Does post-mining rehabilitation on the Weipa bauxite plateau restore bird habitat values?

Transcription

1 Does post-mining rehabilitation on the Weipa bauxite plateau restore bird habitat values? by Susan F. Gould Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy of the Australian National University February 200

2 Candidate's Declaration ii This thesis contains no material which has been accepted for the award of any other degree or diploma in any university. To the best of the author s knowledge, it contains no material previously published or written by another person, except where due reference is made in the text. Susan F. Gould Date: June 2009

3 Dedication iii This work was inspired by love and gratitude for the birds of the Weipa bauxite plateau, who make the forests ring with song, but have no voice in shaping the plans of mankind for their forest homes. This thesis is dedicated to them.

4 Acknowledgements iv It has been a privilege and a pleasure to be mentored by people who are experts in their respective fields, clear thinkers, and fine people to boot. I thank all of the members of my supervisory panel for their generosity in mentoring me through this process: Professor Brendan Mackey (chair) - Fenner School of Environment and Society at ANU, Professor Henry Nix - Fenner School of Environment and Society at ANU, Dr John Ludwig - CSIRO Sustainable Ecosystems Tropical Forest Research Centre at Atherton, David Tongway - Visiting Fellow Fenner School of Environment and Society at ANU, and Dr Penny Olsen - Visiting Fellow Botany and Zoology at ANU. Dr Ross Cunningham also provided early advice on research design before retiring from ANU. Conducting this research was made possible through financial support from an ANU Graduate School Scholarship. Rio Tinto Aluminium Weipa provided substantial financial and in-kind support. Thanks are due to Neale Dahl, Sinead Kaufman, Owen Fisher, Flo New, Sandy Eldridge, Des Land and Haakon Nielsen from the Rio Tinto environment team. Thanks too, to the crew leaders, superintendents and operators in the mine who provided direction and practical assistance as needed. CSIRO Sustainable Ecosystems Tropical Forest Research Centre at Atherton also supported this research by providing affiliated researcher status, supervision and library assistance. Special thanks are due to the team of volunteers who assisted with vegetation sampling under arduous conditions making the work not only bearable but enjoyable: Rod Thorn, Merinda Nash, Amelia Elgar, Michael Barnett, David Wingrove, Nyree Gay, John Doherty, Chris Masters and Owen Pittman generously volunteered their time. Sandy Eldridge and Flo New from Rio Tinto also provided valued assistance with field work. Thanks to the many other people who provided assistance in various ways during the course of this research: Emlyn Williams and Bob Forrester (ANU Statistical Consulting Unit); Colin Pain (GeoScience Australia); Peter Thompson and Aaron Crosby (CYPDA-Sustainable Fire Project); Gary Wilson, John Clarkson, John Neldner, Rosemary Niehus and Paula Peeters (Queensland EPA); Clive Hilliker, Steve Leahy and Karl Nissen (ANU Fenner School of Environment and Society); the support staff at the Fenner School; and the ANU library staff who provided a very efficient off-campus document delivery service. Conversations with other Fenner school students were also very helpful. Thanks to Adam Leavesley, Nicola Munro and Julian Reid. Finally, special thanks are due to Henry and Katharine Nix who generously shared their home and family, and even provided a dog to walk and cuddle when I was visiting Canberra. None of this would have been possible without the extraordinary support of my husband Rod.

