacum officinale * Asteraceae Themeda triandra Poaceae Thysanotus tuberosus Anthericaceae Trachymene incisa Apiaceae Trema tomentosa Ulmaceae Tricoryne elatior Anthericaceae Urochloa decumbens * Poaceae Velleia spathulata Goodeniaceae Wikstroemia indica Thymelaeaceae Woollsia pungens Ericaceae Xanthorrhoea johnsonii Xanthorrhoeaceae Xanthorrhoea latifolia Xanthorrhoeaceae

264

Scientific name Family Zieria smithii Rutaceae Zornia dyctiocarpa Fabaceae

265

Supplementary Table 3.7: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native species presence/absence for RE 12.11.5e.

Cluster Description Sites Important associations 1 Mixed BRIS1, GYM6, GYM13,  Presence of : Brisbane and GYM14, BRIS4 and Acacia concurrens Gympie sites BRIS5 Hardenbergia violacea Allocasuarina torulosa Hybanthus stellarioides  Absence of: Acacia leiocalyx

2 Large remnant GYM1, GYM2, GYM5  Presence of : patches east and GYM7 Acacia maidenii and south of Carissa ovata Gympie Eucalyptus acmenoides Guioa semiglauca Hybanthus stellarioides Ottochloa gracillima Oxalis chnoodes  Similar to cluster 3 but with an additional suite of species associated with wetter conditions

3 Majority of GYM3, GYM9, GYM10,  Presence of : Gympie sites GYM11, GYM15, GYM4, Eucalyptus acmenoides GYM8, GYM16, GYM18, Hybanthus stellarioides GYM17 and GYM12 Oxalis chnoodes Cheilanthes distans Paspalidium distans Phyllanthus virgatus 4 Northern BRIS2, BRIS9, BRIS14,  Presence of : Brisbane and BRIS3, BRIS7, BRIS8, Arundinella nepalensis Gold Coast BRIS13 and BRIS19 Cassytha glabella sites Cheilanthes distans Eucalyptus carnea Persoonia sericea  Complex of species more often found in woodland with an open and sometimes heathy understorey 5 Southern BRIS10, BRIS11, BRIS12,  Presence of : suburban BRIS17, BRIS15, BRIS16, Eragrostis brownii Brisbane sites BRIS18, BRIS23, BRIS20, Corymbia trachyphloia BRIS24 and BRIS21 Eucalyptus carnea Pultenaea spinosa

266

Cluster Description Sites Important associations 6 Dry rainforest BRIS6  Presence of : site1 Nyssanthes diffusa Aristida gracilipes Capparis sarmentosa Claoxylon australe Cyclophyllum longipetalum Hibiscus heterophyllus Marsdenia fraseri Pipturus argenteus Psychotria loniceroides  Absence of: Aristida queenslandica Entolasia stricta Eremochloa bimaculata Acacia disparrima Digitaria parviflora

7 Dry rainforest BRIS22  Presence of : site 2 Alchornea ilicifolia Alectryon tomentosus Capparis velutina Hypoestes floribunda Macrozamia lucida Myrsine variabilis Nyssanthes diffusa Ozothamnus bidwillii Pittosporum viscidum  Absence of: Aristida queenslandica Entolasia stricta Eremochloa bimaculata Corymbia citriodora Acacia disparrima Digitaria parviflora

267

Supplementary Table 3.8: CCA eigenvalues and scores for significant axes for RE 12.11.5e species presence/absence data and environmental and disturbance variables. * indicates exotic species.

Axis 1 2 3 4 5

Eigen. 0.24 0.19 0.15 0.13 0.10 % var. 16.04 12.88 10.24 8.73 7.12 Sig. 0.001 0.001 0.001 0.001 0.010 Scores Ageratum 3.09 Alyxia ruscifolia -3.90 Aristolochia 3.15 Chrysocephalum -5.03 Arundinella 4.08 conyzoides * meridionalis apiculatum nepalensis Astrotricha -3.64 Astrotricha -4.48 Astrotricha -5.04 Crotalaria -4.20 Chrysocephalum -3.80 latifolia latifolia latifolia montana apiculatum Cardiospermum 3.33 Cayratia -3.02 Cyperus -5.04 Gompholobium -6.62 Crotalaria -3.19 grandiflorum * clematidea tetraphyllus pinnatum montana Cupaniopsis 3.85 Cyperus -4.48 Eucalyptus -5.04 Pittosporum 3.16 Gompholobium -4.86 parvifolia tetraphyllus cloeziana revolutum pinnatum Cyperus -3.64 Desmodium -4.38 Phyllanthus -5.04 Senecio -5.03 Parsonsia -3.10 tetraphyllus uncinatum * gunnii hispidulus leichhardtii Desmodium 4.38 Eucalyptus -4.48 Pittosporum -3.19 Xanthorrhoea -3.21 Parsonsia 3.30 uncinatum * cloeziana revolutum latifolia straminea Eucalyptus -3.64 Guioa -4.03 Plectranthus -5.04 BRIS7 -0.94 Psychotria -3.08 cloeziana semiglauca parviflorus daphnoides Hippocratea 4.38 Hippocratea -4.38 Urochloa -3.17 BRIS8 -0.93 BRIS1 0.79 barbata barbata decumbens * Marsdenia sp. -3.14 Nyssanthes -3.87 Velleia -3.03 PS -0.65 BRIS2 0.90 diffusa paradoxa Nyssanthes 4.43 Opercularia 4.04 BRIS20 -0.74 LR 0.65 GYM7 0.72 diffusa diphylla Phyllanthus -3.64 Phyllanthus -4.48 GYM1 -1.20 AAR 0.62 gunnii gunnii Plectranthus -3.64 Plectranthus -4.48 GYM2 -0.74 parviflorus parviflorus Pterocaulon -3.41 Zieria smithii -3.21 TCV 0.60 redolens

268

Axis 1 2 3 4 5 Eigen. 0.24 0.19 0.15 0.13 0.10 % var. 16.04 12.88 10.24 8.73 7.12 Sig. 0.001 0.001 0.001 0.001 0.010 Scores BRIS3 0.94 BRIS6 -0.78

BRIS6 0.87 BRIS7 0.70

BRIS22 1.24 BRIS12 0.79

GYM2 -1.06 BRIS22 -1.01

N -0.75 GYM1 -1.45

RCV 0.76 GYM2 -0.82

AAT -0.59 RR -0.45

ALT -0.72

269

Axis 6 7 8 9 10 Eigen. 0.10 0.10 0.08 0.07 0.07 % var. 7.01 6.48 5.65 5.07 4.49 Sig. 0.001 0.001 0.001 0.002 0.010 Scores Acacia 3.61 Ageratum 3.01 Acacia 5.03 Ageratum -4.47 Celtis sinensis * 3.46 amblygona conyzoides * penninervis conyzoides * Acacia 3.53 Brachychiton -3.61 Brachychiton 5.91 Allocasuarina -3.02 Cirsium vulgare * -4.87 penninervis populneus populneus littoralis Aristolochia -3.59 Cassytha -3.13 Desmodium 5.20 Cayratia -3.77 Dichelachne 3.43 meridionalis pubescens uncinatum * clematidea micrantha Brunoniella -3.15 Daviesia ulicifolia 3.04 Euroschinus 4.11 Chrysocephalum 3.02 Mentha -4.82 australis falcatus apiculatum diemenica Centella asiatica -3.71 Dichelachne -3.28 Hibbertia linearis -3.83 Desmodium -3.02 Opuntia stricta * -3.40 micrantha varians Daviesia ulicifolia 4.66 Eragrostis 3.11 Hippocratea 5.20 Eucalyptus -3.10 Parsonsia 3.67 sororia barbata tereticornis leichhardtii Desmodium 3.19 Euroschinus -3.01 Hovea acutifolia -3.18 Nyssanthes 4.11 Petalostigma -3.21 varians falcatus diffusa pubescens Eragrostis 4.00 Flindersia 3.40 Marsdenia sp. 3.31 Xanthorrhoea 6.39 Polymeria 4.94 sororia australis latifolia calycina Hovea acutifolia 3.25 Hovea acutifolia -5.24 Opercularia -4.14 BRIS14 0.90 Psychotria 3.63 diphylla daphnoides Polyscias elegans 3.06 Indigofera -5.80 Polymeria -3.62 LR -0.31 GYM17 0.66 australis calycina Pratia concolor 5.45 Marsdenia sp. -5.00 Polyscias elegans -3.01 BD -0.31 ASP 0.42 GYM10 0.72 BRIS5 -1.04 BRIS3 0.80 PAW 0.54 GYM12 0.71 GYM5 -0.78 PAW -0.32 GYM13 -0.72 GYM12 0.80 GYM14 -0.86 AAR 0.46 GNE -0.58

270

Supplementary Table 3.9: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native large tree species abundance for RE 12.11.5e.

Cluster Description Sites Important associations 1 Majority of BRIS1, BRIS2, BRIS9, Many common species Brisbane sites BRIS15, BRIS6, Eucalyptus carnea restricted to with many BRIS17, BRIS11, Brisbane sites common canopy BRIS4, BRIS8, BRIS13, Eucalyptus microcorys absent or species BRIS19, BRIS10, in low numbers BRIS12, BRIS18, BRIS14, BRIS24, BRIS3 and BRIS21

2 Brisbane sites BRIS5 and BRIS22 Moderate numbers of with few Eucalyptus microcorys Corymbia Low numbers of Corymbia citriodora citriodora Corymbia trachyphloia absent

3 Brisbane site BRIS16 Canopy dominated by with Corymbia Eucalyptus microcorys citriodora Corymbia citriodora absent absent

4 Small southern BRIS20 and BRIS23 Eucalyptus microcorys present in suburban low numbers Brisbane sites High abundance of Corymbia trachyphloia Eucalyptus crebra/siderophloia absent

5 Southern BRIS7 Absence of many common suburban canopy species Brisbane sites High numbers of Eucalyptus crebra/siderophloia

271

Cluster Description Sites Important associations 6 Majority of GYM1, GYM2, Many common species Gympie sites GYM17, GYM3 , Eucalyptus acmenoides with many GYM7, GYM6, restricted to Gympie sites common canopy GYM8, GYM11, Only cluster where Eucalyptus species GYM18, GYM4, cloeziana is present Absence of Lophostemon GYM9, GYM15, confertus GYM16, GYM10, GYM5 and GYM12

7 Mount Mothar GYM13 and GYM14 Absence of Corymbia citriodora sites with High numbers of Corymbia Corymbia intermedia, Eucalyptus citriodora propinqua and Eucalyptus absent from acmenoides canopy

272

Supplementary Table 3.10: CCA eigenvalues and scores for significant axes for RE 12.11.5e large tree species abundance data and environmental and disturbance variables.

