Routine Turtle and Dugong Monitoring Program Report Dredging Report 3

Transcription

1 Routine Turtle and Dugong Monitoring Program Report Dredging Report 3 Ichthys Nearshore Environmental Monitoring Program L384-AW-REP Prepared for INPEX November 2013

2 Document Information Prepared for INPEX Project Name File Reference L384-AW-REP-10247_0_Turtle and Dugong Monitoring Dredging Report 3.docm Job Reference L384-AW-REP Date November 2013 Contact Information Cardno (NSW/ACT) Pty Ltd Cardno (WA) Pty Ltd Cardno (NT) Pty Ltd Level 9, The Forum 11 Harvest Terrace Level 6, 93 Mitchell Street 203 Pacific Highway West Perth WA 6005 Darwin NT 0800 St Leonards NSW 2065 Telephone: Telephone: Telephone: Facsimile: Facsimile: Facsimile: International: International: International: Document Control Version Date Author Author Initials Reviewer Reviewer Initials A 10/10/2013 Michelle Blewitt MLB Will Macbeth WM Megan Rice MR Lachlan Barnes LB B 31/10/2013 Michelle Blewitt MLB Will Macbeth Craig Blount Joanna Lamb WM CB JL 0 13/11/2013 Michelle Blewitt MLB Joanna Lamb JL Megan Rice MR This document is produced by Cardno solely for the benefit and use by the client in accordance with the terms of the engagement for the performance of the Services. Cardno does not and shall not assume any responsibility or liability whatsoever to any third party arising out of any use or reliance by any third party on the content of this document. Prepared for INPEX Cardno Page ii

3 Executive Summary The Turtle and Dugong Monitoring Program (TDMP) has been developed to monitor the abundance and distribution of turtles and dugongs in waters around Darwin during dredging and spoil disposal activities associated with the Ichthys Gas Field Development Project (the Project). The TDMP incorporates aerial and land-based survey techniques to estimate and monitor turtle and dugong populations in Darwin Harbour and surrounding inshore waters. Finer-scale aspects of abundance and distribution of dugongs and turtles within Darwin Harbour in relation to potential habitat use are also assessed. This report outlines the findings of the second during dredging survey (D3), undertaken between 27 July 2013 and 11 August 2013 and approximately eight months after the commencement of Cutter Suction Dredger (CSD) operations on 4 November It should be noted that the third and final Baseline survey (B3), completed in October 2012 and initially designated as Dredging survey 1 (D1), was undertaken after the commencement of the Backhoe Dredger (BHD) operations, but prior to the start of CSD operations. During B3, sediment plumes from BHD operations were minimal and localised to a very small area and, as such, it was considered part of the Baseline Phase despite its initial designation. The first during dredging survey, which was designated as Dredging survey 2 (D2), was undertaken in May 2013, while the second during dredging survey which is outlined in this report was designated as Dredging survey 3 (D3). Data collected via standardised aerial transects during D3 has enabled a comparison of turtle and dugong sightings, population estimates and density at three locations (or blocks) Bynoe Harbour (Control location 1 - C1), Darwin Harbour/Hope Inlet region (Impact location - IM), and the Vernon Islands and Melville Island (Control location 2 - C2) with data collected prior to the initiation of CSD activities (Baseline Phase). During D3, each block was further sub-divided into sections relating to habitat and bathymetry. The addition of land observations during the Dredging Phase has enabled examination of alternate observational methods to identify finer-scale turtle and dugong habitat associations at two locations within Darwin Harbour. During D3, an average of 1,322 km of linear transect was flown during aerial surveys, equating to an average survey area of approximately 547 km 2 over an approximate 23 hour flying period. The overall monitoring area (i.e. area within the total transect boundaries) during D3 equated to 2,850.5 km 2 across all three blocks. Fewer dugongs were sighted during D3 (71 individuals), compared to the average observed per survey throughout the Baseline Phase (88 individuals). Based on raw sighting data, the mean number of dugongs (± SE) sighted per flight within each block during D3 (C1 = 6.3 ± 1.2, IM = 13.0 ± 2.1, C2 = 4.3 ± 0.3; n = 3 flights) were lower than the corresponding averages for the Baseline Phase (C1 = 7.0, IM = 16.6, C2 = 9.3), particularly in the case of C2. Similarly, the means for D3 were also lower than those recorded during B2 at the same time period last year (C1 = 7.0 ± 4.5, IM = 14.3 ± 8.3, C2 = 9.3 ± 5.8). Dugong population estimates (± SE) during D3, derived from the Marsh and Sinclair (1989) and Pollock et al. (2006) methods, were 134 ± 58 and 48 ± 12 in C1, respectively, and 316 ± 121 and 167 ± 39 in IM, respectively. No population estimates could be calculated for C2 as there were too few dugong sightings per replicate flight. These D3 estimates were generally lower than the average of survey estimates recorded during the Baseline Phase for C1 (190 ± 103 SE and 167 ± 168 SE) and IM (359 ± 154 SE and 231 ± 138 SE). The relatively high variability among replicate population estimates for all surveys may have been as a result of the inherent behaviour of dugongs (i.e. highly mobile and constantly submerging), migration or seagrass seasonality, as well as variable environmental conditions experienced during aerial surveys, such as fluctuating sea states and high turbidity. During D3, no dugongs were observed in Darwin Harbour Inner, while at the same time of year during the Baseline Phase surveys (i.e. B2) some (n=18) dugongs were sighted around Blaydin Point, Channel Island and other areas around the inner harbour. The greatest density of dugong sightings during D3 in IM was again within the Shoal Bay and Lee Point areas, similar to other surveys. More recent seagrass mapping completed within Shoal Bay has recorded Halophila decipiens, which is a preferred diet of dugongs, in the area. Given that an observed relationship between the distribution of dugongs and H. decipiens was also noted during the Baseline Phase and D2, the presence, distribution, density and species composition of seagrass habitats is significant. Five hundred and twenty nine turtles were sighted in D3 in comparison to an average of 598 turtles recorded during the Baseline Phase. The mean numbers of turtle sightings per replicate flight (± SE) within survey Prepared for INPEX Cardno Page iii

4 blocks during D3 (C1 = 58 ± 0.3, IM = 54 ± 2.0, C2 = 64 ± 10.4) were lower than the means of those recorded across Baseline Phase surveys (C1 = 65 ± 12.3, IM = 86 ± 14.4, C2 = 87 ± 13.6) and in the case of IM, lower than that reported at the same time last year during the second Baseline Phase survey (B2) (69 ± 8.5). In contrast, mean numbers of turtles were similar between D3 and B2 for blocks C1 (62 ± 20.7 in B2) and C2 (64 ± 7.6 in B2). The density of sighted turtles during D3 was at least 0.2 turtles per km 2 higher or lower than the mean of Baseline Phase densities in 24% and 61% of IM grid cells, respectively. A relatively small change (between and 0.2 turtles per km 2 ) was observed for 14% of IM grid cells. No turtles were sighted in approximately 2% of IM during D3, the Baseline Phase, or both. In comparison with B2 (July 2012), D3 densities were 0.2 turtles per km 2 higher in 12% of IM grid cells and lower in 23.3% of IM grid cells, with little change or no turtles recorded for the remaining cells. Similarly to dugongs, while a number of turtles were sighted within the Upper East Arm of Darwin Harbour during the Baseline Phase surveys (B2), there was a relative decrease in turtle sightings in that area during D3. Similar decreases were also evident for other areas outside the inner harbour, such as around Shoal Bay. Temporal and spatial variation in dugong and turtle distribution and abundance may be a result of movement in and out of specific areas, possibly due to avoidance behaviour and/or the pursuit of more optimal foraging areas. Further surveys have the potential to provide greater insight into the ongoing use of these areas by turtles within Darwin Harbour and surrounding regions. Aerial survey blocks were further partitioned into non-overlapping sub-blocks comprised of relatively homogeneous characteristics (i.e. habitat type and depth profile), with the aim of reducing imprecision associated with block population estimates. Variances associated with dugong block population estimates derived via sub-blocks were similar to those derived from whole block estimates regardless of the population estimation method used. Conversely, the non-uniform distribution and abundance of turtles within blocks facilitated an observed reduction in variance associated with estimated population sizes based on sub-blocks compared to whole block population estimates. If the distribution of turtles remains non-uniform during future surveys, combining sub-block population estimates may prove beneficial in improving precision associated with turtle block population estimates. Land observations during D3 were completed at Channel Island and Cullen Bay, which have been previously recognised as hotspots for marine megafaunal sightings. At Channel Island, 280 turtles were sighted, with 95% of these turtles confirmed as green turtles and mostly juveniles. No dolphins or dugongs were observed from Channel Island Bridge during D3. It is recognised that Channel Island provides important habitat and foraging grounds for turtles, and possibly also for dugongs and inshore delphinids despite the lack of sightings. Reef and foraging grounds extend below the observation platform (bridge) bringing turtles closer to observers and increasing the potential for observers to detect and identify turtle species. In contrast, at Cullen Bay no turtles were sighted during D3, while one dugong was sighted off the rock wall. Dugongs recorded at Cullen Bay during the Dredging Phase surveys were sighted close to a highconfidence seagrass habitat (Geo Oceans 2013a), and were displaying long dive times suggesting possible foraging behaviour. Since the commencement of the Project, ten turtles have been taken to the Arafura Timor Research Facility or Department of Land Resource Management (DLRM) for rehabilitation. No satellite tagging of these turtles was undertaken by Cardno due to their unsuitability. Reasons included the small number of targeted species rehabilitated to date and the nature of injuries preventing effective tagging. Prepared for INPEX Cardno Page iv

5 Glossary Term or Acronym Availability bias ACF B1 B2 B3 BACI BHD BSS C1 C2 CI CL CSD CV D1 D2 D3 Double platform Definition A bias in the data affected by an animal being concealed by an environmental factor so that is not visible to the observer Availability Correction Factor Baseline survey 1: 1 June 2012 to 8 June 2012 (2 replicate flights) Baseline survey 2: 18 July 2012 to 5 August 2012 (3 replicate flights) Baseline survey 3: 30 September 2012 to 21 October 2012 (3 replicate flights) Before, After, Control, Impact Backhoe Dredger Beaufort Sea State is an empirical measure (0 to 12) for the intensity of the wind based mainly on sea-state or wave conditions with 0 refering to calm, mirror like seas and 12 refers to hurricane conditions Control block 1 (Bynoe Harbour) Control block 2 (Vernon Islands to Melville Island) Confidence Interval Condifence Level Cutter Suction Dredger Coefficient of Variation is used to compare the standard deviations between populations with different means and it provides a measure of variation that is independent of the measurement units Completed during Bachhoe Dredger (BHD) operations only where impacts are considered minimal and thus D1 survey is referred to as Baseline survey 3 (B3) Dredging survey 2: 11 May 2013 to 26 May 2013 (3 replicate flights) Dredging survey 3: 27 July 2013 to 11 August 2013 (2-3 replicate flights) A sampling method using two observers to sample a given species in the same area at the same time. The two observers must be independent of each other and isolated visually and acoustically. On aerial surveys, the two observers are normally arranged as a front and rear observer Prepared for INPEX Cardno Page v

6 Term or Acronym DLPE DLRM DSDMP EBPC Act GEP GIS Group size h HSE IM IPDEP km/h kn LAT MSL min MNES n NEMP NT NTC Paj PB PCF PERMANOVA Perception bias Definition Department of Lands, Planning and the Environment Department of Land Resource Management Dredging and Spoil Disposal Management Plan East Arm The Commonwealth Environment Protection and Biodiversity Conservation Act (1999) Gas export pipeline Geographic Information System Mean group size of a group of dugongs or turtles observed during aerial surveys Hour/s Health Safety Environment Impact block (Darwin Harbour/Hope Inlet to Gunn Point) Ichthys Project Dredging Expert Panel Kilometres per hour Knots: Nautical miles per hour Lowest Astronomical Tide Mean Sea Level Minute/s Matter of National Environmental Significance Sample size Ichthys Project Nearshore Environmental Monitoring Plan Northern Territory National Tidal Centre Probability of availability which is used for population estimates Port observers - both front and rear Perception Correction Factor Permutational Analysis of Variance A bias in the data caused by an observer not seeing an animal despite it being visible Prepared for INPEX Cardno Page vi

7 Term or Acronym PF Population estimate p-perm PR QA/QC Recapture SB SE SF SR TDMP Turbidity Definition Port observer - front only An estimate of the number of individual animals living in one place at the one time Equivalent to the traditional p-value but derived from a permutational analyses (PERMANOVA) Port observer - rear only Quality Assurance/Quality Control The event of one individual or group of animals being counted (and thus recorded) a subsequent time, usually by the rear observer during aerial surveys Starboard observers - both front and back observers Standard error of the mean Starboard observer - front only Starboard observer - rear only Turtle and Dugong Monitoring Program Turbidity gives an indication of water clarity Prepared for INPEX Cardno Page vii

