Coral Monitoring Report Dredging Report 12

Transcription

1 . Coral Monitoring Report Dredging Report 12 Ichthys Nearshore Environmental Monitoring Program L384-AW-REP L384-AW-REP Prepared for INPEX June 2014

2 Document Information Prepared for INPEX Project Name File Reference L384-AW-REP-10091_0_Coral Monitoring Dredging Report 12.docm Job Reference L384-AW-REP Date June 2014 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 02/06/2014 Andrea Nicastro David Cummings AN DC Craig Blount Joanna Lamb CB JL B 10/06/2014 Andrea Nicastro AN Craig Blount Joanna Lamb 0 12/06/2014 Andrea Nicastro AN Craig Blount Joanna Lamb CB JL CB JL 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 ii

3 Executive Summary A Coral Monitoring Program has been developed to detect potential changes in coral health indicators and to infer whether any changes to these indicators are the result of dredging and/or spoil disposal activities associated with the Ichthys Project in Darwin Harbour. A Baseline Phase of three coral surveys was undertaken in June 2012 to August 2012 prior to the commencement of dredging. During the Dredging Phase, the monitoring program involves regular sampling of individually tagged coral colonies and repeat surveys of permanent transects to determine trends in the condition of corals at sites throughout Darwin Harbour. These datasets are collected in parallel with the Water Quality and Subtidal Sedimentation Monitoring Program (WQSSMP) to provide a suite of early warning indicators for assessing the health of corals. Season One dredging commenced on 27 August 2012 and ceased on 30 April Season Two dredging commenced on 23 October 2013 along the gas export pipeline (GEP) route, then in East Arm (EA) on 1 November This report outlines the findings of the twelfth routine coral monitoring survey during the Dredging Phase (D12; 8 May 2014 to 10 May 2014) and covers the intervening period since the end of survey D11 (10 April 2014). Results are presented for reactive coral sites only (i.e. Channel Island, Weed Reef 1 and Weed Reef 2). East Arm sites South Shell Island and Northeast Wickham Point were not required to be surveyed in D12 in accordance with the monthly coral monitoring requirements of the East Arm Dredging and Spoil Disposal Management Plan (DSDMP) (INPEX 2013). During the D12 monitoring period there was a marginal dry season Level 1 water quality Duration and Frequency trigger exceedances at Weed Reef 1, which was primarily attributed to natural causes (INPEX 2014b). Measured coral mortality at Channel Island, Weed Reef 1 and Weed Reef 2 following the trigger exceedances was below the 95% confidence interval of the generalized linear mixed model (GLMM) predicted mortality, thus there were no Level 2 trigger exceedances. The measured mortality at Channel Island in D12 was very similar to the mortality predicted by the GLMM, whereas mortality at Weed Reef 1 and Weed Reef 2 in D12 was less than predicted. Increased mortality at Channel Island and Weed Reef 1 was mainly due to increased cover of immobile fauna (sponges and hydroids) and sediment. Conversely, decreased mortality at Weed Reef 2 in D12 was due to a decrease in sediment and immobile fauna compared with the previous survey (D11). No Level 2 coral bleaching triggers were exceeded during D12. The measured bleaching at Channel Island and Weed Reef 1 were below the average bleaching of the Baseline Phase surveys, while Weed Reef 2 was slightly above. This resulted in gross bleaching of -5.1%, -2.6% and 0.9% at Channel Island, Weed Reef 1 and Weed Reef 2, respectively. It is important to note that tagged corals are useful as an indicator of potential rates of mortality, but are not suitable for measuring the overall condition of a coral community. This is because an individual tagged coral colony will inevitably show partial or complete mortality, given a long enough monitoring program. If the mortality of tagged corals was occurring at an unnaturally high rate, it would be expected that the amount of coral present in the whole reef/site would also be declining. Mortality greater than replacement through recruitment and growth can be captured through transects. In D12, there was no significant change in sitewide coral cover (along transects) at any site, or for all sites combined. The absence of a clear decrease in coral cover suggests that the mortality observed in the tagged corals is natural and is likely being compensated at a site-wide level by growth and, possibly, albeit to a lesser extent, by recruitment. Sediment on coral, which constitutes up to half of the partial mortality of tagged corals recorded, has generally increased through time but it has also shown a large degree of temporal variability. This suggests that the sediment may be settling on living coral, which can then be actively and passively cleared by the coral before subsequent surveys. The increase in coral mortality since D11 at Channel Island and Weed Reef 1 was partially associated with increases in sediment on coral However, this trend was not evident at a community level as sand and silt cover along transects at all reactive sites decreased significantly compared to D11 and was similar to Baseline Phase levels. Prepared for INPEX Cardno iii