Axis 1 2 3 4 Eigen. 0.36 0.19 0.16 0.11 % var. 32.03 17.05 14.36 9.97 Sig. 0.001 0.005 0.003 0.047 Scores Eucalyptus 2.00 Allocasuarina -2.49 Cupaniopsis 4.87 Cupaniopsis 4.13 acmenoides torulosa parvifolia parvifolia

Eucalyptus 2.59 Corymbia -2.33 Eucalyptus 5.22 Eucalyptus -3.08 cloeziana intermedia cloeziana cloeziana

Eucalyptus 2.21 Cupaniopsis -2.95 GYM2 1.37 Eucalyptus -4.26 exserta parvifolia tereticornis

Eucalyptus 2.28 Eucalyptus 4.08 LR -0.33 BRIS3 -1.05 longirostrata cloeziana

GYM2 1.00 Eucalyptus -4.06 ALT 0.38 GYM1 -0.67 tereticornis GYM4 1.03 BRIS3 -1.11 GYM2 -0.80

GYM9 1.04 GYM1 1.04 GYM4 1.39

GYM10 1.27 GYM2 1.34 PS -0.26

GYM14 0.93 GYM14 -1.17 L 0.24

N 0.92 GNE -0.54 RCV -0.28

L 0.37 LR 0.03 BD -0.32

AAT 0.58

PAW 0.49

273

Supplementary Table 3.11: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native tree species abundance for RE 12.11.5e.

Cluster Description Sites Important associations 1 Majority of BRIS1, BRIS4, BRIS13, Many common Eucalyptus and Brisbane sites BRIS15, BRIS17, Corymbia species BRIS3, BRIS2, BRIS9, Eucalyptus carnea restricted to BRIS14, BRIS10, Brisbane sites BRIS11, BRIS12, BRIS6, BRIS20 and BRIS16

2 Brisbane site BRIS5 High numbers of Corymbia with steep scree intermedia slope in Moggill High numbers of Eucalyptus Conservation carnea Park Low numbers of Corymbia citriodora

Corymbia trachyphloia, Eucalyptus crebra/siderophloia and Eucalyptus propinqua absent

3 Southern BRIS7, BRIS19, BRIS8 Low numbers of Corymbia Brisbane and and BRIS24 citriodora Gold Coast sites High abundance of Eucalyptus crebra/siderophloia Low to moderate abundance of Lophostemon confertus Tamborine National park sites (BRIS7 and BRIS8) have high numbers of Acacia disparrima and Allocasuarina spp.

4 Southern BRIS18 and BRIS23 Absence of Corymbia citriodora suburban Brisbane sites

274

Cluster Description Sites Important associations 5 Mixed Brisbane BRIS21, GYM6, High numbers of Eucalyptus and Gympie BRIS22, GYM13 and propinqua, Corymbia sites GYM14 intermedia and Lophostemon confertus Low numbers of Corymbia citriodora, Eucalyptus carnea (Brisbane) and Eucalyptus crebra/siderophloia Absence of Corymbia trachyphloia 6 Majority of GYM1, GYM2, GYM4, Presence of Eucalyptus Gympie sites GYM7, GYM10, acmenoides and Eucalyptus GYM12, GYM3, cloeziana GYM9, GYM18, GYM5, GYM17, GYM11 and GYM15

7 Gympie sites GYM8 and GYM16 Absence of Eucalyptus located in small propinqua, Eucalyptus suburban and crebra/siderophloia, peri-urban Corymbia intermedia and patches Lophostemon confertus

275

Supplementary Table 3.12: CCA eigenvalues and scores for significant axes for RE 12.11.5e all tree species abundance data and environmental and disturbance variables.

Axis 1 2 3 Eigen. 0.35 0.18 0.13 % var. 35.05 18.26 12.88 Sig. 0.001 0.009 0.017 Scores Angophora leiocarpa 2.17 Allocasuarina torulosa -2.86 Acacia disparrima 3.84

Angophora -1.90 Eucalyptus moluccana 2.13 Allocasuarina littoralis 5.40 woodsiana

Corymbia -1.58 BRIS17 0.63 Angophora -3.60 trachyphloia woodsiana

Eucalyptus cloeziana 2.48 BRIS19 -0.83 Eucalyptus cloeziana 3.97

Eucalyptus 2.47 BRIS21 -0.68 Euroschinus falcata -3.13 longirostrata BRIS6 -0.66 GYM1 0.78 BRIS7 1.46

BRIS8 -0.66 GYM2 1.04 BRIS16 -0.66

BRIS10 -0.66 GYM13 -1.36 BRIS18 -0.70

BRIS16 -0.82 GYM16 0.72 BRIS19 1.29

BRIS18 -0.89 PS 0.31 BRIS20 -0.64

BRIS23 -0.86 GNE -0.55 BRIS23 -0.77

GYM1 0.95 ALT 0.37 GYM1 0.71

GYM2 0.83 GYM2 0.76

GYM3 0.70 LR -0.33

GYM4 0.71 AAT -0.31

GYM7 0.72

GYM8 0.78

GYM11 0.84

GYM18 0.86

N 0.93

RCV -0.66

AAT 0.61

276

Supplementary Table 3.13: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native shrub species abundance for RE 12.11.5e

Cluster Description Sites Important associations 1 Brisbane sites BRIS1 and BRIS2 Low to high numbers of Acacia located in concurrens, Acacia falcata, D’Aguilar Acacia fimbriata and National Park Eucalyptus crebra

2 Exclusive GYM1, GYM5, Low to high numbers of Gympie cluster GYM15, GYM2, Eucalyptus acmenoides, GYM16, GYM17, Acacia concurrens, Acacia GYM6, GYM18 and fimbriata and Leucopogon GYM7 juniperinus Absence of Acacia leiocalyx

3 Brisbane sites BRIS 4 and BRIS5 Low to high numbers of Acacia located in fimbriata and Indigofera Moggill australis Conservation Absence of Corymbia citriodora Park

4 Mixed Brisbane BRIS19, GYM13 and Dominance of shrub layer by and Gympie GYM14 Allocasuarina torulosa and group 1 Lophostemon confertus Absence of all Eucalyptus and Corymbia species except Eucalyptus acmenoides

5 Dry rainforest BRIS6 and BRIS22 Dominance of shrub layer by cluster Alyxia ruscifolia, Cupaniopsis parvifolia, Guioa semiglauca and Jagera pseudorhus Absence of eucalypt recruitment

277

Cluster Description Sites Important associations 6 Exclusive BRIS3, BRIS11, Dominance of shrub layer by Brisbane cluster BRIS12, BRIS17, Eucalyptus BRIS9 and BRIS10 crebra/siderophloia, Acacia disparrima and high numbers of Corymbia citriodora when present Most sites have Eucalyptus carnea present Absence of Alphitonia excelsa and many Acacia species

7 Mixed Brisbane BRIS7 , BRIS8, High numbers of Eucalyptus and Gympie BRIS21, BRIS24, propinqua, Acacia group 2 BRIS13, GYM4, disparrima and Jacksonia BRIS16, BRIS18, scoparia across Brisbane BRIS23, BRIS20, and Gympie sites Eucalyptus carnea, Corymbia BRIS14, BRIS15, trachyphloia and Eucalyptus GYM3, GYM9, GYM8, microcorys present in GYM10, GYM11 and Brisbane sites GYM12

278

Supplementary Table 3.14: CCA eigenvalues and scores for significant axes for RE 12.11.5e shrub species abundance data and environmental and disturbance variables. * indicates exotic species.

Axis 1 2 3 4 5 6 8 Eigen. 0.32 0.32 0.28 0.24 0.19 0.15 0.11 % var. 15.89 15.52 13.94 11.92 9.17 7.32 5.04 Sig. 0.144 0.005 0.001 0.001 0.003 0.011 0.016 Scores Alyxia ruscifolia 3.02 Indigofera 5.00 Acacia 2.26 Acacia -2.33 Acacia -2.45 Acacia 4.51 Eucalyptus 3.97 australis penninervis penninervis penninervis penninervis moluccana Cupaniopsis 4.23 Persoonia 2.18 Acrotriche 2.78 Acrotriche 2.85 Acrotriche -2.34 Allocasuarina -3.68 Flindersia 2.02 anacardioides sericea aggregata aggregata aggregata torulosa australis Eucalyptus -1.21 Pittosporum -2.02 Corymbia -2.12 Pittosporum 3.67 Allocasuarina 2.92 Celtis sinensis * 2.70 Petalostigma -2.54 exserta revolutum trachyphloia revolutum torulosa pubescens Eucalyptus -1.30 Xanthorrhoea 2.96 Zieria smithii 2.58 Pultanaea 3.70 Pultanaea 2.72 Eucalyptus 3.78 Polyscias -2.37 moluccana johnsonii spinosa spinosa exserta elegans Guioa 4.53 BRIS4 2.19 BRIS1 0.64 Zieria smithii 4.31 Xanthorrhoea -3.28 Euroschinus 4.67 Senna pendula 2.65 semiglauca latifolia falcata var. glabrata * Jagera 2.34 BRIS5 1.95 BRIS5 1.11 BRIS1 0.89 BRIS14 -0.75 Myoporum 4.26 BRIS2 0.65 pseudorhus montanum Ochna serrulata 2.25 BRIS19 -0.79 BRIS10 -0.96 BRIS4 0.86 BRIS19 0.98 BRIS2 -0.64 BRIS6 -0.70 * Polyscias 1.04 GYM7 -1.75 BRIS16 -0.73 BRIS20 1.70 BRIS20 0.89 BRIS20 0.66 BRIS14 0.73 elegans Senna pendula 4.89 GYM11 -0.65 GYM2 0.70 GYM7 1.54 GYM6 0.73 GYM1 0.93 GYM2 -0.62 var. glabrata * Solanum 2.67 GYM14 -0.72 GYM7 1.01 GYM10 -0.64 GYM7 -0.66 GYM5 0.97 GYM3 0.65 mauritanum * BRIS6 1.62 N -0.31 GYM13 0.73 AAT -0.41 GYM13 1.72 GYM13 -0.96 GYM8 0.64

BRIS22 2.69 PS 0.45 N 0.51 BD 0.41 GYM18 0.81 GNE -0.52 GYM9 0.67

GYM10 -0.68 L -0.33 LR -0.62 L 0.46 GYM12 0.65

GYM17 0.73 AAR -0.59 GNE 0.74 FH 0.34

RR 0.75 TCV 0.52

279

Axis 1 2 3 4 5 6 8 Eigen. 0.32 0.32 0.28 0.24 0.19 0.15 0.11 % var. 15.89 15.52 13.94 11.92 9.17 7.32 5.04 Sig. 0.144 0.005 0.001 0.001 0.003 0.011 0.016 Scores FH -0.29

RCV 0.29

TCV 0.45

280

Supplementary Table 3.15: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native ground species cover abundance for RE 12.11.5e.