8 Table of Contents Executive Summary iii Glossary 1 Introduction Overview Background Aims and Objectives 1 2 Methodology Monitoring of Rehabilitated Turtles Field Monitoring Schedule Field Sampling Locations Aerial Surveys Land Surveys Field Methods Aerial Surveys Land Surveys Data Analysis Aerial Surveys Land Surveys Assumptions and Constraints Aerial Surveys Land Surveys Quality Assurance and Quality Control 11 3 Results Aerial Surveys Survey Effort Dugongs Turtles Land Surveys Survey Effort Channel Island Bridge Cullen Bay Rock Wall Rehabilitated Turtle Update Quality Assurance and Quality Control 52 4 Discussion Aerial Surveys Dugongs Turtles Block vs. sub-block population and variance estimates Land Surveys Rehabilitated Turtle Update 55 5 Conclusions 56 6 Acknowledgements 57 7 References 58 v Prepared for INPEX Cardno Page viii

9 Tables Table 2-1 Monitoring schedule for the TDMP 3 Table 3-1 Table 3-2 Table 3-3 Details of group size estimates and correction factors used in dugong population estimates for D3 (July 2013) aerial surveys (based on Marsh and Sinclair (1989) method) 22 Sighting histories for dugong groups sighted during D3 aerial surveys in July 2013 by (a) individual observers for each replicate flight; and (b) all combined* 23 Comparison of dugong population estimates for each block during D3 and the Baseline Phase based on Marsh and Sinclair (1989) and Pollock et al. (2006) 23 Table 3-4 Results of univariate PERMANOVA testing for differences in the density of dugongs per km 2 based on raw sightings 27 Table 3-5 Table 3-6 Table 3-7 Table 3-8 Table 3-9 Dugong population estimates for each sub-block during D3 based on average Marsh and Sinclair (1989) and Pollock et al. (2006) methods across replicate flights for all sub blocks where sightings occurred 28 Dugong density estimates (per km 2 ) for each sub-block during D3 based on Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods 29 Details of turtle group size estimates for whole blocks and correction factors used in population estimates for the D3 aerial survey 40 Sighting histories for turtles sighted during D3 aerial surveys in July 2013 by (a) individual observers for each replicate flight and (b) all combined* 41 Comparison of turtle population estimates within each block during D3 and the Baseline Phase, based on Marsh and Sinclair (1989) and Pollock et al. (2006) methods 41 Table 3-10 Results of univariate PERMANOVA tests for differences in the density of turtles per km based on: a) the Marsh and Sinclair (1989) population estimate method; b) the Pollock et al. (2006) population estimate method; and c) raw sightings between two Phases (Baseline and Dredging), Phase (B1, B2 and B3 within Baseline Phase and D3 within Dredging Phase), two Treatments (Control and Impact) and three Blocks (C1 and C2 within Control and IM within Impact) 45 Table 3-11 Turtle population estimates for each sub-block during D3 based on average Marsh and Sinclair (1989) and Pollock et al. (2006) methods across replicate flights for each sub-block during D3 46 Table 3-12 Turtle density estimates (per km 2 ) for each sub-block during D3 based on Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods 47 Table 3-13 Survey effort for land-based observations during D3 49 Table 3-14 Number of megafaunal sightings and number of individual animals observed at Channel Island Bridge and Cullen Bay rock wall during D3 50 Figures Figure 2-1 Aerial survey blocks and sub-divisions encompassing Bynoe Harbour (C1), Darwin Harbour/Hope Inlet region (IM) and Vernon Islands (C2) 6 Figure 2-2 Land survey sites in Darwin Harbour during D3 Cullen Bay Rock Wall (left) and Channel Island Bridge (right) 7 Figure 3-1 Distribution of dugongs (based on raw sightings) during D3 aerial survey 14 Figure 3-2 Mean counts per replicate (±SE) of dugongs in each block (C1, IM and C2) during D3 aerial survey and the same time period last year (B2). The mean number of dugongs recorded in the Baseline Phase in each block is indicated by the solid lines 15 Figure 3-3 Density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids at IM during D3 16 Figure 3-4 Relative difference in the density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the Baseline Phase 17 Prepared for INPEX Cardno Page ix

10 Figure 3-5 Relative difference in the density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids at IM during D3 compared to the same time period last year (B2, July 2012) 18 Figure 3-6 Proportion of dugong sightings within different depth ranges (m) in: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase 20 Figure 3-7 Proportion of dugong sightings with respect to benthic habitat types at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase 21 Figure 3-8 Mean population estimates for dugongs in each block during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methodologies. N.B. insufficient data were available for population estimates to be made for C2 in D3 24 Figure 3-9 Mean dugong density per km 2 (±SE) in each location during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989) population estimate method; b) Pollock et al. (2006) population estimate method; and c) raw sightings 26 Figure 3-10 Mean population estimates (±SE) for dugongs during D3 derived from whole block and subblock calculations based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methods in C1 and IM (no dugong population estimates were calculated for C2 during D3) 30 Figure 3-11 Distribution of turtles (based on raw sightings) during D3 aerial survey 32 Figure 3-12 Mean counts per replicate (±SE) of turtles in each block (C1, IM and C2) during D3 and the same time period last year (B2). The mean number of turtles recorded during the Baseline Phase at each block is indicated by the solid line 33 Figure 3-13 Density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 34 Figure 3-14 Relative difference in the density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the Baseline Phase 35 Figure 3-15 Relative difference in the density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the same time period last year (B2, July 2012) 36 Figure 3-16 Proportion of turtle sightings within different depth ranges (m) at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase 38 Figure 3-17 Proportion of turtle sightings with respect to benthic habitat types at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase 39 Figure 3-18 Mean population estimates of turtles within each block during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methodologies42 Figure 3-19 Mean turtle density per km 2 (±SE) at each location during D3 and the Baseline Phase based on: a) the Marsh and Sinclair (1989) population estimate method; b) the Pollock et al. (2006) population estimate method; and c) raw sightings 44 Figure 3-20 Mean population estimates (±SE) for turtles derived from whole block and sub-block calculations based on the: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methods at each location during D3 48 Figure 3-21 Distribution of species sighted during land surveys in D3 (July 2013). N.B. Sightings of dolphins (blue dots) at Cullen Bay are a group of between 4 and 10 individuals sighted over a 2 hour period 51 Prepared for INPEX Cardno Page x

11 Appendices Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F Appendix G Appendix H Appendix I Appendix J Appendix K Appendix L Appendix M Appendix N Summary of Aerial Survey Effort During Baseline Phase and Dredging surveys Summary of Dugong Sightings across Baseline Phase and Dredging surveys Summary of Turtle and Dugong Sightings Overlaid with confirmed seagrass habitat during B2 and D3 surveys Dugong Population Size Estimates During Baseline Phase and Dredging Surveys Corrected Dugong Densities During Baseline Phase and Dredging Surveys Dugong Sub-block Population Size Estimates during D3 Aerial Surveys Corrected Dugong Sub-block Densities during D3 Aerial Surveys Summary of Turtle Sightings Across Baseline Phase and Dredging Surveys Turtle Population Size Estimates during Baseline Phase and Dredging Surveys Corrected Turtle Densities during Baseline Phase and Dredging Surveys Turtle Sub-block Population Size Estimates during D3 Aerial Surveys Corrected Sub-block Turtle Densities during D3 Aerial Surveys Land-based Observation Sightings Data Rehabilitated Turtle Summary Data Prepared for INPEX Cardno Page xi

12 1 Introduction 1.1 Overview INPEX is the operator of the Ichthys Gas Field Development Project (the Project). The Project comprises the development of offshore production facilities at the Ichthys Field in the Browse Basin, some 820 km westsouth-west of Darwin, an 889 km long subsea gas export pipeline (GEP) and an onshore processing facility and product loading jetty at Blaydin Point on Middle Arm Peninsula in Darwin Harbour. To support the nearshore infrastructure at Blaydin Point, dredging works will be carried out to extend safe shipping access from near East Arm Wharf to the new product loading facilities at Blaydin Point which will be supported by piles driven into the sediment. A trench will also be dredged to seat and protect the GEP for the Darwin Harbour portion of its total length. Dredged material will be disposed at the spoil ground located approximately 12 km north-west of Lee Point. A detailed description of the dredging and spoil disposal methodology is provided in Section 2 of the Dredging and Spoil Disposal Management Plan East Arm (DSDMP) (INPEX 2012). 1.2 Background This report outlines the findings of the second Dredging Phase survey of the Turtle and Dugong Monitoring Program (TDMP) undertaken from 27 July 2013 to 11 August 2013 (D3). It includes results from aerial and land surveys and where appropriate, comparisons were also undertaken with Baseline Phase survey 2 (B2; 18 July 2012 to 5 August 2012), which was completed at the corresponding time of year in Baseline survey 3 (B3) was undertaken after the commencement of the Backhoe Dredger (BHD) operations on 27 August 2012 and originally termed Dredging survey 1 (D1); however, given this was prior to the start of the Cutter Suction Dredger (CSD) on 4 November 2012, it was renamed B3. This was further supported by statistical analyses, which found no statistical difference between the three surveys. The first during dredging survey undertaken for the TDMP was in May 2013 and was termed D2. Results presented in this report are for aerial and land surveys undertaken in July 2013 and August 2013 and referred to as D3. The field protocol utilised in this program follows the methodology outlined in the Nearshore Environmental Monitoring Plan (NEMP) (Rev 2) (Cardno 2013a). 1.3 Aims and Objectives The key objectives of the TDMP are to: > Quantify the variability of turtle and dugong relative abundance and distribution in the inshore waters of Darwin Harbour and nearshore waters between and within surveys. Observe changes in these parameters over the duration of the Dredging and Post Dredging Phases; and > Assess fine scale aspects of dugong and turtle populations within Darwin Harbour in relation to potential habitat use and identify foraging areas (if possible). The aim of this Report is to continue the collection of robust data on turtle and dugong distribution and abundance around the monitoring area during the Dredging Phase that can then be used for temporal comparison with data collected in previous surveys done during the Baseline and Dredging Phases. Prepared for INPEX Cardno Page 1

13 2 Methodology Full details on methodology for aerial surveys and turtle tagging for the TDMP can be found in Appendix K of the NEMP (Rev 2) (Cardno 2013a) and the TDMP Baseline Report and Dredging Report 1 (Cardno 2012a, 2013b). Where methodology has been refined, or when there has been additional analyses utilised, details are provided in this report. 2.1 Monitoring of Rehabilitated Turtles Around Darwin, turtles are occasionally found stranded as a result of predatory attack, vessel strike, entanglement or more frequently, from unknown injuries. Many of these turtles are rehabilitated at either the Animal Ark Hospital or at Charles Darwin University. During the TDMP, any turtles bought in for rehabilitation were assessed for possible tagging as part of the monitoring program. Results are summarised in Section 3.3 and in Appendix N. 2.2 Field Monitoring Schedule D3 was undertaken in July 2013 and August 2013, following three Baseline Phase surveys undertaken in 2012 and the first during dredging survey (D2) in May Table 2-1 outlines the monitoring schedule for the TDMP during the Baseline and Dredging Phases to date. Prepared for INPEX Cardno Page 2

14 Table 2-1 Monitoring schedule for the TDMP Sampling Periods Baseline Phase Routine Turtle and Dugong Monitoring Program Report Dredging Report 3 Dredging Phase Field Method Site B1 B2 B3 D2 D3 Aerial surveys Darwin Harbour and surrounding waters Aerial Survey May/June 2012 Training and Safety Inductions 7 days 2 replicate flights Aerial Survey July/August 2012 Training and Safety Inductions 14 days 3 replicate flights Aerial Survey September/October 2012 Training and Safety Inductions 17 days 3 replicate flights Aerial Surveys May 2013 Training and Safety Inductions 16 days 2 to 3 replicate flights* Aerial Surveys July 2013 Training and Safety Inductions 16 days 3 replicate flights Boat surveys Darwin Harbour and surrounding waters Transects and Point- Sampling June/ July 2012 Training and Safety Inductions 6 days, 1 replicate of 15 point sampling sites and transects between sites Transects and Point- Sampling August 2012 Training and Safety Inductions 11 days, 1 replicate of 20 point-sampling sites and transects between sites, 2 nd replicate of area from Lee Point to Vernon Islands Point-Sampling October 2012 Training and Safety Inductions 10 days, 2 replicate of 20 point sampling sites Land surveys Land surveys May 2013 Channel Island Bridge Cullen Bay Rock Wall 4 days Land surveys July 2013 Channel Island Bridge Cullen Bay Rock Wall 4 days Prepared for INPEX Cardno 3