4 Glossary Term or Acronym Actual mortality ADAS AIMS BACI Benthic assemblages Bleached Bray-Curtis dissimilarity matrix B1 B2 B3 CHI CHP Contingency management CPCe D1 D2 Definition Partial mortality in a coral colony where the tissue is dead Australian Diver Accreditation Scheme Australian Institute of Marine Science Before After Control Impact Biota (living) and abiota (non-living) components of the sea bed Corals that have lost their symbiotic algae due to stress and the live tissue of which appears pale or white An index of dissimilarity between samples in the types and relative abundance of species First Baseline Phase survey prior to commencement of dredging activities (16 June 2012 to 18 July 2012) Second Baseline Phase survey prior to commencement of dredging activities (27 July 2012 to 30 July 2012) Third Baseline Phase survey prior to commencement of dredging activities (11 August 2012 to 14 August 2012) Channel Island Charles Point Management based on defaulting to more environmentally secure operations when required Coral Point Count with Excel extensions First Dredging Phase survey after commencement of dredging activities (22 October 2012 to 26 October 2012) Second Dredging Phase survey after commencement of dredging activities (5 December 2012 to 9 December 2012) Prepared for INPEX Cardno iv

5 Term or Acronym D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 DSDMP EA GEP GLMM Informative monitoring MAN Definition Third Dredging Phase survey after commencement of dredging activities (17 February 2013 to 22 February 2013) Fourth Dredging Phase survey after commencement of dredging activities (17 April 2013 to 21 April 2013) Fifth Dredging Phase survey after commencement of dredging activities (16 June 2013 to 20 June 2013) Sixth Dredging Phase survey after commencement of dredging activities (14 August 2013 to 19 August 2013) Seventh Dredging Phase survey after commencement of dredging activities (26 October 2013 to 30 October 2013) Eighth Dredging Phase survey after commencement of dredging activities (9 December 2013 to 14 December 2013) Ninth Dredging Phase survey after commencement of dredging activities (23 February 2014 to 28 February 2014) Tenth Dredging Phase survey after commencement of dredging activities (8 March 2014 to 14 March 2014) Eleventh Dredging Phase survey after commencement of dredging activities (7 April 2014 to 10 April 2014) Twelfth Dredging Phase survey after commencement of dredging activities (8 May 2014 to 10 May 2014) Dredging and Spoil Disposal Management Plan East Arm Gas Export Pipeline Generalized Linear Mixed Model Monitoring programs designed to measure environmental responses to dredging and spoil disposal activities and to provide textual information on effects of sedimentation and turbidity on sensitive receptors Mandorah Prepared for INPEX Cardno v

6 Term or Acronym NEW NT EPA NTU Pale bleached PAR Peak flux Perceived mortality PERMANOVA Reactive monitoring SCUBA SSBA SSI TARP TSHD UCL White bleached Definition Northeast Wickham Point Northern Territory Environment Protection Authority Nephelometric Turbidity Units Bleached tissue that appears to be a lighter hue than healthy tissue Photosynthetically active radiation In relation to PAR measurements Mortality is assumed as part of a coral is obscured by sediment, mobile fauna or algal fronds. Permutational analysis of variance Monitoring programs that include triggers that initiate targeted monitoring and adaptive and contingency management responses to manage impacts within the limits of acceptable loss Self-contained underwater breathing apparatus Surface supply breathing apparatus South Shell Island Trigger Action Response Plan Trailer suction hopper dredger Upper confidence limit Bleached tissue that appears white or near-white WR1 Weed Reef 1 WR2 Weed Reef 2 Zooxanthellae Symbiotic algae that live in coral tissue and provide nutrition to coral hosts Prepared for INPEX Cardno vi

7 Table of Contents Executive Summary Glossary iii iv 1 Introduction Background Requirement to Monitor Corals Results of the Previous Surveys Baseline Surveys (B1 to B3) Season One Dredging Surveys (D1 to D4) and Dry Season Hiatus (D5 and D6) Objectives 3 2 Methodology Overview Vessels, Diving, Safety and Environmental Management Sites, Timing and Frequency of Surveys Tagged Colonies Mortality and Bleaching Sediment Accumulation Coral Colour and Area Fixed Transects Percentage Composition Recruitment (Juvenile Corals) Data Analysis Multivariate and Univariate Analyses of Variance Regression Analysis Physical Environment Quality Control 14 3 Dredging Operations 15 4 Results Tagged Colonies Mortality Bleaching Colony Surface Area Fixed Transects Composition of Sites (Multivariate Analyses) Percentage Cover Analyses (Univariate Analyses) Recruitment (Juvenile Corals) Comparative Analysis of Tagged Corals and Fixed Transects Sediment on Tagged Corals and Transects Mortality of Tagged Corals and Transect Coral Cover Physical Environment 37 5 Discussion Trigger Assessment Level 1 Water Quality Coral Mortality Coral Bleaching Interpretation and Attributability 40 Prepared for INPEX Cardno vii