Cluster Description Sites Important associations 1 Brisbane sites BRIS1, BRIS2 and High cover of Arundinella located in BRIS3 nepalensis and Themeda D’Aguilar triandra National Park Absence of Eremochloa bimaculata, Dianella caerulea, Entolasia stricta and Lobelia purpurascens Arundinella nepalensis almost exclusively dominant among these sites

2 Mixed BRIS5, BRIS18, GYM5, High cover of Cymbopogon Brisbane and BRIS13, GYM15, refractus, Imperata Gympie sites GYM8, GYM12, cylindrica and Lobelia GYM16, GYM17, purpurascens GYM18, BRIS17, GYM14, GYM6, GYM13, BRIS15, BRIS7, BRIS8, BRIS19, GYM2 and GYM11

3 Gympie sites GYM3, GYM9, GYM10 High cover of Alloteropsis associated and GYM4 semialata, Goodenia with large rotundifolia, Dianella state forests caerulea and Paspalidium distans Absence of Imperata cylindrica, Lomandra confertifolia and Lobelia purpurascens

4 Predominantly BRIS4, BRIS14, High cover of Themeda triandra, Brisbane sites BRIS16, BRIS24, Entolasia stricta, Lomandra BRIS20, BRIS23, multiflora and Lepidosperma GYM7, BRIS10, laterale BRIS12, BRIS11 and Absence of Lobelia purpurascens

BRIS21

281

Cluster Description Sites Important associations 5 Brisbane site BRIS9 High cover of Arundinella with nepalensis, Lomandra depauperate multiflora, Aristida ground layer queenslandica, Cymbopogon refractus and Lepidosperma laterale Absence of most common ground species

6 Brisbane sites BRIS6 and BRIS22 High cover of Claoxylon australe, with shaded Ottochloa gracillima, Smilax understorey australis and Lomandra confertifolia Absence of most common ground species

7 Gympie site GYM1 High cover of Smilax australis, with shaded Carissa ovata and understorey Oplismenus hirtellus Many other common species absent, except for Dianella caerulea and Entolasia stricta

282

Supplementary Table 3.16: CCA eigenvalues and scores for significant axes for RE 12.11.5e ground species cover abundance data and environmental and disturbance variables. * indicates exotic species.

Axis 2 3 4 5 6 7 Eigen. 0.27 0.22 0.18 0.17 0.15 0.12 % var. 14.81 12.01 10.20 9.44 8.33 6.77 Sig. 0.001 0.005 0.003 0.001 0.001 0.001 Scores Cheilanthes distans -1.30 Cheilanthes -2.61 Arundinella 2.36 Arundinella 3.69 Arundinella -2.07 Drynaria rigidula 2.32 distans nepalensis nepalensis nepalensis Digitaria parviflora -1.30 Drynaria rigidula -3.73 Carissa ovata 6.19 Carissa ovata -2.23 Carissa ovata -2.51 Hardenbergia -3.63 violacea Drynaria rigidula 2.00 Hardenbergia -3.37 Lantana -2.14 Lepidosperma 2.34 Cheilanthes 2.12 Lantana 2.72 violacea montevidensis * laterale distans montevidensis * Eragrostis brownii -3.07 Oplismenus. 2.09 Mentha diemenica -2.97 Smilax australis -2.75 Lepidosperma 2.02 Mentha diemenica -3.15 imbecillis laterale Lantana 1.52 Pseudoranthenum 2.60 Smilax australis 2.09 Urochloa -3.60 Mentha diemenica -2.86 Paspalidium -2.02 montevidensis * variable decumbens * distans Lepidosperma -2.61 Smilax australis -2.12 Urochloa -2.28 BRIS1 0.85 Ottochloa 2.81 Urochloa -3.05 laterale decumbens * gracillima decumbens * Lobelia 1.31 Xanthorrhoea -3.35 BRIS9 1.05 BRIS2 1.32 Pseudoranthenum -3.36 Xanthorrhoea 4.93 purpurascens johnsonii variable johnsonii Mentha diemenica 2.16 BRIS5 -1.22 BRIS21 -1.24 BRIS3 1.25 Urochloa -2.90 BRIS3 -0.77 decumbens * Oplismenus hirtellus 2.28 BRIS6 -2.20 BRIS22 -0.83 BRIS6 -1.16 BRIS2 -0.98 BRIS5 1.02 ssp. imbecillis Ottochloa gracillima 1.26 BRIS21 0.93 GYM1 1.98 BRIS9 0.97 BRIS6 1.43 BRIS15 0.93

Panicum effusum -2.40 GYM3 -0.97 GYM2 0.95 BRIS19 -0.77 BRIS11 1.15 BRIS21 -1.20

Passiflora suberosa 2.85 GYM5 -0.86 GYM14 -0.87 BRIS21 -1.07 BRIS21 -0.92 BRIS22 -0.92 * Pseudoranthenum 1.30 GYM16 0.75 ALT 0.51 BRIS22 -0.78 BRIS22 1.45 GYM5 0.80 variabile Smilax australis 3.06 PS -0.41 GYM1 -0.74 GYM14 -0.71 FH -0.21

283

Axis 2 3 4 5 6 7 Eigen. 0.27 0.22 0.18 0.17 0.15 0.12 % var. 14.81 12.01 10.20 9.44 8.33 6.77 Sig. 0.001 0.005 0.003 0.001 0.001 0.001 Scores Urochloa -2.39 FH 0.46 LR -0.40 FH -0.41 RR -0.25 decumbens * Xanthorrhoea 1.05 TCV -0.47 AAR 0.42 RR 0.49 AAR -0.21 johnsonii BRIS3 0.88 RCV -0.26

BRIS6 2.05 PAW 0.21

BRIS9 -0.83

BRIS10 -1.04

BRIS11 -0.98

BRIS12 -0.98

BRIS22 0.86

GYM1 1.35

GYM6 0.71

GYM17 0.76

GYM18 0.80

FH -0.45

RR 0.50

284

Supplementary Table 3.17: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native species presence/absence for RE 12.5.3.

Cluster Description Sites Important associations 1 Large northern 1253_1 and 5  Presence of: heathland sites Daviesia umbellulata Ptilothrix deusta Petrophile shirlyae Strangea linearis  Absence of: Acacia disparrima Acacia leiocalyx

2 Large and small 1253_2, 3 and 20  Presence of: heathland sites Billardia scandens Dampiera sylvestris Hakea actites Pultenaea villosa  Absence of: Acacia disparrima

3 Predominantly 1253_4, 13, 12, 14, 7  Presence of: central and 11 Glycine clandestina woodland/open Cheilanthes distans forest sites Acacia disparrima (1253_7 located Acacia leiocalyx south) Banksia oblongifolia

4 Predominantly 1253_6, 15, 17, 16  Presence of: small, disturbed and 18 Digitaria parviflora woodland/open Lophostemon suaveolens forest patches Eremochloa bimaculata located Goodenia rotundifolia centrally and to Leptocarpus tenax Acacia disparrima and Acacia the south with leiocalyx scattered across a moist soil few sites conditions  Absence of: Lophostemon confertus

285

Cluster Description Sites Important associations

5 Centrally 1253_8, 10, 19 and 9  Presence of: located Glochidion sumatranum woodland/open Acacia disparrima forest sites Acacia leiocalyx Lophostemon confertus Stephania japonica Jagera pseudorhus  Absence of: Many heath species

286

Supplementary Table 3.18: CCA eigenvalues and scores for significant axes for RE 12.5.3 species presence/absence data and environmental and disturbance variables. * indicates exotic species.

Axis 1 2 Eigen. 0.42 0.33 % var. 15.74 12.41 Sig. 0.042 0.008 Scores Abutilon grandifolium * -3.47 Abutilon grandifolium * 4.76

Acacia complanata 2.06 Acacia complanata 2.34

Ageratum houstonianum * -2.28 Acacia fimbriata -2.42

Agiortia pedicellata 2.06 Ageratum houstonianum * 2.87

Angophora leiocarpa -3.47 Agiortia pedicellata 2.34

Asparagus aethiopicus * -2.12 Angophora leiocarpa 4.76

Austromyrtus dulcis 2.94 Asparagus aethiopicus * 2.53

Baloskion tetraphyllum -3.47 Baloskion tetraphyllum 4.76

Brachyloma daphnoides 2.94 Callitris collumellaris 2.34

Callitris collumellaris 2.06 Centratherum punctatum ssp. 2.34 australianum Centratherum punctatum ssp. 2.06 Cinnamomum camphora * -2.14 australianum Convolvulus erubescens -3.47 Convolvulus erubescens 4.76

Dampiera stricta 3.35 Desmodium gunnii -2.14

Desmodium uncinatum * -3.47 Desmodium uncinatum * 4.76

Epacris obtusifolia 3.35 Eucalyptus propinqua 4.76

Eragrostis interrupta 2.94 Jacksonia scoparia -2.42

Eucalyptus propinqua -3.47 Leptospermum juniperinum 4.76

Hakea actites 2.19 Lindsaea incisa 4.76

Leptospermum juniperinum -3.47 Melaleuca decora 4.76

Leptospermum liversidegei 2.94 Schefflera actinophylla * 2.87

Lindsaea incisa -3.47 Senna pendula var. glabrata * 2.53

Lomandra spicata 3.35 Solanum torvum * 4.76

Marsdenia fraseri 2.94 Sphagnum cuspidatum 4.76

Melaleuca decora -3.47 Wikstroemia indica -2.14

Palm sp. * -2.28 Xanthorrhoea johnsonii 3.15

Persoonia virgata 2.53 Zieria smithii 2.34

Petrophile shirlyae 3.35 1253_1 0.74

287

Axis 1 2 Eigen. 0.40 % var. 15.61 Sig. 0.020 Scores Philotheca queenslandica 3.35 1253_3 0.48

Platysace linearifolia 3.35 1253_7 -0.74

Petrophile shirleyae 2.94 1253_9 -0.60

Schefflera actinophylla * -2.28 1253_11 -0.41

Senna pendula var. glabrata * -2.12 1253_13 -0.43

Solanum torvum * -3.47 1253_17 -0.53

Sphagnum cuspidatum -3.47 1253_18 1.62

Strangea linearis 2.78 1253_20 0.77

Trema tomentosa 2.06 N 0.39

Woollsia pungens 2.94 FP -0.57

Zieria smithii 2.06 TCV -0.35

1253_1 1.27 BD -0.30

1253_2 0.69

1253_3 0.72

1253_5 1.40

1253_18 -1.44

1253_19 -0.61

1253_20 1.00

N 0.59

PS 0.38

FP 0.49

RCV -0.67

TCV -0.50

288

Supplementary Table 3.19: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native large tree species for RE 12.5.3.