15 Sampling Periods Baseline Phase Dredging Phase Field Method Site B1 B2 B3 D2 D3 1 x satellite tagging event (hand netting) September/ October 2012 Turtle satellite tagging Channel Island Bridge days 1 x satellite tagging event (net capture via boat) November days 1 x satellite tagging event Casuarina Beach September/ October nights Rehabilitated turtle update 2 turtles^ 3 turtles^ 5 turtles^ * D2 surveys: C1 and IM blocks - 3 replicate flights; C2-2 replicate flights ^ turtles in rehabilitation that were released during this period since the end of the previous survey (not suitable for tagging) Prepared for INPEX Cardno 4

16 2.3 Field Sampling Locations Aerial Surveys Aerial surveys were used to collect data over a broad spatial scale within three geographical regions (or blocks) (Figure 2-1): > Block 1 Bynoe Harbour (Control location 1 C1); > Block 2 encompassing Darwin Harbour/Hope Inlet to Gunn Point (Impact location IM); and > Block 3 located to the east, from Gunn Point and the Vernon Islands and across to Melville Island (Control location 2 C2) Land Surveys Baseline Phase boat surveys undertaken during 2012 using transects and point sampling methods resulted in few turtle and dugong observations compared to aerial surveys. Therefore, boat-based surveying was not yielding sufficient data to meet its main objective which was the assessment of finer-scale data concerning behaviour and habitat use. As a consequence, it was proposed by the Ichthys Project Dredging Expert Panel (IPDEP) that land surveys from recognised hotspots around Darwin Harbour (discrete areas with potentially the highest number of marine fauna sightings) be introduced to replace vessel transects and point sampling. These hotspots would allow for longer observation periods and greater likelihood of encountering turtles or dugongs than vessel-based surveys, thus allowing more insight into finer-scale behaviour of turtles and dugongs. A review of historical sighting data, as well as consultation with local experts in the region, identified potential hotspots from which suitable locations for land-based observations for turtles and dugongs were selected. In order to assess finer scale aspects of turtle and dugong populations in relation to frequently utilised habitats within Darwin Harbour during D3, land surveys during D3 were undertaken from two locations around Darwin Harbour Channel Island Bridge and Cullen Bay rock wall (Figure 2-2). Prepared for INPEX Cardno 5

17 Figure 2-1 Aerial survey blocks and sub-divisions encompassing Bynoe Harbour (C1), Darwin Harbour/Hope Inlet region (IM) and Vernon Islands (C2) Prepared for INPEX Cardno 6

18 Figure 2-2 Land survey sites in Darwin Harbour during D3 Cullen Bay Rock Wall (left) and Channel Island Bridge (right) Prepared for INPEX Cardno 7

19 2.4 Field Methods Aerial Surveys Aerial surveys of marine megafauna during D3 were flown from 24 July 2013 to 11 August Three replicate sets of transects (termed replicate flights ) were completed in each of the three main blocks (C1, IM, C2) (Figure 2-1). The aerial survey methodology followed the traditional strip transect survey design technique as described by Marsh and Sinclair (1989). A Dornier 220, (a high winged turbo engine aircraft), flown at a speed of approximately 110 kn and at an altitude of 500 ft (152 m), was used to complete each transect. Transect width was 200 m on each side of the aircraft, where each observer s window was demarcated into four horizontal subsections (50 m (Low), 100 m (Medium), 150 m (High) and 200 m (Very High)). Sightings on the inside or the outside of these subsections, which refers to sightings inside or beyond the 200 m transect width either side of the plane, were also recorded. Further details of the aerial survey methods are given in the TDMP Baseline Report (Cardno 2012a). During Baseline Phase and D2 surveys, each transect followed an approximate north-east to south-west alignment, spaced equally at 2 km intervals. During D3, transects were spaced at 2 km intervals, although the orientation of transects within the C1 and C2 blocks were altered slightly where necessary to accommodate for glare. Transect orientation varied depending on the time of day surveys were undertaken, with morning transects flown in the original north-east to south-west alignment and afternoon transects flown in a north to south alignment. Each of the three replicate flights flown within the IM block during D3 were flown in a north-east to south-west alignment at 2 km intervals regardless of the time of day. Placement of transects at 2 km intervals minimised the chance of sighting the same animal twice, while maximising the aerial coverage (approximately 19%) of the survey block. Dolphins, dugongs and turtles were sighted and recorded in the aerial surveys; however, dolphins were not reported as part of the TDMP. Raw data pertaining to dolphins sighted during the TDMP aerial surveys will be supplied to the Department of Land Resource Management (DLRM) Land Surveys Land surveys were undertaken from Channel Island Bridge and the rock wall at Cullen Bay during D3. At each location, a total of six hours (from 07:30) of observations were undertaken per day for two days. Survey periods incorporated eight 30-minute observations, each separated by a 15-minute ( off-effort ) window. Surveys were undertaken when the Beaufort Sea State (BSS) was less than or equal to three to maximise observability of marine megafauna (turtles, dugongs and inshore dolphins). During each 30-minute observation period, observers continuously scanned the water (aided by polarised sunglasses) and assessed any surface disturbance with binoculars to identify the presence of any marine megafauna. When megafauna were sighted, the following data was recorded on a data sheet: > Date and time of each sighting (if more than one animal was sighted at any one time, the priority was to gather data for the first animal sighted); > Animal type: turtle, dugong, dolphin or other incidental sightings (including an identification ranking e.g. certain, probable or uncertain). Species identification was also attempted, and then ranked as certain or uncertain; > Number of individuals (where multiple individuals were involved in a sighting); > Bearing of sighting from the observer (using a hand-held compass); > Estimated distance from observer to animal; and > Behavioural activity or sighting cue. Environmental and water variables (BSS, glare and turbidity) were recorded at the start and end of each observation period, while the position of each observer was recorded using a hand-held GPS. Where Prepared for INPEX Cardno 8

20 possible, confirmed cases of multiple sightings of individuals were noted; however, this was generally not reliable for turtles or dugongs Channel Island Bridge Two observers were concurrently used per survey at Channel Island Bridge (Figure 2-2). One observer was positioned on the northern side of the bridge observing north, while the second observer was located on the southern side of the bridge observing to the south. Each observer was situated in the middle of the bridge and observed the width of the channel on their observation side Cullen Bay Rock Wall Two observers were concurrently used per survey at the Cullen Bay Rock Wall (Figure 2-2). The observers were located on the rock wall, with one observing the east to north quadrant around towards Fannie Bay, while the other is observing the west to north quadrant, which included Cullen Bay. 2.5 Data Analysis Aerial Surveys Distribution and Habitat Use For each turtle or dugong sighting, the water depth was determined by estimating the vertical difference between the tidal water level and the underwater bed strata depth. This involved combining the tidal height at the time and position of each sighting (relative to mean sea level (MSL) and the bed strata depth at that position (also relative to MSL). The tidal level at the time of each sighting was obtained through interrogation of the predicted tides for the Darwin area, which were provided by the National Tidal Centre (NTC). Spatial variation among blocks in the heights and timing associated with the tidal cycle was assumed to be negligible, although it is acknowledged that such variations exist. The bed strata depths were determined through interrogation of two different bathymetric data sets. The primary source of bathymetric data was via nautical charts provided by the Australian Hydrographical Service (AUS Charts 24-28). For sightings that occurred outside the boundaries of these charts, depths were sourced from the Geoscience Australia Bathymetry and Topography Grid (Whiteway 2009), which has a 250 m grid cell resolution Dugong and Turtle Population Estimates and Densities Dugong and turtle population estimates were calculated separately for each survey block (C1, IM and C2) and were based on the estimated number of animals (turtles or dugongs) for each tandem team per transect, corrected by the appropriate correction factors, mean group size sighted by the observers and density. Population estimates were calculated using methods described in Marsh and Sinclair (1989) and Pollock et al. (2006). Both methods attempt to correct for availability bias (animals not available to observers because of water turbidity) by applying Availability Correction Factors (ACFs) and perception bias (animals visible in the survey transect but missed by observers) by applying Perception Correction Factors (PCFs). However, spatial heterogeneity in sighting conditions among observations were only addressed by the ACF associated with the Pollock et al. (2006) method, while the Marsh and Sinclair (1989) ACF method averages conditions within each replicate flight for a spatial block (or sub-block) and only corrects for differences in availability bias between spatial or temporal partitions (i.e. blocks or surveys). Blocks were then further partitioned into sub-blocks (C1, 4 sub-blocks; IM, 6 sub-blocks; C2, 4 sub-blocks), each of which comprised areas of similar bottom type and bathymetry, such as intertidal, inshore or offshore (Figure 2-1). Whilst population estimates can be calculated based on any number of sightings (including n = 1), it was recognised that the variance associated with estimates calculated using such data may be large, reducing their reliability. Therefore, only replicate flights in which five or more sightings were recorded were used for sub-block population estimation, which has been consistent across all flights throughout the Baseline and Dredging Phase aerial surveys. Where possible, dugong and turtle population estimates were calculated for each block (n = 3) and each sub-block (n = 14) using the Marsh and Sinclair (1989) and Pollock et al. (2006) methods. Prepared for INPEX Cardno 9

21 For calculation of PCF values using the Marsh and Sinclair (1989) method and for modelling of perception bias in MARK (a program that provides parameters used to estimate population size in the Pollock et al. (2006) method), data were pooled by block instead of by replicate flight. The latter was done in previous surveys. D3 aerial transect data were pooled within blocks for D3 calculations as, unlike during previous surveys, the observer team was consistent across all replicate flights and the order of transects flown was haphazardly determined (usually based on weather considerations) rather than in a strict, structured order. Furthermore, the orientation of the flights in C1 and C2 varied depending on the time of day (to minimise the negative influence of glare) and similar environmental conditions were experienced within each block and among replicate flights. As with previous analyses conducted within other surveys, group size and corresponding coefficients of variation (CV) values were determined for each replicate flight. Dugong and turtle density estimates were calculated by dividing the derived population estimates or raw sightings data by the total transect area sampled over water within each of the three blocks, 14 sub-blocks and 43 individual 6 km x 6 km grid cells overlayed within the IM block Statistical Analysis Population density estimates (number per km 2 ) for dugongs and turtles were derived, where appropriate, for replicate flights undertaken in blocks and sub-blocks using three methods: > Marsh and Sinclair (1989) population estimates; > Pollock et al. (2006) population estimates; and > Raw sightings data. Datasets derived via these three methods were statistically analysed separately. Density estimates, derived from raw sightings in D3, were calculated for only replicate flights (within each survey block) with five or more sightings per flight. Spatial and temporal differences in dugong and turtle densities were investigated via permutational univariate analyses of variance (PERMANOVA; Anderson 2001), using Euclidean dissimilarity resemblance matrices with unrestricted data permutation methodology. The four factors used in the analysis of each density dataset were: > Phase (fixed, orthogonal): 2 levels (Baseline Phase and Dredging Phase, equivalent to the Before After comparison in BACI); > Survey (random, nested within Phase): B1, B2 and B3 within Baseline Phase and D2 and D3 within Dredging Phase; > Treatment (fixed, orthogonal): 2 levels (Impact and Controls); and > Block (random, nested within Treatment): IM within Impact treatment and C1 and C2 within Controls treatment. Pooling of lower-level terms with the residual term was undertaken where appropriate (i.e. if the p-value of the pooled term was 0.25). Results for both the pooled and the un-pooled versions were presented where applicable. Where a term (factor or interaction) was significant at p 0.05, pair-wise comparisons were used in post-hoc analyses to identify which levels of factors or interaction strata were significantly different. Blocks in surveys for which less than two replicate flights were available for analysis (i.e. due to < 5 sightings in two or more of the three replicate flights) were not included in analyses. As very few dugongs were sighted during the D3 aerial surveys and only one population estimate could be calculated for IM and C2, statistical analyses of the density of dugongs were only undertaken using density data derived from raw sightings only Land Surveys Variability in the land surveys arising from several sources, including variability in sea state, observing conditions (e.g. sea state, glare) and temporal changes in the distribution and movements of the animals themselves, compromised the potential for formal analyses to detect any spatial differences and/or temporal changes in abundance of dugongs and turtles. Land-based data have, therefore, been presented in a tabular format with spatial representation of sightings in the form of GIS mapping. Prepared for INPEX Cardno 10