8 5.2.1 Comparison of Mortality Rates Coral Mortality Sediment on Coral Bleaching Coral Recruits 44 6 Conclusions 45 7 Acknowledgements 46 8 References 47 Tables Table 1-1 Channel Island coral trigger values (INPEX 2013, 2014a) 5 Table 2-1 Field sampling dates for coral monitoring in the Baseline and Dredging Phases 8 Table 2-2 Summary of the number of tagged colonies included in the mortality estimate (max = 75) and fixed transects sampled (max = 4) during D12 8 Table 2-3 Coral health categories for tagged coral CPCe analysis 9 Table 2-4 Explanation of factors used in statistical analyses 13 Table 2-5 Terms used in describing the outcomes of statistical analyses 13 Table 3-1 East Arm dredge footprint summary 15 Table 4-1 Table 4-2 Measured mortality, predicted mortality and trigger values for D12 from GLMM trend analysis using data from B1 to D7 16 Comparison of the predicted mean coral mortality and associated 95% UCL at the end of Season Two dredging based on GLMM trend analyses using data from B1 to D7 and data from B1 to D12 18 Table 4-3 Recorded gross bleaching and trigger values for D12 24 Table 4-4 Table 4-5 Table 4-6 Number of corals showing an increase in coral brightness (positive coral class change), decrease in coral brightness (negative coral class change), and no change in coral brightness between surveys D11 and D12 and between surveys B3 and D12 26 Summary of multivariate PERMANOVA analyses of entire benthic composition and coral assemblages at the monitoring sites 28 Summary of univariate PERMANOVA analyses of hard coral, sand and silt, and turf algae at the monitoring sites 31 Table 4-7 Summary of univariate PERMANOVA analyses of the number of coral recruits for all sites 33 Table 4-8 Summary of daily-averaged turbidity, daily-averaged water temperature, PAR peak flux and PAR daily dose (mean, minimum and maximum) at monitoring stations during the current reporting period, previous reporting period and the period corresponding to the current reporting period from the previous year 38 Figures Figure 2-1 Figure 4-1 Figure 4-2 Locations of coral monitoring sites 7 GLMM trend analysis of mean coral mortality (red line) and 95% confidence limits (green lines) for surveys B1 to D7 (blue dots). Observed mean partial mortality (red dots) for D8 to D12 was not included in GLMM trend analysis 17 GLMM trend analysis of mean coral mortality (red line) and 95% confidence limits (green lines) for all surveys to date (B1 to D12) 19 Prepared for INPEX Cardno viii

9 Figure 4-3 Figure 4-4 Figure 4-5 Figure 4-6 Figure 4-7 Figure 4-8 Figure 4-9 Figure 4-10 Contribution (%) of each of the coral mortality categories to the mean mortality per coral colony at the monitoring sites in each survey and the remaining number of the original 75 tagged corals included in the mortality estimate (i.e. photograph obtained and of suitable quality) 21 Sediment depth on pavers (mm) from survey D2 to D12 and sediment cover (%) on tagged corals from B1 to D12 22 Relationship between sediment cover on corals (%) and sediment depth on adjacent pavers (mm) during D12 23 Average (±SE) cover (%) of sediment on different growth forms during B1 to D12 for all sites combined 23 Mean (± SE) partial bleaching (%) of tagged corals at each monitoring site for surveys B1 to D12 24 Gross (mean ± SE) coral bleaching for surveys D1 to D12 25 Mean (± SE) coral brightness class for surveys B1 to D12 for: a) darkest coral tissue; and b) lightest coral tissue 26 Proportional change in coral area (%) during D1 to D12 relative to area recorded in B3 27 Figure 4-11 Contribution (%) of main categories to the composition of benthic assemblage along transects at the monitoring sites 29 Figure 4-12 nmds ordinations based on: a) the entire benthic composition; b) coral families; and c) coral growth forms at the monitoring sites 30 Figure 4-13 Figure 4-14 Figure 4-15 Figure 4-16 Mean (±SE) hard coral, sand/silt and turf algae cover for all sites 32 Mean (±SE) number of coral recruits at all sites 33 Sediment on tagged corals and total sediment cover along transects from B1 to D12 35 Partial mortality of tagged corals and coral cover along transects from B1 to D12 36 Figure 4-17 Temperature, turbidity and PAR daily dose data (adjusted to -3 m LAT) from all coral monitoring sites between 1 August 2012 and 10 May Appendices Appendix A Type and Form of Coral Colonies Appendix B Statistical Analyses Appendix C Raw Data Appendix D Tagged Corals Condition: Supplementary Material Appendix E Water Quality Data Appendix F Quality Control Results Prepared for INPEX Cardno ix