Cluster Description Sites Important associations 1 Sites with high 1253_1, 10 and 2 High numbers of Eucalyptus numbers of racemosa, Melaleuca Melaleuca quinquenervia and quinquenervia 1 Lophostemon suaveolens (associated with Absence of Allocasuarina littoralis lower, eroded slopes)

2 Sites with high 1253_6 and 18 High numbers of Melaleuca numbers of quinquenervia Melaleuca Low numbers of Eucalyptus quinquenervia 2 racemosa (associated with Absence of Eucalyptus tindaliae or Corymbia gummifera lower, eroded slopes)

3 Southern sites 1253_7, 16 and 15 High numbers of Acacia disparrima, Banksia integrifolia, Corymbia trachyphloia, Allocasuarina littoralis and Eucalyptus microcorys Absence of Eucalyptus tindaliae and Corymbia gummifera

4 Central sites 1253_3, 13, 14 High numbers of Eucalyptus and 17 tindaliae Absence of Melaleuca quinquenervia and Acacia disparrima

5 Central and 1253_4, 8, 11, 5, High abundance of Eucalyptus northern sites 9, 12, 20 and 19 racemosa and Corymbia intermedia

289

Supplementary Table 3.20: CCA eigenvalues and scores for significant axes for RE 12.5.3 large tree species abundance data and environmental and disturbance variables. * indicates exotic species.

Axis 1 2 4 Eigen. 0.47 0.37 0.26 % var. 22.65 17.87 12.35 Sig. 0.017 0.004 0.002 Scores Angophora leiocarpa 4.08 Acacia disparrima -2.00 Angophora leiocarpa -2.32

Eucalyptus propinqua 4.08 Angophora leiocarpa 2.88 Banksia serrata 2.60

Pinus sp. * 3.37 Banksia integrifolia -2.13 Callitris collumellaris 2.60

1253_4 -0.95 Eucalyptus crebra -2.92 Eucalyptus propinqua -2.32

1253_8 -0.77 Eucalyptus propinqua 2.88 Pinus sp. * -2.34

1253_18 2.00 Eucalyptus resinifera -2.12 1253_1 1.05

1253_15 1.03 Melaleuca sieberi -3.59 1253_6 0.94

FP -0.66 Pinus sp. * 2.16 1253_13 -0.71

1253_7 -1.16 1253_19 -0.72

1253_15 -1.13 1253_20 0.78

1253_16 -1.23 PS 0.53

1253_18 1.09 ALT -0.59

N 0.61

LR -0.58

PAW -0.75

290

Supplementary Table 3.21: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native tree species abundance for RE 12.5.3.

Cluster Description Sites Important associations 1 Coastal heath 1253_1, 5,19 12 and 20 High numbers of Eucalyptus sites at low racemosa and Corymbia altitudes intermedia across all sites High numbers of Acacia leiocalyx, Callitris columellaris and Eucalyptus pilularis for specific sites Absence of Acacia disparrima

2 Central and 1253_2, 3, 11, 14, 13 Moderate to high numbers of northern sites and 4 Corymbia gummifera and Eucalyptus tindaliae Low numbers of Corymbia intermedia

3 Central 1253_78, 9 and 10 High numbers of Acacia disparrima, woodland/open Corymbia intermedia and forest sites Lophostemon confertus

4 Central and 1253_6, 7, 15, 16 and High numbers of Allocasuarina southern sites 17 littoralis Presence of Banksia integrifolia and Corymbia trachyphloia Low numbers of Eucalyptus racemosa

5 Small fire- 1253_18 High abundance of Melaleuca excluded patch quinquenervia Absence of Eucalyptus racemosa and Corymbia intermedia

291

Supplementary Table 3.22: CCA eigenvalues and scores for significant axes for RE 12.5.3 all tree species abundance data and environmental and disturbance variables. * indicates exotic species.

Axis 1 2 Eigen. 0.54 0.48 % var. 22.92 20.11 Sig. 0.006 0.002 Scores Angophora woodsiana 1.37908 Allocasuarina littoralis 1.77

Eucalyptus propinqua -3.72168 Angophora woodsiana -1.25

Melaleuca -2.54158 Corymbia gummifera -1.40 quinquenervia Pinus sp. * -2.78342 Eucalyptus microcorys 1.92

1253_4 0.856355 Melaleuca sieberi 2.72

1253_18 -2.13755 1253_3 -1.15

FP 0.703559 1253_7 0.97

1253_15 1.53

1253_16 1.39

N -0.64

LR 0.64

PAW 0.56

292

Supplementary Table 3.23: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native shrub species abundance for RE 12.5.3.

Cluster Description Sites Important associations 1 Coastal heath 1253_1, 5 and 20 High numbers of Acacia sites complanata, Austromyrtus dulcis, Brachyloma daphnoides, Epacris obtusifolia, Banksia oblongifolia and Corymbia intermedia Absence of Acacia leiocalyx

2 Mixed central 1253_3, 4 and 15 High numbers of Acacia and northern hubbardiana, Allocasuarina sites littoralis and Banksia oblongifolia Absence of Acacia leiocalyx

3 Mixed northern, 1253_2, 14, 6, 17, 8, High numbers of Acacia central and 9, 13, 19, 10, 12, 18 disparrima, Acacia leiocalyx southern sites and 11 and Alphitonia excelsa

4 Single southern 1253_7 High numbers of Acacia site 1 fimbriata, Corymbia trachyphloia, Eucalyptus crebra and Banksia integrifolia 5 Single southern 1253_16 High numbers of Banksia site 2 spinulosa and Banksia integrifolia

293

Supplementary Table 3.24: CCA eigenvalues and scores for significant axes for RE 12.5.3 shrub species abundance data and environmental and disturbance variables. * indicates exotic species.

Axis 5 Eigen. 0.20 % var. 9.74 Sig. 0.026 Scores Acacia hubbardiana -1.36

Acacia maidenii -1.88

Banksia integrifolia -1.37

Corymbia gummifera -1.34

Corymbia intermedia 1.10

Eucalyptus crebra -2.63

Eucalyptus tindaliae -1.45

Hakea actites 1.35

Hakea florulenta -1.30

Lantana camara * -1.21

Lophostemon confertus -1.12

Lophostemon suaveolens 0.32

Ochna serrulata* 1.68

Persoonia sericea -1.02

Pinus sp. * 2.71

Pipturus argenteus -1.70

Schefflera actinophylla * 4.91

Strangea linearis 1.36

Xanthorrhoea johnsonii 3.74

1253_1 0.64

1253_4 -0.43

1253_7 -0.55

1253_9 -0.56

1253_11 -0.45

1253_18 1.49

294

Axis 5 Eigen. 0.20 % var. 9.74 Sig. 0.026 Scores FP -0.50

ALT -0.45

295

Supplementary Table 3.25: Cluster numbers, cluster descriptions, associated sites and important species for UPGMA analysis of native ground species cover abundance for RE 12.5.3.

Cluster Description Sites Important associations 1 Heathland sites 1253_1, 3, 14, 2, 20, High cover of Ptilothrix deusta with seasonally 12, 13 and 16 across all sites moist substrates 2 Experimental 1253_5 High cover of Daviesia fire plot burnt umbellulata, Strangea every 3 years linearis, Leucopogon juniperinus and Pultenaea myrtoides

3 Woodland/open 1253_4, 7, 11 and 19 High cover of Imperata forest with cylindrica, Pteridium grassy esculentum and Lobelia understorey purpurascens

4 Damp sites with 1253_6, 18 and 15 High cover of Leptocarpus tenax reduced fire across all sites activity and high Most grasses, except Entolasia canopy cover stricta, were absent or had low covers

5 Depauperate 1253_8, 17, 9 and 10 High cover of Lomandra ground layers confertifolia across most due to recent or sites absent fire Low cover of common grass activity species Absence of Themeda triandra

296

Supplementary Table 3.26: CCA eigenvalues and p-values for RE 12.5.3 ground species cover abundance data and environmental and disturbance variables.

Axis Eigenvalue p-value

1 0.51 0.10

2 0.40 0.07

3 0.28 0.63

4 0.22 0.85

5 0.21 0.40

6 0.18 0.43

7 0.14 0.78

8 0.13 0.33

9 0.11 0.23

10 0.09 0.33

11 0.07 0.51

12 0.04 0.70

297

Supplementary Figure 4.1: Flowchart for ANCOVA and linear regression process for ecological condition attributes and environmental and disturbance variables for RE 12.11.5e and RE 12.5.3.

Test attribute and variable (covariate) for homogeneity of variance using Levene’s test

If attribute or variable If attribute or variable (covariate) exhibits homogeneity of (covariate) violates variance then test for homogeneity of regression, i.e. homogeneity of interaction between covariate (environmental, management variance then discard or disturbance variable) and fixed factor (subregion)

If homogeneity of regression is violated (if If homogeneity of regression assumption there is an interaction) then use a linear holds then run full factorial ANCOVA regression for each subregion

Test model for homogeneity of variance using Levene’s test

If model exhibits If model violates homogeneity of homogeneity of variance then variance then discard continue with full factorial ANCOVA

298

Supplementary Table 4.1: Descriptive statistics for raw vegetation condition attribute values for RE 12.11.5e Brisbane and Gympie subregions.