5 Abstract v Rehabilitation is seen as a key strategy for minimising biodiversity losses. Although most rehabilitation strategies aim to provide habitat for fauna, they usually focus entirely on establishing vegetation. Successful vegetation establishment, however, does not necessarily provide habitat to the same species that are threatened by habitat loss. Improved understanding of faunal response to rehabilitation is required if rehabilitation techniques are to be refined and deliver the hoped for biodiversity outcomes. This study aimed to assess to what extent postmining rehabilitation on the Weipa bauxite plateau has restored the bird habitat values of the pre-mining native forest. Bird assemblages, vegetation, and landscape functionality were compared between: () Eucalyptus tetrodonta open forest reference sites representative of the pre-mining native forest; (2) two reference land units of Eucalyptus tetrodonta tall woodland that have previously been nominated as ecologically appropriate analogues for the post-mining landscapes; and (3) a chronosequence of post-mining rehabilitation sites up to 23 years old. Bird species richness and mean bird abundance increased with rehabilitation age. Bird species composition also changed and became more similar to native forest bird assemblages with increasing age. Significant differences remained, however, in mean bird abundance and composition of the bird assemblages between the oldest age class of mine rehabilitation and reference native forest land units. The mean bird species shortfall index in the oldest age class of mine rehabilitation was 63%, compared to a mean species shortfall index of 27% for premining native forest sites. There were also significant differences in vegetation composition and structure between reference native forest land units and post-mining rehabilitation sites. Most importantly, the framework plant species that dominate the native vegetation community occurred at much lower densities in mine rehabilitation than in reference land units. Site detection rates of birds were strongly related to vegetation composition and structure. It is concluded that mining and post-mining rehabilitation on the Weipa bauxite plateau has so far resulted in habitat conversion rather than habitat restoration. It therefore contributes to the causes of biodiversity decline. Post-mining rehabilitation created new habitat for 8 bird species not sourced from the pre-mining native forest. It also provided partial habitat for many of the generalist native forest bird species recorded, although their presence in the landscape remained dependent on access to native forest. However, rehabilitation did not provide the habitat resources that are required by habitat specialists and foraging specialists. This study found that the native forest bird species most sensitive to habitat loss, and most in need of habitat restoration, may be the last to return to rehabilitation if they return at all. The findings of this study have implications for rehabilitation practices, biodiversity conservation on the Weipa bauxite plateau, as well as broader implications for policies that rely on the assumption that rehabilitation can offset biodiversity losses.

6 Table of Contents vi Candidate's Declaration... ii Dedication... iii Acknowledgements... iv Abstract... v Table of Contents... vi List of Figures... xi List of Tables... xii Chapter : Introduction.... Research rationale and aim....2 Research questions Thesis outline... 3 Chapter 2: The study area Location Geology Climate Hydrology Vegetation Fire in the tropical savannas Effects of fire on vegetation in the tropical savannas Birds of the Eucalyptus tetrodonta open forest Mining on the Weipa bauxite plateau Conclusion Chapter 3: Bird-environment relationships Introduction The amount of habitat resources Plant productivity, habitat resources and bird species richness Area The variety of habitat resources Environmental heterogeneity and bird species richness Habitat diversity and bird species richness Bird-habitat relationships in Eucalyptus formations Vegetation structural complexity and bird species richness The effect of disturbance on environmental heterogeneity Partitioning of habitat resources... 4

7 vii 3.4. Stability of habitat resources Stability of habitat resources and Australian bird assemblages Habitat resources: a conceptual framework The habitat templet The effects of spatial and temporal scale on pattern and process Conclusion Chapter 4: Succession in post-mining rehabilitation Introduction Vegetation succession Succession and mine rehabilitation Bird assembly in successional environments Bird species richness Turnover in bird species composition Changes in bird community structure Habitat value of rehabilitation plantings Bird succession in post-mining rehabilitation Birds in post-mining rehabilitation at Weipa Limitations of previous research Research design Definitions and inconsistencies in terminology Conclusions and relevance to this study Chapter 5: Research design and methods Introduction Assumptions of the research design Research design Grain and extent of the study Selection of native forest reference sites Selection of mine rehabilitation sites Sampling procedures Bird assemblages Vegetation composition and structure Landscape functionality Data analysis Analysis of bird data Analysis of landscape function and vegetation data Conclusion Chapter 6: Bird results Introduction... 03

8 viii 6.. Bird survey completeness Bird assemblages of the pre-mining native forest Comparison of bird assemblages Mean bird abundance Species richness Composition of bird assemblages Site species shortfall index Comparison by individual species Comparison by foraging group Temporal patterns in bird succession Temporal patterns in individual bird species Temporal patterns in bird community structure Bird species associated with mature framework trees Bird response to vegetation and landscape Bird response to vegetation structure Bird response to vegetation composition Foraging group response to vegetation structure Bird response to landscape variables Effect of proximity to remnant native vegetation Native forest bird response to variation in vegetation structure Highly correlated variables Summary of bird results Chapter 7: Vegetation and landscape functionality results Introduction Reference vegetation and landscape functionality values Pre-mining native forest - Land unit 2B Post-mining landscape analogue - Land unit 2C Post-mining landscape analogue - Land unit 5K Summary of native forest reference values Vegetation and landscape functionality in mine rehabilitation sites Mine rehabilitation - to 2 years old Mine rehabilitation - 3 to 4 years old Mine rehabilitation - 5 to 8 years old Mine rehabilitation - 9 to 6 years old Mine rehabilitation - > 6 years old Comparison of landscape functionality Stability index values Infiltration index values Nutrient cycling index values... 74