22 2.6 Assumptions and Constraints Aerial Surveys Routine Turtle and Dugong Monitoring Program Report Dredging Report 3 The average wind strength during D3 was BSS = 2.2, which was consistent with conditions and visibility throughout the Baseline Phase (average of BSS = 2; Cardno 2012a). Despite this, strong wind during peak D3 flight periods was evident. This resulted in the flights having to be flown in either early morning or late afternoon, which then increased the effect of glare due to the angle of the sun. To minimise the effect of glare, which could potentially impede the ability to sight animals, the orientation of transects changed from a north-east/south-west orientation to a more north/south orientation Land Surveys During land surveys, most individual turtles are not recognisable and it is likely that some animals are counted more than once. Therefore, the total recorded number of turtle sightings may be higher than the actual number of individuals present at each site during each sampling event. 2.7 Quality Assurance and Quality Control Sighting and environmental data were transcribed and entered into an Access database post flight by each observer and the environmental data collector. As part of the process, 10% of all entries were randomly checked by each team. If any errors were located, a further 10% of records were checked. Any errors were subsequently rechecked and amended in the database before data compilation and analysis. Once all datasets were compiled, the final QA process was undertaken on data in a master database by the survey leader, where a further 10% of the total number of transects flown were randomly checked. Land survey data were recorded on datasheets in the field and entered into an Excel spread sheet at the end of each survey period by each team of observers. All entries were checked for completeness against field datasheets and any missing cells or information was added or amended where necessary. Ten percent of edited files were then checked for errors by random selection. Prepared for INPEX Cardno 11

23 3 Results 3.1 Aerial Surveys Survey Effort During D3 (27 July 2013 to 11 August 2013), 7,262 km were flown over an approximate 57 hours, including 22:45 hours on linear transects. The cumulative total length of transects flown over water during each replicate flight (blocks combined) was between 1,296.4 km and 1,346.9 km (depending on transect orientation), equating to an area of between km 2 and km 2 (based on a transect width of 400 m). The overall area surveyed (variable due to tidal level) was 2,850.5 km 2 across all three geographical blocks (Appendix A). Approximately 19% of the available area was surveyed across each of the three blocks. Conditions during replicate flights were variable (average BSS = 2.2, range 1 to 4). Turbidity, based on a standardised turbidity scale used in aerial surveys (Hodgson et al. 2011), ranged between 3 and 4 and averaged approximately 4. The orientation of transects flown over the C1 and C2 blocks were selected according to conditions on each flight day, allowing for the best possible sampling conditions (Figure 2-1). Surveys in the IM block were flown in a north-east to south-west alignment regardless of the time of day Dugongs Sightings and Distribution The distribution of dugongs within blocks during D3 is presented in Figure 3-1. Seventy-one dugongs were sighted across all three survey blocks (C1: 19, IM: 39, C2: 13). This total was slightly lower than the total for B2 (92 dugongs) which was undertaken at the same time of year (July) in 2012, and the mean of 88 dugongs per survey recorded during the Baseline Phase (B1, n = 58; B2, n = 92; B3, n = 113), (Appendix B). It was however, considerably higher than the total number of dugongs recorded during D2 (May 2013, n = 33 dugongs). Appendix C shows the distribution of dugongs in Darwin Harbour (IM), overlaid with confirmed seagrass habitat for D3 and the same time in 2012 (B2; seagrass data: Geo Oceans 2013a). In B2, dugongs were sighted sporadically throughout the inner harbour region; however, no dugongs were sighted in these areas during D3. Based on raw sighting data, the mean number of dugongs (± SE) sighted per flight within each block during D3 (C1 = 6.3 ± 1.2, IM = 13.0 ± 2.1, C2 = 4.3 ± 0.3; n = 3 flights) were lower than the corresponding averages (across surveys) for the Baseline Phase (C1 = 7.0, IM = 16.6, C2 = 9.3), particularly in the case of C2. Similarly, the means for D3 were also lower than those recorded during B2, but within the associated error for that survey (C1 = 7.0 ± 4.5, IM = 14.3 ± 8.3, C2 = 9.3 ± 5.8) (Figure 3-2). The highest densities of sighted dugongs recorded during D3 (based on raw sightings within designated 6 km x 6 km grids) were within Hope Inlet and around Lee Point (grid cells 20, 22 and 23, Figure 3-3), which was similar to spatial patterns in density reported during the Baseline Phase, including the same time of the year (B2). These grid cells were within the vicinity of recently mapped seagrass for this area (Figure 3-3) (Geo Oceans 2013a, b). The relative density of sighted dugongs during D3 increased by more than 0.2 dugongs per km 2 from the mean of Baseline Phase relative densities in 4.7% of IM grid cells (grid cells 22 and 23), and decreased by more than 0.2 dugongs per km 2 from mean Baseline Phase densities in 27.9% of IM grid cells (Figure 3-4). A relatively small change (between -0.2 and 0.2 dugongs per km 2 ) was observed in the relative density of dugong sightings in 26% of IM grid cells during D3 compared to the mean of the Baseline Phase. No dugongs were sighted in 42% of IM grid cells during either D3 or the Baseline Phase (Figure 3-4). Similarly, 7% of IM grid cells exhibited an increase in the relative density of dugong sightings by more than 0.2 dugongs per km 2 during D3 compared to the same time during the Baseline Phase (B2, July 2012), while the relative density of dugong sightings decreased by more than 0.2 dugongs per km 2 in 23.3% of IM grid cells between the same time period. While no dugongs were observed in the Upper East Arm of Darwin Harbour during D3, some dugongs were sighted around Blaydin Point and Channel Island during sampling undertaken at the corresponding time of the year during the Baseline Phase (i.e. B2). This disparity resulted in the decrease in density in grid cells 39 and 40 shown in Figure 3-5. However, the dugong density estimates were based on very small numbers of Prepared for INPEX Cardno 12

24 sightings (if any) within relatively small areas (grid cell 39: B2, n = 4 and D3, n = 0 dugongs; grid cell 40: B2, n = 3 and D3, n = 0 dugongs). An apparent decrease was also evident for grid cell 34, located around the west arm area and away from the vicinity of the dredging activities. Dugongs sighted during D3 were predominantly adults, which has also been the case for all previous surveys. One cow/calf pair was sighted in C1, with another pair sighted in IM, while a further two pairs were recorded in C2. Similarly, during the Baseline Phase, calves were recorded in each block with more calves observed around IM. Prepared for INPEX Cardno 13

25 Figure 3-1 Distribution of dugongs (based on raw sightings) during D3 aerial survey Prepared for INPEX Cardno 14

26 Oct Mean dugong count SE B2 D3 0 C1 IM C2 Location Figure 3-2 Mean counts per replicate (±SE) of dugongs in each block (C1, IM and C2) during D3 aerial survey and the same time period last year (B2). The mean number of dugongs recorded in the Baseline Phase in each block is indicated by the solid lines Prepared for INPEX Cardno 15

27 Figure 3-3 Density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids at IM during D3 Prepared for INPEX Cardno 16

28 Figure 3-4 Relative difference in the density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the Baseline Phase Prepared for INPEX Cardno 17

29 Figure 3-5 Relative difference in the density of dugong sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids at IM during D3 compared to the same time period last year (B2, July 2012) Prepared for INPEX Cardno 18

30 Dugong Bathymetric Distribution and Habitat Associations During D3, dugongs were sighted in waters up to a depth of 35 m. Forty-five per cent of dugong sightings occurred in waters < 5 m deep, 35% in waters 6 m to 10 m deep, 15% occurring in water 11 m to 20 m deep and only 5% in waters > 20 m deep. Compared to the Baseline Phase, a lower proportion of dugong sightings were recorded within the 0 m to 5 m depth range during D3 for the C1 and IM blocks (Figure 3-6). In contrast, a relatively higher proportion of sightings were recorded within the 6 m to 10 m depth range for those blocks, while fewer than 20% of sightings were recorded in depths > 10 m. In C2, dugongs were most frequently sighted in the 11 m to 20 m depth range during D3 (54% of sightings). The proportion of sightings recorded in the < 5 m, 11 m to 20 m and > 30 m depth categories were slightly greater during D3 compared to the Baseline Phase in C2, while no sightings were recorded during D3 in the cases of the 6 m to 10 m or the 21 m to 30 m depth ranges (Figure 3-6). In C1, the vast majority (95%) of dugongs were sighted over gravel substrata during D3, with the remainder sighted over mud (Figure 3-7). During the Baseline Phase, C1 dugongs were most frequently observed over gravel, sand and mangroves, while none were sighted over mud. In IM, a higher proportion of sightings were recorded over mud during D3 compared to the Baseline Phase (Figure 3-7). In contrast, proportionally fewer IM dugongs were sighted over reef, sand, seagrass and gravel during D3 compared to the Baseline Phase. In C2, relatively higher proportions of sightings were recorded over reef and unsurveyed substrata during D3 compared to the Baseline Phase, while fewer sightings were recorded over gravel (Figure 3-7). While there were sightings recorded over sand and mud in C2 during the Baseline Phase, no dugongs were recorded over these substrata during D3. Prepared for INPEX Cardno 19

31 1 0.9 a) C1 Baseline Phase D Proportion of dugong sightings b) IM Baseline Phase D c) C2 Baseline Phase 0.5 D > 30 Depth range (m) Figure 3-6 Proportion of dugong sightings within different depth ranges (m) in: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase Prepared for INPEX Cardno 20

32 1 a) C1 Baseline Phase 0.9 D Proportion of dugong sightings b) IM Baseline Phase D c) C2 Baseline Phase 0.4 D Habitat Type Figure 3-7 Proportion of dugong sightings with respect to benthic habitat types at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase Prepared for INPEX Cardno 21

33 Estimates of Dugong Population Size Whole Blocks Mean group sizes and correction factors (PCF and ACF) used for estimating the size of the dugong populations (using the Marsh and Sinclair (1989) method) in aerial survey blocks C1, IM and C2 during D3 are presented in Table 3-1. Due to the small number of dugong sightings, PCFs and ACFs were calculated using all sightings combined across blocks, while group sizes and CV were calculated for each replicate flight. Parameter estimates in Table 3-1 were based on all sightings including those in the inside and outside zones (refer to Section 2.4.1). Table 3-1 Details of group size estimates and correction factors used in dugong population estimates for D3 (July 2013) aerial surveys (based on Marsh and Sinclair (1989) method) Survey Block Replicate Mean group size (CV) PCF (CV) ACF (CV) Total mean group size Starboard Port (0.3) C (0.3) (0.1) (0.1) D3 IM (0.1) 1.20 (0.1) 1.41 (0.17) 1.06 (0.05) 2.99 (0.17) 1.37 (0.66) (0.2) C (0.3) (0.0) During D3, five or more dugongs were sighted in four of the nine replicate flights (Reps 1, 2 and 3 in IM; Rep 2 in C1) (Table 3-2); while during B2, undertaken at the same time of year during the Baseline Phase, population estimates were calculated for again, only four of the nine replicate flights (Rep 1 and 2 in IM; Rep 3 for C1 and C2). The estimated population size (± SE) of dugongs within IM during D3 ranged among replicate flights from 226 ± 113 to 391 ± 248 using the Marsh and Sinclair (1989) method and from 139 ± 54 to 195 ± 71 using the Pollock et al. (2006) method (Table 3-3). The mean of these population estimates was 316 ± 121 and 167 ± 39 for the two methods, respectively. Dugong population estimates for C1 during D3 were 134 ± 58 and 48 ± 12 based on the Marsh and Sinclair (1989) and Pollock et al. (2006) methods, respectively (Table 3-3). Dugong population size estimates derived for C1 and IM during D3 were lower than the corresponding estimates during the Baseline Phase (B2), although there was considerable imprecision associated with most of those estimates as indicated by the large, overlapping SE values (Figure 3-8, Appendix D). During B2, the calculated population estimate for IM was 411 ± 139 based on the Marsh and Sinclair (1989) method and 249 ± 150 using the Pollock et al. (2006) method. Prepared for INPEX Cardno 22

34 Table 3-2 a) Sighting histories for dugong groups sighted during D3 aerial surveys in July 2013 by (a) individual observers for each replicate flight; and (b) all combined* Block Rep Number of dugong groups seen by: # dugongs sighted/ replicate Total # dugong groups sighted % dugong groups sighted by both observers SF SR SB PF PR PB Starboard Port C IM C b) Survey Number of dugong groups seen by: # dugongs sighted/ replicate Total # dugong groups sighted % dugong groups sighted by both observers SF SR SB PF PR PB Starboard Port D Starboard front (SF); starboard rear (SR); starboard both (SB); port front (PF); port rear (PR) and port both (PB) * All dugong sightings combined were used for estimates for PCF and MARK analysis in D3 survey. Table 3-3 Comparison of dugong population estimates for each block during D3 and the Baseline Phase based on Marsh and Sinclair (1989) and Pollock et al. (2006) Baseline Phase D3 Estimation Method Block Estimate SE n Estimate SE n Marsh and Sinclair C IM C Pollock et al. C IM C Prepared for INPEX Cardno 23

35 600 a) Marsh and Sinclair Baseline Phase 500 D Mean dugong population estimate + SE b) Pollock et al. Baseline Phase D C1 IM C2 Location Figure 3-8 Mean population estimates for dugongs in each block during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methodologies. N.B. insufficient data were available for population estimates to be made for C2 in D3 Prepared for INPEX Cardno 24