Brisbane Gympie

Attribute N Minimum Maximum Mean Median Standard N Minimum Maximum Mean Median Standard Deviation Deviation Patch size (ha) 24 1.37 44110.12 17260.03 13825.79 18537.96 18 5.97 17328.75 6294.97 952.09 7382.71 Number of large trees 24 25.00 61.00 41.50 39.50 9.78 18 18.00 58.00 33.61 30.00 11.35 Canopy height (m) 24 20.25 37.00 27.83 27.75 4.48 18 21.50 30.50 26.97 27.25 2.16 CWD (m) 24 16.60 91.72 36.81 30.49 20.22 18 8.33 157.00 65.59 59.39 41.95 Native grass cover (%) 24 1.40 60.00 34.53 33.80 18.03 18 7.00 64.00 27.97 23.50 18.03 Litter (%) 24 20.00 73.00 40.83 35.60 16.37 18 27.00 65.00 48.31 49.00 10.23 Weed cover (%) 24 0.00 35.00 8.88 2.50 12.01 18 1.00 35.00 7.28 2.00 10.61 Native canopy cover(m) 24 33.40 104.30 65.74 63.60 20.43 18 33.70 115.50 68.37 68.70 21.73 Native shrub cover (m) 24 11.00 116.10 44.59 40.28 22.24 18 12.95 79.15 39.53 38.53 19.22 Canopy richness 24 2.00 7.00 4.71 5.00 1.02 18 2.00 8.00 5.00 5.00 1.67 Tree richness 24 4.00 10.00 6.96 7.00 1.54 18 4.00 10.00 6.44 6.00 1.57 Shrub richness 24 5.00 19.00 10.42 9.50 3.53 18 5.00 19.00 10.28 9.50 3.46 Grass richness 24 1.00 10.00 7.04 8.00 2.03 18 3.00 12.00 8.11 8.00 2.26 Forbs other richness 24 5.00 29.00 14.92 14.00 5.38 18 13.00 26.00 17.06 17.00 3.21

299

Supplementary Table 4.2: Descriptive statistics for raw vegetation condition attribute values for all RE 12.11.5e sites.

All 12.11.5e Sites

Attribute N Minimum Maximum Mean Median Standard Deviation

Patch size (ha) 42 1.37 44110.12 12560.72 4038.98 15976.74 Number of large trees 42 18.00 61.00 38.12 38.50 11.32 Canopy height (m) 42 20.25 37.00 27.46 27.25 3.74 CWD (m) 42 8.33 157.00 49.15 38.78 34.93 Native grass cover (%) 42 1.40 64.00 31.72 30.80 18.54 Litter (%) 42 20.00 73.00 44.03 46.40 14.72 Weed cover (%) 42 0.00 35.00 8.19 2.00 11.60 Native canopy cover(m) 42 33.40 115.50 66.87 65.15 21.29 Native shrub cover (m) 42 11.00 116.10 42.42 39.98 21.40 Canopy richness 42 2.00 8.00 4.83 5.00 1.36 Tree richness 42 4.00 10.00 6.74 6.50 1.59 Shrub richness 42 5.00 19.00 10.36 9.50 3.55 Grass richness 42 1.00 12.00 7.50 8.00 2.22 Forbs other richness 42 5.00 29.00 15.83 15.50 4.76

300

Supplementary Table 4.3: Descriptive statistics for vegetation condition attributes for RE 12.5.3 sites.

12.5.3 Sites

Attribute N Minimum Maximum Mean Median Standard Deviation

Patch size (ha) 20 1.70 13501.07 2140.75 178.88 4468.40 Number of large trees 20 25.00 65.00 42.50 39.50 10.93 Canopy height (m) 20 15.00 27.00 23.70 25.00 3.13 CWD (m) 20 6.40 61.60 33.70 33.75 13.50 Native grass cover (%) 20 0.00 53.00 22.69 19.50 16.90 Litter (%) 20 11.40 87.80 44.31 42.20 21.29 Weed cover (%) 20 0.00 30.00 3.80 1.00 6.98 Native canopy cover(m) 20 11.00 94.70 53.66 52.90 23.68 Native shrub cover (m) 20 24.30 132.95 55.28 47.55 26.27 Canopy richness 20 2.00 7.00 4.10 4.00 1.67 Tree richness 20 2.00 11.00 6.90 7.00 2.26 Shrub richness 20 2.00 19.00 12.30 11.00 4.82 Grass richness 20 2.00 9.00 5.95 6.00 2.06 Forbs other richness 20 7.00 23.00 15.90 16.00 3.81

301

Supplementary Table 4.4: Results for Shapiro-Wilk tests untransformed vegetation attributes for RE 12.11.5e. Shaded values indicate non-normally distributed data.

Attribute Subregion Statistic df Sig.

Number of large Brisbane .963 24 .498 trees Gympie .928 18 .177 Canopy height Brisbane .975 24 .778 Gympie .938 18 .267 Coarse woody Brisbane .848 24 .002 debris Gympie .937 18 .256 Native grass Brisbane .906 24 .029 cover Gympie .897 18 .051 Litter cover Brisbane .913 24 .041 Gympie .949 18 .407 Weed cover Brisbane .693 24 .000 Gympie .613 18 .000 Native canopy Brisbane .957 24 .381 cover Gympie .959 18 .586 Native shrub Brisbane .905 24 .027 cover Gympie .951 18 .438 Tree species Brisbane .954 24 .323 richness Gympie .923 18 .148 Forbs and other Brisbane .964 24 .531 species richness Gympie .900 18 .058

302

Supplementary Table 4.5: Results for Levene’s tests for untransformed vegetation attributes for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance.

Attribute Levene df 1 df 2 Sig. statistic Number of large .771 1 40 .385 trees Canopy height 10.604 1 40 .002

Coarse woody 6.479 1 40 .015 debris Native grass .052 1 40 .820 cover Litter cover 3.806 1 40 .058

Weed cover .293 1 40 .591

Native canopy .018 1 40 .894 cover Native shrub .001 1 40 .974 cover Tree species .002 1 40 .966 richness Forbs and other 2.881 1 40 .097 species richness

303

Supplementary Table 4.6: Results for Shapiro-Wilk tests for transformed vegetation attributes for RE 12.11.5e. Shaded values indicate non-normally distributed data.

Attribute Subregion Statistic df Sig.

Canopy height Brisbane .980 23 .909 (sqrt) Gympie .930 18 .196 Coarse woody Brisbane .911 23 .042 debris (log) Gympie .913 18 .098 Native grass Brisbane .903 23 .029 cover (arcsin) Gympie .895 18 .047 Litter cover Brisbane .892 23 .018 (arcsin) Gympie .959 18 .583 Weed cover Brisbane .692 23 .000 (arcsin) Gympie .612 18 .000 Native shrub Brisbane .953 23 .345 cover (arcsin) Gympie .940 18 .293

304

Supplementary Table 4.7: Results for Levene’s tests for transformed vegetation attributes for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance.

Attribute Levene df 1 df 2 Sig. statistic Canopy height 8.273 1 39 .006 (sqrt) Coarse woody .114 1 39 .737 debris (log) Native grass .114 1 39 .737 cover (arcsin) Litter cover 2.493 1 39 .122 (arcsin) Weed cover .337 1 39 .565 (arcsin) Native shrub .389 1 39 .536 cover (arcsin)

305

Supplementary Table 4.8: Results for Shapiro-Wilk tests for untransformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate non-normally distributed data.

Attribute Subregion Statistic df Sig.

Patch size Brisbane .762 24 .000 Gympie .729 18 .000 Fire height Brisbane .616 24 .000 Gympie .803 18 .002 Logging Brisbane .833 24 .001 Gympie .554 18 .000 Rural residential Brisbane .567 24 .000 cover Gympie .724 18 .000 Grazing cover Brisbane .337 24 .000 Gympie .623 18 .000 Local road cover Brisbane .857 24 .003 Gympie .904 18 .068 Aspect Brisbane .898 24 .020 Gympie .966 18 .711 Altitude Brisbane .909 24 .034 Gympie .845 18 .007 Average annual Brisbane .950 24 .276 rainfall Gympie .747 18 .000 Rainfall Brisbane .947 24 .229 variability Gympie .893 18 .043 Average annual Brisbane .959 24 .416 temperature Gympie .872 18 .019 Temperature Brisbane .881 24 .009 variability Gympie .947 18 .381 Bulk density Brisbane .835 24 .001 Gympie .901 18 .060 Plant available Brisbane .915 24 .046 water Gympie .793 18 .001

306

Supplementary Table 4.9: Results for Levene’s tests for untransformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance.

Attribute Levene df 1 df 2 Sig. statistic Patch size 11.895 1 40 .001

Fire height .842 1 40 .364

Logging 3.244 1 40 .079

Rural residential .385 1 40 .539 cover Grazing cover 4.960 1 40 .032

Local road cover 10.566 1 40 .002

Aspect 1.065 1 40 .308

Altitude .454 1 40 .504

Average annual 3.666 1 40 .063 rainfall Rainfall 1.127 1 40 .295 variability Average annual 2.416 1 40 .128 temperature Temperature .036 1 40 .851 variability Bulk density 2.644 1 40 .112

Plant available 4.809 1 40 .034 water

307

Supplementary Table 4.10: Results for Shapiro-Wilk tests for transformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate non-normally distributed data.

Attribute Subregion Statistic df Sig.

Patch size (log) Brisbane .835 21 .002 Gympie .897 13 .123 Fire height (sqrt) Brisbane .465 21 .000 Gympie .592 13 .000 Logging (sqrt) Brisbane .818 21 .001 Gympie .311 13 .000 Rural residential Brisbane .510 21 .000 cover (arcsin) Gympie .823 13 .013 Grazing cover Brisbane .352 21 .000 (arcsin) Gympie .644 13 .000 Local road cover Brisbane .818 21 .001 (log) Gympie .976 13 .954 Aspect (cos) Brisbane .885 21 .018 Gympie .825 13 .014 Altitude (log) Brisbane .964 21 .607 Gympie .874 13 .059 Plant available Brisbane .926 21 .114 water (log) Gympie .827 13 .015

308

Supplementary Table 4.11: Results for Levene’s tests for transformed environmental and disturbance variables for RE 12.11.5e. Shaded values indicate data that violates homogeneity of variance.