9 ix 7.5 Comparison of vegetation composition Plant species composition Density of framework species Density of destabilising species Comparison of vegetation structure Vegetation height Canopy volume Basal area Vegetation layering Confounding of variables Summary of vegetation and landscape function results Chapter 8: Discussion of results Introduction Temporal patterns in bird assemblages Bird species richness and abundance Rate of change Succession and 'specialisation' Bird community structure Bird assemblages, vegetation and landscape Composition of bird assemblages Habitat-specific bird assemblages The landscape context of rehabilitation The trajectory of post-mining rehabilitation Development of landscape functionality Vegetation development Is mine rehabilitation likely to restore habitat similar to that of pre-mining native forest? Conservation implications of habitat conversion for the native forest birds Chapter 9: Conclusions Introduction Birds as indicators of rehabilitation success Findings in relation to key research questions How do the bird assemblages of post-mining rehabilitation compare with native forest bird assemblages? What is the temporal pattern of bird succession in post-mining rehabilitation on the Weipa bauxite plateau? What environmental factors determine bird occupancy in pre-mining native forest and post-mining rehabilitation on the Weipa bauxite plateau? Is the post-mining rehabilitation process likely to restore habitat similar to that of pre-mining native forest on the Weipa bauxite plateau?

10 x 9.3 Overall conclusions Implications of findings Implications for rehabilitation practices Biodiversity management implications for the Weipa bauxite plateau Broad policy implications Limitations of this study and directions for future research References Appendix Land units of the Weipa region Appendix 2 Birds of the eucalypt forests and woodlands of Cape York Peninsula Appendix 3 Conservation significance of the birds of CYP eucalypt forest Appendix 4 Mine rehabilitation seed mix used at Weipa Appendix 5 Variables that may affect post-mining rehabilitation plots Appendix 6 Native forest reference sites Appendix 7 Mine rehabilitation sites Appendix 8 Soil surface assessment indicators Appendix 9 Bird response variables Appendix 0 Landscape context and vegetation variables Appendix Unrecorded species Appendix 2 Bird summary values by site Appendix 3 Bird Abundance ANOVA Results Appendix 4 Pairwise differences in bird species composition Appendix 5 Foraging groups Appendix 6 CCA Site bird detection rates and vegetation structure Appendix 7 Correlations between vegetation structure and composition Appendix 8 CCA Site bird detection rates and vegetation composition Appendix 9 CCA of foraging group abundance and vegetation structure Appendix 20 CCA site bird detection rates and landscape variables... 35