36 Dugong Densities Whole Block As per the population estimates (refer to Section ), the density of dugongs per km 2 was calculated for the second replicate flight at C1 and for each of the three replicate flights at IM during D3. The estimated mean density (± SE) of dugongs in C1 was 0.05 ± 0.01 per km 2 and 0.17 ± 0.07 per km 2 based on Pollock et al. (2006) and Marsh and Sinclair (1989) population estimate methods, respectively (Figure 3-9). At IM the estimated mean density (± SE) of dugongs was 0.15 ± 0.01 per km 2 and 0.28 ± 0.04 per km 2 based on Pollock et al. (2006) and Marsh and Sinclair (1989) population estimate methods, respectively, during D3 (Figure 3-9). Based on raw sightings, the mean density (± SE) of dugongs was 0.05 ± 0.01 per km 2 at C1, 0.06 ± 0.01 per km 2 in IM and 0.02 ± per km 2 in C2 during D3 (Figure 3-9). The density of dugongs per km 2 based on Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods and raw counts for each replicate flight and survey period within each of the three blocks are presented in Appendix E. Prepared for INPEX Cardno 25

37 a) Marsh and Sinclair Baseline Phase D Mean dugong density per km + SE b) Pollock et al. Baseline Phase D c) Raw sightings Baseline Phase D C1 IM C2 Location Figure 3-9 Mean dugong density per km 2 (±SE) in each location during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989) population estimate method; b) Pollock et al. (2006) population estimate method; and c) raw sightings Prepared for INPEX Cardno 26

38 No significant Phase x Treatment interaction was detected for the density of dugongs derived from raw sightings, indicating that the mean density of dugongs did not significantly differ between Impact and Control Treatments in either Phase or between Baseline and Dredging Phases for either Treatment (Table 3-4). There were also no other significant interactions between Phases, or Surveys within Phases and Treatments, or Blocks within Treatments, nor significant differences between Phases (Treatments combined), Treatments (Phases combined), Blocks (within Treatments; Phases combined), or among Surveys (within Phases; Treatments combined). The analysis shows that the density of dugongs has not significantly differed through time or between different areas for the duration of the TDMP (Table 3-4). Table 3-4 Results of univariate PERMANOVA testing for differences in the density of dugongs per km 2 based on raw sightings Analysis compares two Phases (Baseline and Dredging), two Treatments (Control and Impact), among Surveys (B1, B2 and B3 within Baseline Phase and D3 within Dredging Phase) and between Blocks (C1 and C2 within Control and IM within Impact); pooled with those for the Residual/Pooled term in the final analysis Source of Variation df MS Pseu-F p-(perm) Raw sightings Phase 1 < Treatment 1 < Phase x Treatment 1 < Survey (Phase) 3 < Block (Treatment) 1 < 0.01 < Phase x Block (Treatment) 1 < Treatment x Survey (Phase) 3 < Survey (Phase) x Block (Treatment) (3) (< 0.01) (0.65) 0.58 Residual/ Pooled 26 < 0.01 Total 40 Prepared for INPEX Cardno 27

39 Dugong Sub-block Population Estimates During D3, replicate sub-block population estimates for dugongs were calculated for all sub-blocks where dugongs were sighted within IM and C1. No sub-block estimates were done for C2 as block population estimates were not calculated for this whole block. Population estimates within IM-4 were the highest recorded among sub-blocks. Within IM, mean dugong sub-block population estimates (± SE; averaged across replicate flights) among the six sub-blocks ranged between 19 ± 18 and 215 ± 64 dugongs and between 12 ± 4 and 130 ± 34 based on Marsh and Sinclair (1989) and Pollock et al. (2006) methods respectively (Table 3-5). Within C1, mean population estimates among the four sub-blocks were only possible for sub-blocks two and three. Mean population estimates (± SE) ranged between 29 ± 27 and 12 ± 3 dugongs and between 85 ± 45 and 76 ± 19 dugongs using the Marsh and Sinclair (1989) and Pollock et al. (2006) methods respectively (Table 3-5, Appendix F). Table 3-5 Dugong population estimates for each sub-block during D3 based on average Marsh and Sinclair (1989) and Pollock et al. (2006) methods across replicate flights for all sub blocks where sightings occurred Marsh and Sinclair (1989) Pollock et al. (2006) Block Sub-block n Population Estimate SE Population Estimate SE C1 IM C n = number of replicate flights where sightings were recorded Prepared for INPEX Cardno 28

40 Dugong Sub-block Densities In D3, the highest mean density of dugongs across sub-blocks was recorded within IM-4. Estimates of mean density of dugongs (± SE) among IM sub-blocks ranged between 0.1 ± 0.1 and 1.3 ± 0.4 per km 2 based on Marsh and Sinclair (1989) population estimates and between < 0.01 ± < 0.01 and 0.8 ± 0.2 per km 2 based on Pollock et al. (2006) population estimates (Table 3-6). Mean density of dugongs (± SE) among C1 sub-blocks ranged between 0.2 ± 0.2 and 0.3 ± 0.2 per km 2 based on Marsh and Sinclair (1989) population estimates and between 0.1 ± 0.0 and 0.3 ± 0.1 per km 2 based on Pollock et al. (2006) population estimates (Table 3-6, Appendix G). As per the population estimates, no densities were calculated within any sub-blocks in C2. Table 3-6 Dugong density estimates (per km 2 ) for each sub-block during D3 based on Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods Marsh and Sinclair (1989) Pollock et al. (2006) Block Sub-block Area (km 2 ) Density Estimate SE Density Estimate SE C1 IM C < 0.01 < Dugong Whole Block vs. Sub-block Population Estimates Dugong block population estimates for D3 derived from whole block estimates and via the combination of replicate sub-block estimates were compared for each block and using the two estimation methods. Data used included all sightings (i.e. 1 sighting per replicate flight within each sub-block). Based on the Marsh and Sinclair (1989) method, the mean population size estimates (± SE) for D3 derived from whole block estimates were 134 ± 58 for C1 and 215 ± 64 for IM. Estimates derived from the combination of sub-block estimates via the Marsh and Sinclair (1989) method were 114 ± 52 for C1 and 258 ± 64 for IM (Figure 3-10). Based on the Pollock et al. (2006) method, the mean population size estimates derived from whole block estimates were 48 ± 23 for C1 and 130 ± 34 for IM, which were both slightly lower than those derived from the combination of sub-block estimates (88 ± 19 for C1 and 155 ± 31 for IM; Appendix D and F. Prepared for INPEX Cardno 29

41 a) Marsh and Sinclair Block Sub-block Mean dugong population estimate + SE b) Pollock et al. Block Sub-block C1 IM C2 Location Figure 3-10 Mean population estimates (±SE) for dugongs during D3 derived from whole block and sub-block calculations based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methods in C1 and IM (no dugong population estimates were calculated for C2 during D3) Prepared for INPEX Cardno 30

42 3.1.3 Turtles Sightings and Distribution Five hundred and twenty nine turtles were sighted across all three survey blocks during D3 (C1: 173, IM: 161, C2: 195), with observations consisting predominantly of individual turtles. Groups of two or more individuals were recorded in C2, including one group of four turtles (Appendix H). The D3 total was approximately 17% lower than the average number of turtles sighted per survey during the Baseline Phase (634 turtles), and approximately 10% lower than the number recorded during B2 (585 turtles) which was undertaken at the same time of year (July) in Turtle sightings were widely distributed within each of the three survey blocks during D3 (Figure 3-11), with this distribution generally similar to that recorded during D2 and during the Baseline Phase (Cardno 2012a, 2013b). Appendix C shows the distribution of turtles in Darwin Harbour (IM), overlaid with confirmed seagrass habitat during D3 and at the same time in 2012 (B2; seagrass data: Geo Oceans 2013a). It is evident that turtles were sighted throughout the inner harbour region during the B2 survey in July 2013; however, there were no sightings in these areas during July 2013 (D3) survey. The mean numbers of turtle sightings per replicate flight (± SE) within survey blocks during D3 (C1: 58 ± 0.3, IM: 54 ± 2.0, C2: 64 ± 10.4) were lower than the means of those recorded across the Baseline Phase (C1 = 65 ± 12.3, IM = 86 ± 14.4, C2 = 87 ± 13.6) and in the case of IM, lower than that reported during B2 (69 ± 8.5; Figure 3-12). Mean numbers of turtles were similar between D3 and B2 for blocks C1 (62 ± 20.7 in B2) and C2 (64 ± 7.6 in B2). During D3, the highest densities per km 2 of turtle sightings in the IM block were recorded around Lee Point and within Hope Inlet (grid cells 16 and 21 to 24), adjacent to Casuarina Beach (grid cell 20) and on the western edge of Darwin Harbour around Charles Point (grid cells 17 and 43) (Figure 3-13). Within Darwin Harbour Inner, the density per km 2 of turtle sightings was higher within grid cells 36, 38 and 39 compared to adjacent grid cells. The distribution of areas of high and low turtle density was similar to spatial patterns in density reported during the Baseline Phase (Cardno 2012a). The relative density of sighted turtles during D3 increased by more than 0.2 turtles per km 2 from the mean Baseline Phase relative densities in 23% of IM grid cells and decreased by more than 0.2 turtles per km 2 from mean Baseline Phase densities in 61% of IM grid cells (Figure 3-14). A relatively small change (between -0.2 and 0.2 turtles per km 2 ) was observed in the relative density of turtle sightings in 14% of IM grid cells during D3 compared to the mean of the Baseline Phase. No turtles have been sighted in 2.3% of IM grid cells during either D3 or the Baseline Phase (Figure 3-14). Similarly, 12% of IM grid cells exhibited an increase in the relative density of turtle sightings by more than 0.2 turtles per km 2 during D3 compared to B2 (July 2012), while the relative density of turtle sightings decreased by more than 0.2 turtles per km 2 in 23% of IM grid cells between the same time period (Figure 3-15). Fewer turtles were observed in the Upper East Arm during D3 compared to the corresponding time of the year during the Baseline Phase (B2). Similarly to dugongs, this resulted in a decrease in turtle density from B2 to D3 that was relatively large compared with most other grid cells inside and outside the inner harbour (Figure 3-15). However, decreases of a similar extent were also apparent for other grid cells outside the inner harbour (i.e. grid cells 7 and 27). Prepared for INPEX Cardno 31

43 Figure 3-11 Distribution of turtles (based on raw sightings) during D3 aerial survey Prepared for INPEX Cardno 32

44 B2 D3 Mean turtle count SE C1 IM C2 Location Figure 3-12 Mean counts per replicate (±SE) of turtles in each block (C1, IM and C2) during D3 and the same time period last year (B2). The mean number of turtles recorded during the Baseline Phase at each block is indicated by the solid line Prepared for INPEX Cardno 33

45 Figure 3-13 Density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 Prepared for INPEX Cardno 34

46 Figure 3-14 Relative difference in the density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the Baseline Phase Prepared for INPEX Cardno 35

47 Figure 3-15 Relative difference in the density of turtle sightings (based on raw counts) per km 2 of transect area sampled over water within 6 km x 6 km grids in IM during D3 compared to the same time period last year (B2, July 2012) Prepared for INPEX Cardno 36

48 Turtle Bathymetric Distribution and Habitat Associations Turtles were sighted across all depth ranges during D3, with most turtle sightings in C1 (64%) and IM (39%) recorded in waters < 5 m deep (Figure 3-16). In contrast, the most turtle sightings in C2 (47%) were recorded in the 11 m to 20 m depth range. There was also a small number of turtles sighted in association with deep channel habitats (> 35 m deep) near the Vernon Islands in C2 (5%). The depth distributions of turtle sightings recorded during D3 were similar to those reported during the Baseline Phase in each of the three blocks (Figure 3-16). During D3, habitat associations of turtles were generally consistent with those observed during the Baseline Phase for all three blocks (Figure 3-17). Turtles were most frequently sighted over gravel in C1 (72% of sightings), sand in IM (63% of sightings) and over unsurveyed substrata in C2 (45% of sightings). In C2, 29% of sightings were over reef, 17% over gravel and 7% over sand (Figure 3-17). Prepared for INPEX Cardno Page 37

49 a) C1 Baseline Phase D Proportion of turtle sightings b) IM Baseline Phase D c) C2 Baseline Phase D > 30 Depth range (m) Figure 3-16 Proportion of turtle sightings within different depth ranges (m) at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase Prepared for INPEX Cardno Page 38

50 a) C1 Baseline Phase D3 Proportion of turtle sightings b) IM c) C2 Baseline Phase D3 Baseline Phase D Habitat Type Figure 3-17 Proportion of turtle sightings with respect to benthic habitat types at: a) C1; b) IM; and c) C2 during D3 and the Baseline Phase Prepared for INPEX Cardno Page 39