Attribute Levene df 1 df 2 Sig. statistic Patch size (log) 2.001 1 32 .167

Fire height (sqrt) .099 1 32 .755

Logging (sqrt) 5.853 1 32 .021

Rural residential .794 1 32 .380 cover (arcsin) Grazing cover 5.170 1 32 .030 (arcsin) Local road cover 4.190 1 32 .049 (log) Aspect (cos) .030 1 32 .863

Altitude (log) 5.174 1 32 .030

Plant available 2.684 1 32 .111 water

309

Supplementary Table 4.12: Results for Shapiro-Wilk tests for untransformed vegetation attributes for RE 12.5.3. Shaded values indicate non-normally distributed data.

Attribute Statistic df Sig.

Number of large .933 20 .178 trees Canopy height .827 20 .002

Weed cover .579 20 .000

Native shrub cover .866 20 .010

Canopy species .882 20 .019 richness Tree species .973 20 .821 richness Shrub species .909 20 .060 richness Grass species .945 20 .300 richness

310

Supplementary Table 4.13: Results for Shapiro-Wilk tests for transformed vegetation attributes for RE 12.5.3. Shaded values indicate non-normally distributed data.

Attribute Statistic df Sig.

Canopy height (sqrt) .816 19 .002

Weed cover (arcsin) .587 19 .000

Native shrub cover .920 19 .115 (sqrt) Canopy species .887 19 .029 richness (sqrt)

311

Supplementary Table 4.14: Results for Shapiro-Wilk tests for untransformed environmental and disturbance variables for RE 12.5.3. Shaded values indicate non-normally distributed data.

Attribute Statistic df Sig.

Northing .923 20 .112

Patch size .501 20 .000

Fire proportion .447 20 .000

Fire height .728 20 .000

Rainfall variability .907 20 .057

Temperature .877 20 .015 variability Plant Available .809 20 .001 Water

312

Supplementary Table 4.15: Results for Shapiro-Wilk tests for transformed environmental and disturbance variables for RE 12.5.3. Shaded values indicate non-normally distributed data.

Attribute Statistic df Sig.

Patch size (log) .939 20 .226

Fire proportion .447 20 .000 (arcsin) Fire height (sqrt) .722 20 .000

Temperature .872 20 .013 variability (sqrt) Plant Available .896 20 .034 Water (log)

313

Supplementary Figure 5.1: Flowchart of statistical tests and outcomes for Chapter 5.

314

Supplementary Table 5.1: A summary of site codes, mean NDSI values, bird species richness and BioCondition attributes for RE 12.11.5e soundscape survey sites.

Site Mean Bird B.C. Patch Patch Patch No. Can. Rec. (%) CWD Grass Litter Weed Tree Shrub Grass Forb Can. Shrub code NDSI sp. score size (ha) conn. cont. large height (m) cover cover cover sp. sp. sp. sp. Cover cover (24 rich. (%) (%) trees (m) (%) (%) (%) rich. rich. rich. rich. (%) (%) hours) BRIS1 0.51 39 0.95 44110.12 93.03 97.40 40 29.00 100.00 45.73 49.00 24.00 0.00 7 10 8 29 47.90 22.60

BRIS2 0.64 28 0.88 44110.12 97.79 97.64 37 35.50 100.00 41.25 52.00 33.40 3.00 6 9 5 14 56.10 34.45

BRIS3 0.57 48 0.81 44110.12 92.25 64.74 38 33.50 86.00 44.15 55.40 20.80 35.00 7 9 5 17 76.60 30.35

BRIS4 0.46 45 0.91 44110.12 82.79 96.45 52 23.50 100.00 31.70 35.00 34.00 0.00 7 8 5 22 104.30 39.00

BRIS6 0.58 36 0.82 44110.12 82.79 92.37 25 25.00 100.00 32.40 21.00 38.00 15.00 6 18 8 15 84.60 42.50

BRIS17 0.39 15 0.84 125.09 8.65 35.77 34 32.50 100.00 29.27 60.00 23.20 1.00 9 11 7 9 74.50 42.65

BRIS18 -0.04 15 0.83 125.09 8.65 23.91 61 27.00 100.00 19.16 47.40 47.00 2.00 10 13 9 8 74.80 50.45

BRIS19 0.20 27 0.65 3.44 5.77 49.12 33 20.25 100.00 70.14 53.00 27.20 25.00 8 5 10 17 57.25 54.60

BRIS20 0.02 30 0.55 3.94 0.00 2.26 39 31.25 100.00 19.22 18.40 73.00 1.00 7 13 5 5 34.80 71.05

BRIS21 0.07 26 0.66 5.18 53.40 8.30 26 33.50 100.00 16.60 54.00 20.00 25.00 8 13 6 13 62.50 71.30

315

Supplementary Table 5.2: A summary of site codes, mean NDSI values, bird species richness and BioCondition attributes for RE 12.5.3 soundscape survey sites.

Site code Mean Bird B.C. Patch Patch Patch No. Can. Rec. (%) CWD Grass Litter Weed Tree Shrub Grass Forb Can. Shrub NDSI sp. score size (ha) conn. cont. large height (m) cover cover cover sp. sp. sp. sp. Cover cover (24 rich. (%) (%) trees (m) (%) (%) (%) rich. rich. rich. rich. (%) (%) hours) 1253_3 0.62 27 0.91 955.72 78.50 60.23 36 24.00 100.00 34.30 42.00 11.40 0.00 5 19 4 15 39.20 47.15

1253_12 -0.10 21 0.94 358.60 57.31 37.78 40 25.00 100.00 33.20 53.00 32.20 0.00 6 8 9 21 55.50 27.30

1253_14 0.58 32 0.82 124.90 89.81 48.86 25 23.00 100.00 26.14 14.00 28.00 0.00 6 18 8 19 47.40 82.30

1253_4 0.48 24 0.96 1350.56 63.16 57.73 42 27.00 100.00 42.00 39.00 45.20 1.00 5 19 6 17 44.50 68.20

1253_18 -0.36 20 0.63 2.13 18.87 6.72 39 15.00 50.00 25.15 12.40 28.60 30.00 8 10 6 18 66.60 35.45

1253_20 -0.27 33 0.64 13.48 41.50 33.12 41 22.00 100.00 6.40 10.00 56.00 5.00 8 11 2 16 53.80 33.10

1253_5 0.56 36 0.87 730.10 100.00 93.42 26 27.00 100.00 50.45 26.60 13.00 0.00 2 10 5 19 54.50 46.30

1253_13 0.60 32 0.88 232.85 55.83 50.65 36 25.00 71.00 27.80 27.00 40.20 1.00 9 19 8 18 36.50 62.20

1253_19 -0.40 11 0.73 5.33 20.02 23.25 39 27.00 100.00 9.70 27.00 46.60 15.00 7 16 5 10 66.10 33.00

316

Supplementary Table 5.3: Table of all bird species including common names, scientific names and site locations for soundscape survey sites in RE 12.11.5e.

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Bird Species Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Australasian Sphecotheres 0 0 1 1 1 0 1 0 0 1 Figbird vieilloti Australian King Alisterus 0 1 1 0 0 0 0 0 0 0 Parrot scapularis Australian Eudynamys 0 0 0 0 0 0 0 0 0 0 Koel orientalis Australian Gymnorhina 1 0 0 1 1 1 1 1 1 0 Magpie tibicen Australian Owlet- Aegotheles 0 0 0 0 0 0 0 0 0 0 Nightjar cristatus Australian Wood Chenonetta 0 1 0 0 0 0 0 0 0 0 Duck jubata Bar-shouldered Geopelia 1 0 0 0 0 0 0 0 0 0 Dove humeralis Bell Miner Manorina 0 1 0 0 0 0 0 0 0 0 melanophrys Black-faced Coracina 1 0 1 1 1 1 1 0 0 1 Cuckoo-shrike novaehollandiae Black-faced Monarcha 1 0 1 1 0 1 0 1 0 0 Monarch melanopsis Blue-faced Entomyzon 0 0 0 0 0 0 1 1 1 1 Honeyeater cyanotis

317

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Common name Scientific name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Brown Cuckoo- Macropygia 1 1 1 0 0 0 0 0 0 0 dove amboinensis Brown Goshawk Accipiter fasciatus 0 0 0 0 0 0 1 0 0 0

Brown Lichmera indistincta 0 0 0 0 0 0 1 0 0 0 Honeyeater Brown Thornbill Acanthiza pusilla 0 0 0 0 1 0 0 0 0 0

Brush Cuckoo Cacomantis 0 0 0 0 0 0 0 0 0 0 variolosus Buff-banded Gallirallus 0 0 0 0 0 0 0 0 0 0 Rail philippensis Bush Stone- Burhinus grallarius 0 0 0 1 0 0 0 0 0 0 curlew Channel-billed Scythrops 0 0 0 0 0 0 0 0 0 0 Cuckoo novaehollandiae Collared Accipiter 1 0 1 0 0 0 1 1 0 0 Sparrowhawk cirrocephalus Common Phaps chalcoptera 0 0 0 1 0 0 0 0 0 0 Bronzewing Common Coracina 0 0 0 0 0 0 0 0 0 0 Cicadabird tenuirostris Crested Pigeon Ocyphaps lophotes 1 0 0 0 0 0 0 0 0 0

Crested Shrike-tit Falcunculus 0 0 0 1 0 0 0 0 0 0 frontatus Eastern Koel Eudynamys 0 0 0 0 1 0 0 0 0 0 orientalis