11 xi Appendix 2 CCA Site bird detection rates and vegetation structure (mine rehabilitation sites only) Appendix 22 CCA Site bird detection rates and vegetation structure (native forest sites only) Appendix 23 Mean canopy volume of woody species m /ha by land unit 3 Appendix 24 Mean canopy volume of grasses (m /ha) by land unit List of Figures Figure 2. Location of study area... 6 Figure 2.2 Exposure of the bauxite profile... 8 Figure 2.3 Cyclone tracks near Weipa over the last 50 years...0 Figure 2.4 Monthly mean rainfall at Weipa... Figure 2.5 Monthly mean temperature at Weipa... Figure 2.6 Distribution of Eucalyptus tetrodonta open forest (R.E )... 4 Figure 2.7 Fire scar map of the Weipa locality... 6 Figure 2.8 Distribution of bauxite mine leases and minerals exploration leases... 2 Figure 2.9 Mining sequence Figure 2.0 Undulations in the post-mining land surface Figure 4. Factors influencing the effectiveness of rehabilitation as habitat Figure 5. Location of study sites Figure 5.2 Variation in pre-mining native forest reference sites Figure 5.3 Variation in pre-mining native forest reference sites Figure 5.4 Variation in pre-mining native forest reference sites Figure 5.5 Post-mining landscape analogue land units Figure 5.6 Vegetation sampling Figure 5.7 Patches and inter-patches created by bank-trough structures Figure 6. Individual- and sample-based rarefaction curves Figure 6.2 Increasing bird abundance with age of site... 0 Figure 6.3 Increasing species richness with age of site... Figure 6.4 Decreasing species shortfall index with age of site... 3 Figure 6.5 Profile graph of percentage of sites in which shared species were recorded... 9 Figure 6.6 Profile graph of mean detection rates of shared species Figure 6.7 Mean foraging group abundance by age class and land unit Figure 6.8 Ordination graph of birds and vegetation structure (sites) Figure 6.9 Ordination graph of birds and vegetation structure (species)... 3 Figure 6.0 Ordination graph of birds and vegetation composition (sites) Figure 6. Ordination graph of birds and vegetation composition (species) Figure 6.2 Ordination graph of foraging groups and vegetation structure (sites) Figure 6.3 Ordination graph of foraging groups and vegetation structure (foraging groups) Figure 6.4 Ordination graph of birds and landscape context (sites)... 4 Figure 6.5 Ordination graph of birds and landscape context (species) Figure 6.6 Detection rate of banded honeyeater in native forest sites Figure 6.7 Detection rate of grey shrike-thrush in native forest sites Figure 6.8 Detection rate of bar-shouldered dove in native forest sites Figure 7. Pre-mining native forest land unit 2B Figure 7.2 Post-mining landscape analogue land units...55 Figure 7.3 Causes of spatial variation in native forest reference sites Figure 7.4 Young mine rehabilitation sites Figure 7.5 Mine rehabilitation sites Figure 7.6 Mine rehabilitation sites Figure 7.7 Regression of stability index values against age of site... 7 Figure 7.8 Regression of infiltration index values against age of site Figure 7.9 Regression of nutrient cycling index values against age of site... 75

12 xii Figure 7.0 Dendrogram of sites based on relativised canopy volume of woody species Figure 7. Canopy volume of framework, desirable and other species Figure 7.2 Contributions of main plant genera to canopy volume... 8 Figure 7.3 Regression of vegetation height against age of site Figure 7.4 Regression of canopy volume against age of site Figure 7.5 Regression of basal area against age of site Figure 7.6 Percent foliage cover of vegetation layers by land unit List of Tables Table 2. Description of the bauxite profile... 7 Table 3. Landscape attributes related to bird distributions... 3 Table 3.2 Partitioning of bird assemblages in temperate eucalypt forests Table 3.3 Eucalypt structural attributes related to bird distributions...38 Table 3.4 Hierarchical scale effects on species richness Table 4. Expression of disturbance at various levels of ecological organisation Table 5. Study overview Table 5.2 Stratification of native forest sites Table 5.3 A comparison of conditions at Gove and Weipa Table 6. Breakdown of the number of bird species recorded Table 6.2 Indicator species in native forest and mine rehabilitation Table 6.3 Community level bird statisitics by age class and land unit Table 6.4 ANOVA results for bird abundance data...09 Table 6.5 MRPP results (bird species composition)... 2 Table 6.6 Bird species recorded exclusively in native forest reference sites... 5 Table 6.7 Bird species recorded exclusively in mine rehabilitation sites... 6 Table 6.8 Bird species recorded in both native forest reference sites and mine rehabilitation sites... 7 Table 6.9 Mean abundances of foraging groups... 2 Table 6.0 Species detection rates by age class Table 6. Number and richness of foraging groups by age class Table 6.2 Mean detection rates of species associated with big framework trees Table 6.3 Correlation coefficients between vegetation structure variables and ordination axes Table 6.4 Correlation coefficients between vegetation composition variables and ordination axes Table 6.5 Correlation coefficients between vegetation structure variables and ordination axes Table 6.6 Correlation coefficients between landscape variables and ordination axes Table 6.7 Linear regressions between bird summary values and landscape variables Table 6.8 Comparison of regressions of highly correlated variables Table 7. Individual vegetation layer statistics by land unit Table 7.2 Total values for woody vegetation by age class and land unit... 6 Table 7.3 Landscape function indices by age class and land unit Table 7.4 Summary statistics for framework and desirable species Table 7.5 ANOVA results for LFA indices in mine rehabilitation Table 7.6 MRPP results (plant species composition) Table 7.7 MRPP results (density of framework species) Table 7.8 ANOVA results for summary vegetation values Table 8. Native forest birds absent from or significantly less abundant in mine rehabilitation... 29