51 Estimates of Turtle Population Size Whole Blocks Mean group sizes and correction factors used for estimating turtle population size in aerial survey blocks C1, C2 and IM during D3 are presented in Table 3-7. The PCFs (CV) and CVs calculated for turtle population size estimations ranged from 1.14 (0.08) for C1, to 1.62 (0.11) for IM for starboard side observations and from 2.09 (0.16) for C2, to 2.73 (0.24) for IM for the port side observations. The PCFs calculated for turtles were generally lower than those calculated for the Baseline Phase (particularly for the port side), mainly as a result of increased recaptures between observers (between 12% and 29%; Table 3-8b). The mean of population size estimates for turtles (± SE) in C1, IM and C2 during D3 derived using the Marsh and Sinclair (1989) method was 650 ± 85, 641 ± 91 and 565 ± 113 turtles respectively, and derived using the Pollock et al. (2006) method was 649 ± 95, 641 ± 60 and 566 ± 107 turtles respectively (Table 3-9). These population estimates were lower than the mean estimates calculated for the Baseline Phase for all three blocks (Figure 3-18, Appendix I). Table 3-7 Details of turtle group size estimates for whole blocks and correction factors used in population estimates for the D3 aerial survey These parameters were used in the Marsh and Sinclair (1989) population estimate methods only; PCF Perception Correction Factor; CV coefficient of variation; Parameter estimates were based upon all sightings in the inside and outside zones Survey Block Replicate Mean group size (CV) PCF (CV) Starboard Port (0) C (0.02) 1.14 (0.08) 2.60 (0.22) (0) D3 IM (0) 1.00 (0) 1.62 (0.11) 2.73 (0.24) (0) (0) C (0.03) 1.28 (0.05) 2.09 (0.16) (0.04) Prepared for INPEX Cardno Page 40

52 Table 3-8 a) Sighting histories for turtles sighted during D3 aerial surveys in July 2013 by (a) individual observers for each replicate flight and (b) all combined* Block Rep. Number of turtles groups seen by: Total # turtle groups sighted % turtle groups sighted by both observers SF SR SB PF PR PB Starboard Port C IM C b) Block Number of turtles groups seen by: Total # turtle groups sighted % turtle groups sighted by both observers SF SR SB PF PR PB Starboard Port C IM C Starboard front (SF); starboard rear (SR); starboard both (SB); port front (PF); port rear (PR) and port both (PB) * Replicates combined per block were used for population estimates and MARK analysis in D3 survey Table 3-9 Comparison of turtle population estimates within each block during D3 and the Baseline Phase, based on Marsh and Sinclair (1989) and Pollock et al. (2006) methods Baseline Phase Estimation Method Block Estimate SE n Estimate SE n Marsh and Sinclair C IM C Pollock et al. C IM C D3 Prepared for INPEX Cardno Page 41

53 2000 a) Marsh and Sinclair Baseline Phase D Mean turtle population estimate + SE b) Pollock et al. Baseline Phase D C1 IM C2 Location Figure 3-18 Mean population estimates of turtles within each block during D3 and the Baseline Phase based on: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methodologies Prepared for INPEX Cardno Page 42

54 Turtle Densities Whole Blocks Mean turtle density (± SE) based on the Marsh and Sinclair (1989) population estimate method was 0.79 ± 0.06 per km 2 in the C1 block, 0.56 ± 0.05 per km 2 in IM and 0.55 ± 0.04 per km 2 in C2 during D3 (Figure 3-19). Based on the Pollock et al. (2006) population estimate method, mean turtle density (± SE) was 0.8 ± 0.1 per km 2 in C1, 0.6 ± 0.03 per km 2 in IM and 0.6 ± 0.03 per km 2 at C2 during B3 (Figure 3-19). The mean density (± SE) of turtles based on raw counts was 0.4 ± per km 2 in C1, 0.3 ± 0.01 per km 2 in IM and 0.3 ± 0.06 per km 2 in C2 during B3. Mean turtle densities were lower during D3 compared to the Baseline Phase (Figure 3-19). The density of turtles per km 2 based on Marsh and Sinclair (1989), Pollock et al. (2006) and raw counts for each replicate flight and survey period within each of the three blocks are presented in Appendix J. Based on data derived using the Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods, no significant Phase x Treatment interaction was detected for the density of turtles, indicating that mean density did not significantly differ between Impact and Control Treatments in either Phase, or between Baseline and Dredging Phases for either Treatment (Table 3-10a and b). There were also no other significant interactions between Phases or Surveys within Phases and Treatments or Blocks within Treatments, nor significant differences between Phases (Treatments combined), Treatments (Phases combined), Blocks (within Treatments; Phases combined), or among Surveys (within Phases; Treatments combined). In contrast, significant differences in turtle density estimates derived from raw observations were detected among Surveys within Phases (blocks combined) (p < 0.01), while no other terms in the analysis were significant (Table 3-10c). Pairwise comparisons revealed that during the Baseline Phase the mean density of turtles was significantly higher in D3 (0.6 ± 0.07 per km 2 ) than in B1 (0.3 ± 0.04 per km 2 ) and B2 (0.4 ± 0.05 per km 2 ), while B1 and B2 did not significantly differ. During the Dredging Phase the mean density of turtles was significantly higher in D2 (0.5 ± 0.08 per km 2 ) than in D3 (0.3 ± 0.03 per km 2 ). Prepared for INPEX Cardno Page 43

55 1 a) Marsh and Sinclair Baseline Phase 0.8 D Mean turtle density per km + SE b) Pollock et al. Baseline Phase D c) Raw sightings Baseline Phase D C1 IM C2 Location Figure 3-19 Mean turtle density per km 2 (±SE) at each location during D3 and the Baseline Phase based on: a) the Marsh and Sinclair (1989) population estimate method; b) the Pollock et al. (2006) population estimate method; and c) raw sightings Prepared for INPEX Cardno Page 44

56 Table 3-10 Results of univariate PERMANOVA tests for differences in the density of turtles per km based on: a) the Marsh and Sinclair (1989) population estimate method; b) the Pollock et al. (2006) population estimate method; and c) raw sightings between two Phases (Baseline and Dredging), Phase (B1, B2 and B3 within Baseline Phase and D3 within Dredging Phase), two Treatments (Control and Impact) and three Blocks (C1 and C2 within Control and IM within Impact) ( ) Pooled with those for the Residual/Pooled term in the final analysis; ^ No. permutations < 100 so Monte Carlo calculated p-value presented Source of Variation df MS Pseu-F p-(perm) df MS Pseu-F p-(perm) df MS Pseu-F p-(perm) (a) Marsh and Sinclair (1989) (b) Pollock et al. (2006) (c) Raw sightings Phase ^ ^ ^0.72 Treatment ^ ^0.34 Phase x Treatment Survey (Phase) > > < 0.01 Block (Treatment) Phase x Block (Treatment) (1) (0.01) (0.06) 0.80 (1) (0.01) (<0.01) 0.82 (1) (0.02) (1.09) 0.30 Treatment x Survey (Phase) (3) (0.03) (0.18) 0.90 (3) (0.04) (0.18) 0.91 (3) (<0.01) (0.29) 0.83 Survey (Phase) x Block (Treatment) (3) (0.21) (0.97) 0.42 (3) (0.18) (0.79) 0.51 (3) (0.02) (1.00) 0.39 Residual/ Pooled Total Pairwise Comparisons (ranked from largest (left) to smallest (right)) Baseline: B2 = B1 Baseline: B3 > B1 Baseline: B3 > B2 Dredging: D2 > D3 Prepared for INPEX Cardno Page 45

57 Turtle Sub-block Population Estimates Turtle sub-block population estimates (± SE; averaged across replicate flights) among the four sub-blocks within C1 ranged between 38 ± 16 and 370 ± 59 turtles based on Marsh and Sinclair (1989) methods and between 38 ± 16 and 362 ± 62 based on Pollock et al (2006) methods (Table 3-11). Population estimates among the four C2 sub-blocks ranged between 116 ± 36 and 205 ± 47 turtles, and between 104 ± 28 and 217 ± 52 turtles for those two estimation methods respectively. In IM, estimates among the six sub-blocks ranged between 72 ± 18 and 155 ± 27 turtles and between 69 ± 10 and 159 ± 29 turtles respectively (Figure 3-11, Appendix K). Table 3-11 Turtle population estimates for each sub-block during D3 based on average Marsh and Sinclair (1989) and Pollock et al. (2006) methods across replicate flights for each subblock during D3 Block Sub-block n C1 IM C2 Marsh and Sinclair (1989). Pollock et al. (2006) Population estimate SE Population estimate n = number of replicate flights where 1 turtle sighting was recorded SE Prepared for INPEX Cardno Page 46

58 Turtle Sub-block Densities Estimates of mean density of turtles (± SE) among C1 sub-blocks ranged between 0.4 ± 0.2 and 1.2 ± 0.2 per km 2 based on Marsh and Sinclair (1989) population estimates, and between 0.4 ± 0.1 and 1.2 ± 0.2 per km 2 based on Pollock et al. (2006) population estimates (Table 3-12). Mean density among C2 sub-blocks ranged between 0.4 ± 0.1 and 1.0 ± 0.2 per km 2 and between 0.4 ± 0.1 and 1.0 ± 0.3 based on the two estimation methods respectively, while among IM sub-blocks the mean density ranged between 0.3 ± 0.1 and 1.1 ± 0.3 per km 2 and between 0.3 ± 0.1 and 1.1 ± 0.2 per km 2 respectively (Table 3-12, Appendix L). Table 3-12 Turtle density estimates (per km 2 ) for each sub-block during D3 based on Marsh and Sinclair (1989) and Pollock et al. (2006) population estimate methods Block Sub-block Area (km²) C1 IM C2 Marsh and Sinclair (1989) Pollock et al. (2006) Density estimate SE Density estimate Turtle Whole Block vs. Sub-block Population Estimates Population estimates for turtles for D3 and associated standard errors derived from whole block estimates were higher than those derived using the combination of sub-block estimates for all three blocks in the cases of both the Marsh and Sinclair (1989) and Pollock et al. (2006) estimation methods. Based on the Marsh and Sinclair (1989) population estimate method, mean turtle population size estimates (± SE) derived from whole blocks were 650 ± 85 for C1, 641 ± 91 for IM and 565 ± 113 for C2 (Figure 3-20). Estimates derived from the combination of sub-block estimates using the Marsh and Sinclair (1989) method were 552 ± 65 for C1, 587 ± 56 for IM and 524 ± 113 for C2. Based on the Pollock et al. (2006) method, the mean population size estimates derived from whole blocks were 649 ± 95 for C1, 641 ± 60 for IM and 566 ± 107 for C2 (Figure 3-20). Estimates derived from the combination sub-block estimates based on the Pollock et al. (2006) method were 548 ± 65 for C1, 586 ± 44 for IM and 524 ± 107 for C2 (Appendix I and K). SE Prepared for INPEX Cardno Page 47

59 800 a) Marsh and Sinclair Block 700 Sub-block Mean turtle population estimate + SE b) Pollock et al. Block Sub-block C1 IM C2 Location Figure 3-20 Mean population estimates (±SE) for turtles derived from whole block and sub-block calculations based on the: a) Marsh and Sinclair (1989); and b) Pollock et al. (2006) population estimate methods at each location during D3 Prepared for INPEX Cardno Page 48

60 3.2 Land Surveys Survey Effort The mean sea state (BSS ± SE) at the start of 30-minute observation periods during the first and second sampling surveys at Channel Island Bridge was 1.5 ± 0.3 and 2.9 ± 0.2, respectively, while at Cullen Bay rock wall, mean sea state was 1.8 ± 0.1 and 1.3 ± 0.2 for the two surveys respectively (Table 3-13). Land-based surveys undertaken on the first and second surveys commenced during a rising and falling neap tide respectively, while the tidal change over which sampling was conducted was 1.56 m and 1.26 m respectively (Table 3-13). Raw data for each land survey period are presented in Appendix M. Table 3-13 Survey effort for land-based observations during D3 Date 30 July 2013 Location surveyed Channel Island Bridge Tidal window (cm) Survey start time Mean (±SE) sea state at each start time Mean (±SE) turbidity at each start time : (0.3) 4.0 (0.0) 30 July 2013 Cullen Bay rock wall : (0.1) 4.0 (0.0) 3 August 2013 Channel Island Bridge : (0.2) 4.0 (0.0) 3 August 2013 Cullen Bay rock wall : (0.2) 4.0 (0.0) Channel Island Bridge Two hundred and ninety one turtle sightings were recorded from Channel Island Bridge during D3, with the majority off the northern side of the bridge and during the second survey (Figure 3-21, Table 3-14). Ninetysix per cent of sightings were identified as individual turtles (n = 280), with 95% of those identified as green turtles and the remainder unable to be identified to species level. The majority (93%) of turtles observed from Channel Island Bridge were considered to be juvenile. No dolphins or dugongs were observed during either of the two surveys at Channel Island Bridge during D Cullen Bay Rock Wall At Cullen Bay Rock Wall during D3 a single dugong sighting was recorded during the first survey and again, another single dugong was sighted several times over an approximate 50-minute period during the second survey. A pod of dolphins consisting of 4 to 10 adults was sighted to the north-west of the wall during the first survey and observed over a 2-hour period. Dolphins within this pod displayed behaviour consistent with socialising and foraging activities. No turtles were observed during either of the two surveys at Cullen Bay rock wall (Table 3-14). Prepared for INPEX Cardno Page 49