318

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Eastern Whipbird Psophodes 1 1 1 1 1 0 0 1 0 0 olivaceus Eastern Yellow Eopsaltria 1 1 1 1 1 0 0 0 0 0 Robin australis Fairy Gerygone Gerygone 0 0 0 0 0 0 0 0 0 1 palpebrosa Fan-Tailed Cuckoo Cacomantis 0 0 0 1 1 0 1 1 0 0 flabelliformis Forest Kingfisher Todiramphus 1 1 0 0 0 1 0 0 0 0 macleayii Galah Eolophus 0 0 1 1 0 0 0 0 0 0 roseicapilla Golden Whistler Pachycephala 1 0 0 0 1 1 0 0 1 0 pectoralis Grey Butcherbird Cracticus 1 0 1 1 1 0 0 1 1 1 torquatus Grey Fantail Rhipidura 1 1 1 1 1 0 1 1 0 0 albiscapa Grey Goshawk Accipiter 0 0 0 0 0 0 0 0 0 0 novaehollandiae Grey Shrike-thrush Colluricincla 1 1 1 1 0 1 1 0 0 1 harmonica Large-billed Sericornis 0 0 0 0 0 0 0 0 0 0 Scrubwren magnirostra Laughing Dacelo 0 0 1 1 1 0 0 0 1 0 Kookaburra novaeguineae Leaden Flycatcher Myiagra rubecula 1 0 1 1 1 1 0 0 0 0

319

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Lewin's Honeyeater Meliphaga lewinii 1 1 1 1 1 0 0 0 0 0

Little Bronze- Chrysococcyx 0 0 1 1 0 0 0 0 0 0 Cuckoo minutillus Little Corella Cacatua 0 0 0 0 0 0 0 0 0 0 sanguinea Little Friarbird Philemon 0 0 0 0 0 0 0 0 0 0 citreogularis Little Shrike-thrush Colluricincla 0 0 1 1 1 0 0 0 0 0 megarhyncha Magpie-lark Grallina 0 0 1 0 0 0 1 1 0 1 cyanoleuca Masked Lapwing Vanellus miles 0 0 1 0 0 0 0 0 0 0

Mistletoebird Dicaeum 0 0 1 0 0 1 1 1 0 0 hirundinaceum Mystery_1 NA 0 0 1 0 1 0 0 0 0 0

Mystery_2 NA 0 0 0 1 0 0 0 0 0 0

Noisy Friarbird Philemon 1 0 1 1 1 1 1 1 1 1 corniculatus Noisy Miner Manorina 1 0 1 1 1 1 1 1 1 1 melanocephala Noisy Pitta Pitta versicolor 0 0 0 1 0 0 0 0 0 0

Olive-backed Oriole Oriolus sagittatus 1 1 1 1 1 1 1 1 1 1

320

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Pale-headed Platycercus 1 0 1 1 1 1 1 1 0 0 Rosella adscitus Pale-vented Bush- Amaurornis 0 0 0 0 0 0 0 0 0 0 hen moluccana Peaceful Dove Geopelia striata 0 0 0 0 0 0 0 0 0 0

Pheasant Coucal Centropus 0 0 1 0 0 0 0 0 0 0 phasianinus Pied Butcherbird Cracticus 1 1 1 1 1 0 1 1 1 1 nigrogularis Pied Currawong Strepera 1 1 1 1 1 0 1 1 1 1 graculina Rainbow Bee-eater Merops ornatus 0 0 1 1 0 0 0 0 0 0

Rainbow Lorikeet Trichoglossus 1 1 1 1 1 1 1 1 1 1 haematodus Red-backed Fairy- Malurus 1 1 1 1 0 1 0 0 1 0 wren melanocephalus Red-browed Finch Neochmia 1 1 0 1 1 1 0 0 0 0 temporalis Rufous Fantail Pachycephala 0 0 0 0 0 0 0 0 0 0 rufiventris Rufous Shrike- Colluricincla 0 0 0 0 0 0 0 0 0 0 thrush megarhyncha Rufous Whistler Todiramphus 0 1 1 1 0 0 1 0 0 0 sanctus Sacred Kingfisher Trichoglossus 0 0 1 0 0 1 0 1 0 0 chlorolepidotus

321

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Scaly-breasted Myzomela 1 1 1 1 1 1 1 1 1 1 Lorikeet sanguinolenta Scarlet Honeyeater Chrysococcyx 1 1 1 1 1 1 0 1 0 0 lucidus Shining Bronze- Chrysococcyx 1 0 0 0 0 0 0 0 0 0 cuckoo lucidus Silvereye Zosterops lateralis 1 1 1 1 1 1 1 1 0 0

Spangled Drongo Dicrurus 1 0 0 1 1 0 0 0 0 0 bracteatus Spectacled Symposiarchus 0 0 0 0 0 0 0 0 0 0 Monarch trivirgatus Spotted Pardalote Pardalotus 1 0 1 1 0 0 0 0 0 0 punctatus Spotted Turtle- Streptopelia 0 0 0 0 0 1 1 1 0 0 dove chinensis Spotted Quail- Cinclosoma 0 0 0 0 0 0 0 0 0 0 thrush punctatum Striated Pardalote Pardalotus 1 1 1 1 1 1 1 1 1 0 striatus Sulphur-crested Cacatua galerita 1 1 1 1 1 0 1 0 0 0 Cockatoo Superb Fairy- Malurus cyaneus 0 0 0 0 0 0 0 0 0 0 wren Tawny Grassbird Megalurus 0 0 1 0 0 0 0 0 0 0 timoriensis Torresian Crow Corvus orru 1 1 1 1 1 1 1 1 1 1

322

BRIS1 BRIS2 BRIS3 BRIS4 BRIS6 BRIS17 BRIS18 BRIS19 BRIS20 BRIS21

Scientific Name D’Aguilar 1 D’Aguilar 2 D’Aguilar 3 Moggill 1 Moggill 3 Mount Nursery Wellers Hill Whites Hill Whites Hill Warren Road 1 2 Park Varied Triller Lalage leucomela 0 0 0 1 1 0 0 0 0 0

Variegated Fairy- Malurus lamberti 0 1 1 0 1 1 1 0 0 0 wren Wandering Dendrocygna 0 0 0 0 0 0 0 0 0 0 Whistling-duck arcuata White-bellied Coracina 0 0 1 0 0 0 0 0 0 0 Cuckoo-shrike papuensis White-browed Sericornis 0 0 1 0 1 1 1 0 0 0 Scrubwren frontalis White-browed Artamus 0 0 0 0 0 1 0 0 0 0 Woodswallow superciliosus White-throated Gerygone 1 0 0 1 0 0 0 0 0 0 Gerygone olivacea White-throated Melithreptus 1 1 1 1 1 1 1 1 0 0 Honeyeater albogularis White-throated Cormobates 1 1 1 1 1 0 0 0 0 0 Treecreeper leucophaea Willie-wagtail Rhipidura 1 0 0 0 0 0 1 0 0 0 leucophrys Wonga Pigeon Leucosarcia 0 1 1 1 0 0 0 0 0 0 melanoleuca Yellow-faced Lichenostomus 1 1 1 0 0 1 0 1 0 0 Honeyeater chrysops Yellow-tailed Black- Calyptorhynchus 0 0 0 0 0 0 0 0 0 0 Cockatoo funereus

323

Supplementary Table 5.4: Table of all bird species including common names, scientific names and site locations for soundscape survey sites in RE 12.5.3.

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Australasian Sphecotheres 0 0 1 0 1 1 0 0 1 Figbird vieilloti Australian King Alisterus 1 0 0 0 1 0 0 1 0 Parrot scapularis Australian Koel Eudynamys 0 0 1 1 0 0 1 0 1 orientalis Australian Gymnorhina 0 0 0 1 0 1 0 0 0 Magpie tibicen Australian Owlet- Aegotheles 1 0 1 0 0 0 0 0 0 nightjar cristatus Australian Wood Chenonetta 0 0 0 0 0 0 0 0 0 Duck jubata Bar-shouldered Geopelia 0 0 1 0 1 0 0 1 0 Dove humeralis Bell Miner Manorina 0 0 0 0 0 0 0 0 0 melanophrys Black-faced Coracina 0 0 0 0 0 0 0 0 0 Cuckoo-shrike novaehollandiae Black-faced Monarcha 0 0 1 0 1 0 1 1 0 Monarch melanopsis Blue-faced Entomyzon 0 0 0 0 1 1 0 0 0 Honeyeater cyanotis Brown Cuckoo- Macropygia 0 0 0 0 0 0 0 1 0 dove amboinensis

324

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Brown Goshawk Accipiter fasciatus 0 0 0 0 0 0 0 0 0

Brown Lichmera 0 0 0 0 0 0 0 0 0 Honeyeater indistincta Brown Thornbill Acanthiza pusilla 0 0 0 0 1 0 0 0 0

Brush Cuckoo Cacomantis 0 0 0 0 1 1 1 0 0 variolosus Buff-banded Gallirallus 0 0 0 0 0 1 0 0 0 Rail philippensis Bush Stone- Burhinus 0 0 0 0 0 0 0 0 0 curlew grallarius Channel-billed Scythrops 0 0 0 1 0 1 0 0 0 Cuckoo novaehollandiae Collared Accipiter 0 0 0 0 0 0 0 0 0 Sparrowhawk cirrocephalus Common Phaps chalcoptera 0 0 0 0 0 0 0 0 0 Bronzewing Common Coracina 0 1 0 0 0 0 0 0 0 Cicadabird tenuirostris Crested Pigeon Ocyphaps 0 0 0 0 0 0 0 0 0 lophotes Crested Shrike-tit Falcunculus 0 0 0 0 0 0 0 0 0 frontatus Eastern Koel Eudynamys 0 0 0 0 0 0 0 0 0 orientalis Eastern Whipbird Psophodes 0 0 0 0 0 0 0 0 0 olivaceus

325

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Eastern Yellow Eopsaltria 0 1 0 0 1 0 0 1 0 Robin australis Fairy Gerygone Gerygone 0 0 0 0 0 0 0 0 0 palpebrosa Fan-tailed Cacomantis 1 0 0 0 1 0 1 1 0 Cuckoo flabelliformis Forest Kingfisher Todiramphus 0 0 1 0 1 0 1 0 0 macleayii Galah Eolophus 0 1 1 1 0 0 1 0 0 roseicapilla Golden Whistler Pachycephala 0 0 0 0 0 0 0 0 0 pectoralis Grey Butcherbird Cracticus 1 0 0 1 1 1 0 1 1 torquatus Grey Fantail Rhipidura 0 0 1 0 0 0 1 1 0 albiscapa Grey Accipiter 0 0 0 0 0 0 1 0 0 Goshawk novaehollandiae Grey Shrike- Colluricincla 1 1 1 1 0 0 1 1 0 thrush harmonica Large-billed Sericornis 0 0 0 0 1 0 0 0 0 Scrubwren magnirostra Laughing Dacelo 1 0 0 1 0 0 0 0 0 Kookaburra novaeguineae Leaden Myiagra rubecula 1 1 1 1 0 0 1 1 0 Flycatcher Lewin's Meliphaga lewinii 1 0 1 1 1 0 1 1 0 Honeyeater

326

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Little Bronze- Chrysococcyx 0 1 0 0 0 0 1 0 0 cuckoo minutillus Little Corella Cacatua 0 0 0 0 0 0 0 1 0 sanguinea Little Philemon 0 0 0 0 0 0 1 1 0 Friarbird citreogularis Little Shrike- Colluricincla 0 0 0 0 0 0 0 0 0 thrush megarhyncha Magpie-lark Grallina 0 0 0 0 1 1 0 0 1 cyanoleuca Masked Lapwing Vanellus miles 0 0 0 0 0 1 0 0 0

Mistletoebird Dicaeum 0 0 0 1 1 0 1 0 0 hirundinaceum Mystery_1 NA 0 0 0 0 0 0 0 0 0

Mystery_2 NA 0 0 0 0 0 0 0 0 0

Noisy Friarbird Philemon 0 0 0 0 1 0 1 0 0 corniculatus Noisy Miner Manorina 0 1 0 0 1 1 0 0 1 melanocephala Noisy Pitta Pitta versicolor 0 0 0 0 0 0 0 0 0

Olive-backed Oriolus sagittatus 1 1 1 1 1 1 0 1 0 Oriole Pale-headed Platycercus 0 0 0 0 0 0 0 0 0 Rosella adscitus

327

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Pale-vented Amaurornis 0 0 0 0 0 1 0 0 0 Bush-hen moluccana Peaceful Dove Geopelia striata 1 1 1 0 0 1 1 1 0

Pheasant Coucal Centropus 0 0 0 1 0 0 0 1 0 phasianinus Pied Butcherbird Cracticus 1 1 1 1 0 0 0 1 0 nigrogularis Pied Currawong Strepera 0 0 1 0 1 0 0 1 0 graculina Rainbow Bee- Merops ornatus 0 0 0 0 0 0 0 0 0 eater Rainbow Lorikeet Trichoglossus 1 1 0 1 1 1 1 0 1 haematodus Red-backed Malurus 1 0 1 1 0 0 1 1 0 Fairy-wren melanocephalus Red-browed Neochmia 1 0 1 0 1 0 0 1 0 Finch temporalis Rufous Pachycephala 0 0 1 0 0 0 0 0 0 Fantail rufiventris Rufous Shrike- Colluricincla 0 0 0 0 1 0 1 1 0 thrush megarhyncha Rufous Whistler Todiramphus 1 1 1 1 0 0 1 1 0 sanctus Sacred Kingfisher Trichoglossus 1 1 0 0 0 0 1 0 1 chlorolepidotus Scaly-breasted Myzomela 1 1 1 0 1 1 1 1 1 Lorikeet sanguinolenta

328

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Scarlet Chrysococcyx 1 0 1 0 1 0 1 0 0 Honeyeater lucidus Shining Bronze- Chrysococcyx 1 0 1 1 1 0 1 0 0 cuckoo lucidus Silvereye Zosterops lateralis 0 0 1 1 0 0 0 1 0

Spangled Drongo Dicrurus 0 0 0 0 0 0 1 1 0 bracteatus Spectacled Symposiarchus 0 0 1 0 0 0 0 1 0 Monarch trivirgatus Spotted Pardalotus 0 0 0 0 0 0 1 0 0 Pardalote punctatus Spotted Turtle- Streptopelia 0 0 0 0 0 0 0 0 0 dove chinensis Spotted Quail- Cinclosoma 0 0 1 0 0 0 0 0 0 thrush punctatum Striated Pardalotus 1 1 1 1 0 0 1 1 0 Pardalote striatus Sulphur-crested Cacatua galerita 0 1 0 0 1 1 0 0 0 Cockatoo Superb Fairy- Malurus cyaneus 0 0 0 0 0 0 0 0 1 wren Tawny Grassbird Megalurus 1 0 0 0 0 0 1 0 0 timoriensis Torresian Crow Corvus orru 1 1 0 1 1 1 1 1 1

Varied Triller Lalage leucomela 0 0 0 0 0 0 0 0 0

329

1253_3 1253_12 1253_14 1253_4 1253_20 1253_18 1253_5 1253_13 1253_19

Common name Scientific name Caves Road Freshwater Kings Road Murphys Newells Road New SA1 Saddleback Steve Irwin Road Settlement Way Road Variegated Fairy- Malurus lamberti 0 0 1 0 0 0 1 0 0 wren Wandering Dendrocygna 0 0 0 0 0 1 0 0 0 Whistling-duck arcuata White-bellied Coracina 0 0 0 0 0 0 0 0 0 Cuckoo-shrike papuensis White-browed Sericornis 1 0 0 0 1 0 0 0 0 Scrubwren frontalis White-browed Artamus 0 0 0 0 0 0 0 0 0 Woodswallow superciliosus White-throated Gerygone 0 1 0 0 1 0 1 1 0 Gerygone olivacea White-throated Melithreptus 1 1 1 1 1 1 1 0 0 Honeyeater albogularis White-throated Cormobates 1 1 1 1 1 0 1 1 0 Treecreeper leucophaea Willie-wagtail Rhipidura 0 0 1 0 0 1 0 0 1 leucophrys Wonga Pigeon Leucosarcia 0 0 0 0 0 0 0 0 0 melanoleuca Yellow-faced Lichenostomus 1 1 1 1 1 0 1 1 0 Honeyeater chrysops Yellow-tailed Black- Calyptorhynchus 1 0 0 0 0 0 1 0 0 Cockatoo funereus

330

Supplementary Table 5.5: Descriptive statistics for bird species richness and mean NDSI values for soundscape survey sites in RE 12.11.5e and RE 12.5.3.

RE 12.11.5e RE 12.5.3

N Minimum Maximum Mean Median Standard N Minimum Maximum Mean Median Standard Deviation Deviation Bird species richness 10 15.00 48.00 30.00 28.00 10.87 9 11.00 36.00 26.22 27.00 7.54

Mean NDSI (24 hours) 10 -0.13 0.65 0.35 0.41 0.25 9 -0.37 0.60 0.27 0.49 0.36

331

Landscape attributes Tree- and shrub-related structural attributes Ground-related structural attributes Species richness-related attributes

Patch Patch Patch Large Canopy Recruit. Canopy Shrub CWD Grass Litter Weed Tree Shrub Grass Forb size (PS) conn. cont. trees height (R) cover cover cover cover cover rich. rich. rich. rich. (PC) (PX) (LT) (CH) (CC) (SC) (GC) (LC) (WC) (TR) (SR) (GR) (FR) Mean NDSI Z=2.452 Z=2.122 Z=2.382 Z=-0.200 Z=0.091 Z=-0.522 Z=0.381 Z=-2.455 Z=1.727 Z=-0.636 Z=0.236 Z=-0.606 Z=-2.398 Z=-0.148 Z=-0.546 Z=1.587 (24 hrs) p=.006 p=0.030 p=0.009 p=0.868 p=0.936 p=0.799 p=0.739 p= 0.005 p=0.090 p=0.554 p=0.838 p=0.575 p=0.008 p=0.903 p=0.611 p=0.123 %v=48.66 %v=18.73 %v=32.62 %v=90.39 %v=56.95

332

Landscape attributes Tree- and shrub-related structural attributes Ground-related structural attributes Species richness-related attributes

Patch Patch Patch Large Canopy Recruit. Canopy Shrub CWD Grass Litter Weed Tree Shrub Grass Forb size (PS) conn. cont. trees height (R) cover cover cover cover cover rich. rich. rich. rich. (PC) (PX) (LT) (CH) (CC) (SC) (GC) (LC) (WC) (TR) (SR) (GR) (FR) Mean NDSI Z=2.404 Z=2.215 Z=2.640 Z=-1.069 Z=0.771 Z=-0.387 Z=-1.839 Z=1.697 Z=2.498 Z=1.207 Z=-1.980 Z=-1.954 Z=-1.862 Z=1.228 Z=0.000 Z=0.523 (24 hrs) p=.006 p=0.016 p=0.001 p=0.321 p=0.470 p=0.742 p=0.068 p= 0.102 p=0.003 p=0.255 p=0.041 p=0.049 p=0.063 p=0.244 p=1.000 p=0.632 %v=45.46 %v=31.85 %v=22.69 %v=45.47 %v=13.46 %v=35.57

333

Attribute RE Statistic df Sig.

Patch size 12115e .656 10 .000 1253 .850 9 .074 Patch connectivity 12115e .803 10 .016 1253 .950 9 .691 Patch context 12115e .875 10 .114 1253 .973 9 .918 Coarse woody 12115e .908 10 .266 debris 1253 .958 9 .777 Litter cover 12115e .815 10 .022 1253 .951 9 .706 Weed cover 12115e .789 10 .011 1253 .655 9 .000 Tree species 12115e .903 10 .238 richness 1253 .941 9 .588 Native shrub cover 12115e .939 10 .537 1253 .910 9 .315

334

Attribute Levene df 1 df 2 Sig. statistic Patch size 35755.135 1 17 .000

Patch 2.888 1 17 .107 connectivity Patch context 4.127 1 17 .058

Coarse woody .105 1 17 .749 debris Litter cover .033 1 17 .858

Weed cover .711 1 17 .411

Tree species 1.497 1 17 .238 richness Native shrub .245 1 17 .627 cover

335

Attribute RE Statistic df Sig.

Patch size (log) 12115e .760 10 .005 1253 .890 9 .199 Patch connectivity 12115e .847 10 .054 (arcsin) 1253 .933 9 .510 Litter cover (arcsin) 12115e .783 10 .009 1253 .956 9 .751 Weed cover (arcsin) 12115e .789 10 .011 1253 .653 9 .000

336

Attribute Levene df 1 df 2 Sig. statistic Patch size (log) 7.852 1 17 .012

Patch context 1.954 1 17 .180 (arcsin)

Litter cover .008 1 17 .932 (arcsin) Weed cover .712 1 17 .411 (arcsin)

337