61 Table 3-14 Number of megafaunal sightings and number of individual animals observed at Channel Island Bridge and Cullen Bay rock wall during D3 Channel Island Bridge Survey 1 Survey 2 Total sightings Individuals* Total sightings Individuals* Unidentified turtle Green turtle Dugong Dolphin Total sightings Cullen Bay Rock Wall Survey 1 Survey 2 Total sightings Individuals* Total sightings Individuals* Unidentified turtle Green turtle Dugong Dolphin 1^ Total sightings * Individual turtles recognised through recognisable features or observation of dive behaviour ^ Refers to group sighting, with between 4 and 10 individual dolphins in the one group Prepared for INPEX Cardno Page 50

62 Figure 3-21 Distribution of species sighted during land surveys in D3 (July 2013). N.B. Sightings of dolphins (blue dots) at Cullen Bay are a group of between 4 and 10 individuals sighted over a 2 hour period Prepared for INPEX Cardno 51

63 3.3 Rehabilitated Turtle Update A summary of all rehabilitated turtles (current and past) reported to Cardno by the Arafura Timor Research Facility or DLRM since the commencement of the Project is provided in Appendix N. Rehabilitated turtles have included one olive ridley and six flatback turtles. To date, no satellite tagging of rehabilitated turtles has been undertaken by Cardno during the Project. Of the ten turtles rehabilitated to date, only three belonged to species targeted for satellite tagging (one green and two hawksbill turtles), with all unsuitable for tagging due to their existing injuries. One hawksbill turtle was found outside of the Project area (Dundee Beach, west of Bynoe Harbour) and was subsequently released at this location by Arafura Timor Research Facility staff. Another hawksbill turtle had lost its fore flipper and the green turtle had severe head injuries, therefore both were immunocompromised and not suitable for tagging for this Project. 3.4 Quality Assurance and Quality Control Within the aerial survey observer database, a total of 256 line entries were quality checked, with three errors detected (1%). As part of the Environmental database quality control check, 272 line entries were checked, with 19 errors detected (7%). During the master database quality check there were 330 entries checked, with two errors detected (1%). All errors were subsequently corrected. All entries in the land-based survey observer database were checked against the datasheets for quality assurance and no data entry errors were found. Prepared for INPEX Cardno 52

64 4 Discussion The TDMP incorporates various visual survey techniques to estimate and monitor turtle and dugong populations within Darwin Harbour, Hope Inlet, Shoal Bay, Bynoe Harbour and from the Vernon Islands to Melville Island. Sampling for D3 was completed during July 2013 and August 2013, approximately eight months after the commencement of CSD operations. Data collected via standardised aerial transects during D3 has enabled a comparison of sightings, population estimates and density estimates at three locations; Bynoe Harbour (C1), Darwin Harbour/ Hope Inlet (IM) and around the Vernon Islands (C2), with comparable data collected during the Baseline Phase. Land observations undertaken during D3 have provided an opportunity to identify fine-scale turtle and dugong habitat associations within Darwin Harbour. 4.1 Aerial Surveys Dugongs The number of dugongs sightings during D3 aerial surveys (71 individuals in total) was approximately 19% lower than the average for Baseline Phase (mean of 88 individuals), but more than twice the number of individuals sighted during D2 (33 individuals in total) which was undertaken in May Fewer dugongs were sighted during D3 than during the survey undertaken at a similar time of year during the Baseline Phase (B2, July 2012, 92 individuals in total). Dugong population estimates within the Darwin Harbour region (IM) during D3, estimated using two alternative population estimation methods (Marsh and Sinclair (1989) and Pollock et al. (2006)), were 316 ± 121 and 167 ± 39 individuals, respectively. These estimates were 12% and 28% lower, respectively, than the Baseline Phase estimates, but substantially higher than estimates for D2 (90 and 99 individuals for the two methods, respectively). Dugong population estimates for the Bynoe Harbour region (C1) were 134 ± 58 individuals using the Marsh and Sinclair (1989) method and 48 ± 12 individuals based on the Pollock et al. (2006) method. These estimates were lower than Baseline Phase estimates of 190 ± 103 and 167 ± 168 individuals, respectively. Fewer than five dugongs were sighted during any replicate flight completed within the C2 block during D3, preventing the calculation of reliable population estimates. In comparison, during the B2, the calculated population estimate for IM was 411 ± 139 based on the Marsh and Sinclair (1989) method and 249 ± 150 using the Pollock et al. (2006) method based on four of the nine replicate flights. Results from previous surveys (Cardno 2012a; 2013b, c) and from D3 suggest that the sizes of dugong populations in the three blocks have been spatially and temporally variable throughout the monitoring program. Environmental conditions, such as sea state and turbidity, can affect the number of dugong sightings and consequently population estimates (Marsh and Sinclair 1989). As these environmental conditions have been relatively similar among surveys, the observed variability in dugong sightings is likely to be associated with inherent behaviour characteristics of dugongs, including being highly mobile, migratory and frequently submerging (Bayliss and Freeland 1989; Marsh et al. 2002). Seasonal trends in dugong population estimates have been recorded throughout their geographic range, particularly within Moreton Bay, Queensland (Preen 1993). Additional surveys in Darwin may further facilitate examination of the effect of seasonality on dugong population estimates within Darwin Harbour and surrounding locations. Dugongs were distributed throughout the entire survey area, although the frequency of sightings was not uniform among survey blocks. Dugongs were observed in association with shallow areas near the Vernon Islands, off the islands at the mouth of Bynoe Harbour, such as north around Bare Sand Island or around Dum In Mirrie Island, but in far greater frequency throughout Hope Inlet. The density of dugong sightings has not differed statistically through time or among blocks for the duration of the TDMP. While no dugongs were observed in the Upper East Arm of Darwin Harbour during D3, some dugongs were sighted around Blaydin Point and Channel Island during sampling done at the corresponding time of the year during the Baseline Phase (i.e. B2). This disparity resulted in an estimated change in density for that area to be in excess of many other areas inside and outside Darwin Harbour. However, decreases in dugong density of similar extent were also apparent for areas well away from dredging activities (i.e. west arm). This inter- Prepared for INPEX Cardno 53

65 survey variability may be a function of re-distribution of animals resulting from short-term movement in and out of specific areas, possibly due to avoidance behaviour and/or the pursuit of more optimal foraging areas. As observed during previous surveys, the highest densities of dugong sightings during D3 were recorded in Shoal Bay and Hope Inlet in IM, which corresponds with the presence of a large seagrass area dominated by Halophila decipiens (Geo Oceans 2013b). Seagrass habitat mapping surveys in the IM block (conducted every three months as part of the Seagrass Monitoring Program) have indicated that ephemeral H. decipiens seagrass beds are usually present in the area between Charles Point and Lee Point from approximately May to October. This genus of seagrass has been recognised as a preferred food source of dugongs and is possibly actively selected by dugongs over other seagrass species on the basis of higher nutritional value, nitrogen content and less inedible fibre (Preen 1993). The consistently high concentrations of dugongs in habitats with known H. decipiens beds such as around Lee Point and Shoal Bay indicate these areas are likely to be important foraging grounds for dugongs. Lanyon (2003) has suggested that the repeated regrazing of particular areas by dugongs could be described as a form of cultivation grazing that could serve to maintain the nutritional quality of their seagrass diet (Preen 2005) Turtles Five hundred and twenty nine turtles were sighted during D3, which was approximately 17% lower than the average number of turtles sighted per survey during the Baseline Phase (634 turtles), and approximately 16% lower than the number of sightings during D2 in May 2013 (627 turtles). Turtle sightings during D3 were also slightly lower (by ~10%) than the number recorded during B2 (585 turtles), undertaken at the same time of year (July) in The average number of turtle sightings per replicate flight was similar across all surveyed areas (C1: 58, IM: 54, C2: 64) during D3, and generally lower than corresponding averages for Baseline Phase surveys (65, 86, and 87 turtle sightings, respectively). In the case of IM, the mean turtle sightings per flight during D3 was approximately 12% lower compared to the B2 average (69 turtles per flight), while for blocks C1 and C2 the D3 average was similar to that recorded for B2 (62 and 64 turtles, respectively). Mean turtle population estimates for D3 were lower than the average of the Baseline Phase population estimates for all three blocks and in the cases of both population estimation methods (Marsh and Sinclair (1989) and Pollock et al. (2006)). However, population estimates for Baseline Phase surveys was highly variable. Statistical analyses of density of turtle sightings revealed no significant difference in density between Impact and Control Treatments in either Phase or between Baseline and Dredging Phases for either Treatment, indicating that there is no evidence that the spatial and temporal patterns in turtle densities in the region have been affected by dredging. In contrast, the density of turtles based on raw observations was significantly higher during D2 compared to D3 and higher in B3 compared to either B1 or B2. Inter-survey temporal variation in turtle densities could be a result of short-term movements in and out of specific areas or due to the ongoing pursuit of optimal foraging grounds. The smaller-scale spatial distribution of turtles within each block was generally similar during D3 compared with the Baseline Phase. Turtle sightings to date have most frequently been recorded within relatively shallow water habitats. Where benthic habitat type has been identified, turtle sightings have primarily been recorded over reef, sand and seagrass. Known foraging habitats of green turtles include coral and rocky reefs, and inshore areas that support algae and/or seagrass growth (Whiting 2000; Limpus 2008). The spatial distribution of turtle sightings may therefore have been influenced by the distribution of habitats in which they forage. Similarly, the inter-survey variability in turtle density observed for areas inside and outside Darwin Harbour may also be a result of such patterns in behaviour. It is unclear whether the turtles sighted in deeper waters were travelling between shallow water locations or foraging within unsurveyed habitats. Density estimates of turtles in D3 (across both population estimation methods) ranged from approximately 0.4 turtles per km 2 in C2, to approximately 0.8 per km 2 in C1. These values are generally similar to those reported in Shark Bay, Western Australia (0.4 per km 2, Preen et al. 1997), but lower than those reported for the Torres Straight (2.1 per km 2, Marsh and Lawler 1992). Prepared for INPEX Cardno 54

66 4.1.3 Block vs. sub-block population and variance estimates The dugong and turtle population size estimates derived for each block during D3 (and previous surveys (Cardno 2012a, 2013b, c)) have been quite imprecise. Partitioning of survey blocks into non-overlapping sub-blocks that each comprise of relatively homogeneous characteristics (i.e. habitat type and depth profile) has the potential to reduce imprecision associated with these estimates (Cochran 1963). Mean turtle population estimates and associated variances for each block in D3 were lower when calculated by combining sub-block estimates compared to when derived from whole block estimates for both population estimation methods (i.e. Marsh and Sinclair (1989) and Pollock et al. (2006)). In contrast, mean dugong population estimates based on sub-blocks were higher than those based on whole block estimates at both IM and C1 when the Pollock et al. (2006) method was used, but only in IM when the Marsh and Sinclair (1989) method was used. Variances associated with dugong population estimates derived by the combination of sub-blocks were similar to those derived from whole blocks regardless of the population estimation method used. The Marsh and Sinclair (1989) and Pollock et al. (2006) population estimation methods used to determine dugong and turtle population sizes are sensitive to small sample sizes and become increasingly unreliable and variable as fewer animal sightings are recorded. Unlike the whole block population estimates, sub-block population estimates for dugongs were calculated when fewer than five sightings were recorded, and in the cases of C1-3, IM-2, IM-3 and IM-6, calculated when only one individual dugong was sighted. Hence, producing dugong population estimates by combining sub-block estimates with few sightings probably contributed to the relatively higher population estimates compared to whole block estimates, which were only calculated when only five or more dugong sighting were recorded. The precision associated with dugong population estimates derived from the combination of sub-block estimates and whole block estimates were not substantially different. As few sub-blocks within each of the three sampling locations have recorded greater than five dugong sightings during a particular replicate flight, it is unlikely that partitioning of whole blocks into sub-blocks would reduce the variance associated with dugong population estimates or provide more accurate estimates of the dugong population size if applied to data recorded during future surveys. The precision associated with turtle population estimates derived from the combination of sub-block estimates was higher than for those derived via whole block estimates, for all three blocks. Unlike dugongs, turtle sightings were more frequent and, with the exception of C2-1 during replicate flight three, tallied to more than five sightings for any replicate flight within any particular sub-block. The non-uniform distribution and abundance of turtles within blocks likely facilitated the observed reduction in variance associated with estimated population sizes based on sub-blocks compared to whole block population estimates. If the distribution of turtles remains non-uniform during future surveys, combining sub-block population estimates may prove beneficial in improving precision associated with turtle population estimates. 4.2 Land Surveys One dugong was sighted off the rock wall at Cullen Bay on each of the two days of land observations (30 July 2013 and 3 August 2013) during D3, while no turtles were observed during the same period. The dugongs were observed close to mapped Halodule seagrass habitat (Geo Oceans 2013a). The long diving times observed within this area indicated that these dugongs may have been foraging during this observational period. Several dolphins were also sighted in the area, displaying distinctive foraging behaviour. At Channel Island Bridge, 280 individual turtles were sighted during D3. The majority of these turtles were identified as green turtles and many of these classified as juveniles. In contrast to Cullen Bay rock wall observations, no dugongs were observed from Channel Island Bridge. The large number of turtle observations recorded at Channel Island Bridge indicated that Channel Island may provide an important habitat for turtles, with the extensive reef habitat surrounding Channel Island possibly providing important foraging grounds. 4.3 Rehabilitated Turtle Update To date, no satellite tagging of rehabilitated turtles has been conducted as part of the TDMP. The ten turtles that have been rehabilitated and released by Arafura Timor Research Facility or DLRM since the start of the TDMP have proved unsuitable for tagging due to severe injuries or having originated from areas outside Darwin Harbour. Assessments of the suitability of rehabilitated turtles for satellite tagging will continue as part of the TDMP. Prepared for INPEX Cardno 55

67 5 Conclusions It is important to gain a comprehensive spatial and temporal understanding of abundance, density and distribution of dugongs and turtles in Darwin Harbour and surrounding locations, especially during potential disturbance events. Aerial surveys have been appropriate for the collection of broad-scale data to meet this objective and the broad objectives of the TDMP. Aerial surveys during D3 (July 2013 and August 2013) resulted in 71 and 529 individual sightings of dugongs and turtles respectively. The number of dugong and turtle sightings was slightly lower during D3 compared to the Baseline Phase survey completed at the corresponding time of year (B2 July/August 2012) at each of the three locations sampled, although there is no clear evidence to suggest that the distribution or abundance of these animals have changed as a consequence of dredging activities. The temporal and spatial variation in dugong and turtle distribution and abundance observed may be a result of movement in and out of specific areas, possibly due to avoidance behaviour and/or the pursuit of more optimal foraging areas. Dugong sightings within IM were concentrated within the Shoal Bay area between Lee Point and Charles Point. As per previous surveys, the patchy distribution of dugongs within IM may be associated with the distribution of seagrass beds in the area, which could serve as foraging areas. Similarly, many turtle sightings were recorded in association with habitats such as reefs, sand and seagrass, which they may use as foraging areas. These observations indicate that the spatial and temporal distribution of both dugongs and turtles is most likely strongly influenced by the availability of suitable foraging habitats. Aerial survey blocks were further partitioned into non-overlapping sub-blocks comprised of relatively homogeneous characteristics (i.e. habitat type and depth profile), with the aim of reducing imprecision associated with block population estimates. Variances associated with dugong block population estimates derived via sub-blocks were similar to those derived from whole block estimates regardless of the population estimation method used. Conversely, the non-uniform distribution and abundance of turtles within blocks facilitated an observed reduction in variance associated with estimated population sizes based on sub-blocks compared to whole block population estimates. If the distribution of turtles remains non-uniform during future surveys, combining sub-block population estimates may prove beneficial in improving precision associated with turtle block population estimates. Continued monitoring of the distribution and abundance of dugongs and turtles within Darwin Harbour and surrounding locations will provide a strong foundation with which to investigate long-term trends during the dredging and post dredging phases of the Project. Prepared for INPEX Cardno 56

68 6 Acknowledgements This report was written by Dr Michelle Blewitt, Megan Rice and Dr Lachlan Barnes, and reviewed by Dr Will Macbeth and Dr Craig Blount. Fieldwork was undertaken by Megan Rice, Dr Glenn Dunshea, Dr Susan Gibbs, Barry Krueger and Paul Tod. GIS support was provided by Amy Steiger and the Wollongong Cardno team. Data analysis was undertaken by Dr Glenn Dunshea, Dr Lachlan Barnes, Megan Rice and Dr Michelle Blewitt. The figures, tables and appendices were prepared by Dr Lachlan Barnes, Megan Rice and Dr Michelle Blewitt. The TDMP was reviewed by Professor Helene Marsh from James Cook University. The authors would like to thank the Ichthys Project Dredging Expert Panel (IPDEP) members for constructive criticism throughout the ongoing monitoring program. Prepared for INPEX Cardno 57

69 7 References Anderson, M.J. (2001). A new method for non-parametric multivariate analysis of variance. Austral Ecology 26, pp Bayliss, P. and Freeland, W.J. (1989). Seasonal distribution and abundance of dugongs in the Western Gulf of Carpentaria. Australian Wildlife Research 16, pp Cardno (2012a). Turtle and Dugong Monitoring Program Baseline Report.. Prepared for INPEX. November 2012, pp Cardno (2013a). Nearshore Environmental Monitoring Plan (Rev. 2). Prepared for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013b). Routine Turtle and Dugong Monitoring Program Report Dredging Report 1. Ichthys Nearshore Environmental Monitoring Program. Prepared for INPEX. January 2013, pp Cardno (2013c). Routine Turtle and Dugong Monitoring Program Report Dredging Report 2. Ichthys Nearshore Environmental Monitoring Program. Prepared for INPEX. July 2013, pp Chatto, R. and Baker, B. (2008). The Distribution and Status of Marine Turtle Nesting in the Northern Territory. Technical Report 77. Parks and Wildlife Service, Department of Natural Resources, Environment, The Arts and Sport. Cochran, W.G. (1963). Sampling Techniques, 2 nd Ed., John Wiley and Sons, New York. Geo Oceans (2012). : Seagrass Habitat Monitoring Survey, October Technical Report. Prepared for Cardno on behalf of INPEX, November Geo Oceans (2013a). : Seagrass Habitat Monitoring Survey, July Technical Report. Prepared for Cardno on behalf of INPEX, June Geo Oceans (2013b). : Seagrass Habitat Monitoring Survey, October Technical Report. Prepared for Cardno on behalf of INPEX, October Hodgson, A., Marsh, H., Marsh, L., Grech, A., McMahon, A. and Parra, G. (2011). Dugong Aerial Survey Manual. James Cook University, Townsville. INPEX (2012). Dredging and Spoil Disposal Management Plan - East Arm. INPEX Operations Australia Pty Ltd. Lanyon, J.M. (1991). The nutritional ecology of the dugong (Dugong dugon) in tropical north Queensland. Ph.D. Thesis, Monash University, Melbourne. Lanyon, J.M. (2003). Distribution and abundance of dugongs in Moreton Bay, Queensland, Australia. Wildlife Research 30, pp Limpus, C.J. (2008). Green turtle, Chelonia mydas (Linnaeus). In: A Biological Review of Australian Marine Turtles. Queensland Government Environmental Protection Agency. Limpus, C.J. and Chaloupka, M. (1997). Nonparametric regression modelling of green sea turtle growth rates (southern Great Barrier Reef). Marine Ecology Progress Series 149, pp Makowski, C., Seminoff, J.A. and Salmon, M. (2006). Home range and habitat use of juvenile Atlantic green turtles (Chelonia mydas L.) on shallow reef habitats in Palm Beach, Florida, USA. Marine Biology. 148, pp Marsh, H. and Lawler, I. (1992). The Status of Dugongs, Sea Turtles and Dolphins in the Torres Strait Region. Report to The Australian Fish Management Authority., June Marsh, H. and Sinclair, D.F. (1989). Correcting for visibility bias in strip transect aerial surveys of aquatic fauna. Journal of Wildlife Management 53, pp Marsh, H., Penrose, H., Eros, C. and Hughes, J. (2002). Dugong: status reports and action plans for countries and territories in its range. Cambridge, UK: IUCN. Pollock, K., Marsh, H., Lawler, I. R. and Alldredge, M.W. (2006). Estimating animal abundance in heterogeneous environments: an application to aerial surveys for dugongs. Journal of Wildlife Management 70, pp Preen, A.R. (1993). Interactions between dugongs and seagrasses in a sub-tropical environment. Ph.D. Thesis, James Cook University, Townsville. Preen, A. (1995). Impacts of dugong foraging on seagrass habitats: observational and experimental evidence for cultivation grazing. Marine Ecology Progress Series 124, pp Preen A.R., Marsh, H., Lawler, I.R., Prince, R.I.T. and Shepherd, R. (1997). Distribution and abundance of dugongs, turtles, dolphins and other megafauna in Shark Bay, Ningaloo Reef and Exmouth Gulf, Western Australia. Wildlife Research 24, pp Prepared for INPEX Cardno 58

70 Whiteway, T. (2009). Australian Bathymetry and Topography Grid. Geoscience Australia. Whiting, S.D. (2000). The ecology of immature green and hawksbill turtles foraging on two reef systems in northern Australia, Ph.D. Thesis. Northern Territory University, Darwin. Whiting, S.D. (2004). Conserving dugongs in Darwin Harbour. Report to Natural Heritage Trust, Coast and Clean Seas Project, # Prepared for INPEX Cardno 59

71 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Ichthys Nearshore Environmental Monitoring Program APPENDIX A SUMMARY OF AERIAL SURVEY EFFORT DURING BASELINE AND DREDGING PHASE SURVEYS Prepared for INPEX Cardno 1

72 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Appendix A: Summary of aerial survey effort for each block (C1, IM and C2) during Baseline and Dredging Phase surveys Bynoe Harbour (C1) Darwin Harbour (IM) Vernon Islands (C2) Baseline Survey Period B1 B2 B3 D2 D3 B1 B2 B3 D2 D3 B1 B2 B3 D2 D3 Number of transects ^ ^ Total length transectsª (km) ^ ^ Total sampled area (km²) ^ ^ Overall surveyed area (km²) Proportion surveyed* (%) ^ Average survey duration (h: min: sec) 0:54:17 1:23:23 1:27:53 1:06:32 1:28:45 1:25:47 1:25:36 1:26:55 1:36:53 2:21:30 1:36:17 1:12:08 1:20:3 6 1:15:09 1:28:51 ª Transect length over water. * The proportion of surveyed area is averaged in the calculation of total area surveyed. B1 = 2 replicates, B2 and B3 = 3 replicates; D2 = 2/3 replicates. ^ Variation due to flight orientation. Prepared for INPEX Cardno 61

73 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Ichthys Nearshore Environmental Monitoring Program APPENDIX B SUMMARY OF DUGONG SIGHTINGS ACROSS DREDGING AND BASELINE PHASE SURVEYS Prepared for INPEX Cardno

74 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Appendix B: Summary of the number of dugongs recorded during Dredging and Baseline Phase All areas Total sightings Survey period D3 B1 B2 B3 D2 Replication number R1 R2 R3 Total Number of observations Number of animals Number of calves Maximum group size Bynoe Harbour (C1) Total sightings Survey period D3 B1 B2 B3 D2 Replication number R1 R2 R3 Total Number of observations Number of animals Number of calves Maximum group size Darwin Harbour (IM) Total sightings Survey period D3 B1 B2 B3 D2 Replication number R1 R2 R3 Total Number of observations Number of animals Number of calves Maximum group size Vernon Islands (C2) Total sightings Survey period D3 B1 B2 B3 D2 Replication number R1 R2 R3 Total Number of observations Number of animals Number of calves Maximum group size B1= 2 replicate flights; D2= 2-3 replicate flights; B2, B3, D3 = 3 replicate flights Prepared for INPEX Cardno 63

75 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Ichthys Nearshore Environmental Monitoring Program APPENDIX C SUMMARY OF TURTLE AND DUGONG SIGHTINGS OVERLAID WITH CONFIRMED SEAGRASS HABITAT DURING B2 AND D3 SURVEYS Prepared for INPEX Cardno

76 Routine Turtle and Dugong Monitoring Program Report- Dredging Report 3 Appendix C: Distribution of dugongs and turtles in Darwin Harbour (IM) in July 2012 (B2) and July 2013 (D3), overlaid with confirmed seagrass habitat (seagrass data for June 2012 and August 2013, respectively: Geo Oceans 2013) Prepared for INPEX Cardno 65