10 1 Introduction 1.1 Background 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 Bladin Point on Middle Arm Peninsula in Darwin Harbour. To support the nearshore infrastructure at Bladin Point, dredging works is being carried out to extend safe shipping access from near East Arm Wharf to the new product loading facilities at Bladin Point, which is supported by piles driven into the sediment. A trench is also being dredged to seat and protect the GEP for the Darwin Harbour portion of its total length. Dredged material is being 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 East Arm (EA) Dredging and Spoil Disposal Management Plan (DSDMP) (INPEX 2013) and GEP DSDMP (INPEX 2014a). 1.2 Requirement to Monitor Corals Following an Environmental Impact Assessment (INPEX 2011), the Project was approved subject to conditions that included monitoring for potential effects of dredging or spoil disposal on local ecosystems, including corals. Sedimentation can cause stress in corals, as they may need to invest energy into the removal of sediment from their surface to prevent partial mortality and colony death (Cortes and Risk 1985). Turbidity and light attenuation stress corals by reducing the photosynthetic output of their zooxanthellae, thereby reducing the amount of autotrophic nutrition obtained by them (Phillipp and Fabricius 2003). Most corals obtain the majority of their nutrition autotrophically; hence, prolonged and elevated levels of turbidity and light attenuation may lead to mortality. Following cessation of dredging at the end of Season One, a review of the Nearshore Environmental Management Plan (NEMP), including the Coral Monitoring Program, was undertaken. The review resulted in a number of changes being incorporated in the Coral Monitoring Program. The Coral Monitoring Program was originally designed to specifically detect changes at Channel Island as the coral community at this site is listed on the Commonwealth Register of the National Estate and Northern Territory Heritage Register (AHPI 2012). Although sediment plume and sedimentation modelling did not indicate any potential for ecologically significant impacts at the Channel Island coral community (INPEX 2011), given the listed status of the area, it was determined appropriate by the Department of Natural Resources, Environment, the Arts and Sport (NRETAS, now NT EPA) to apply the precautionary principle and to treat Channel Island as a potential Impact site. For this reason, reactive triggers were developed to provide an early warning indicator of potential changes in coral health that would allow for early management actions for protecting the Channel Island coral community. Two other sites in Darwin Harbour with a similar assemblage of corals to Channel Island that were also not predicted to be affected by dredging (Weed Reef 1 and 2) were considered to be appropriate Control sites for monitoring in comparison with Channel Island in a Before After Control Impact (BACI) framework (Underwood 1992). During the review process, the appropriateness of the BACI design was reassessed. As a result, the classification of Channel Island as an impact site based on its listed status was identified as unsuitable given that no impact on coral health was predicted and similar turbidity influences were also predicted for Control sites at Weed Reef. To address this concern, the coral mortality reactive trigger was redesigned to include Weed Reef 1 and Weed Reef 2 as well as Channel Island and to detect potential changes to the rate of sitewide coral mortality above long-term trends for each site (i.e. above what would be expected). The revised NEMP outlines the updated methodology and parameters for the Coral Monitoring Program (Cardno 2013a). In addition to the changes to the design of the Coral Monitoring Program in terms of the reactive triggers, the EA DSDMP was revised to allow adaptive dredge management. To ensure that no increase in coral mortality occurs as a result of adaptive dredge management, the coral monitoring frequency during implementation of adaptive dredge management was increased from bimonthly to monthly. Prepared for INPEX Cardno 1

11 1.3 Results of the Previous Surveys Baseline Surveys (B1 to B3) Baseline sampling for the Coral Monitoring Program was undertaken between June 2012 and August 2012 at Channel Island, two sites in EA (Northeast Wickham Point and South Shell Island), two sites in Middle Harbour (Weed Reef 1 and 2) and two sites in Darwin Outer (Mandorah and Charles Point). At each of the monitoring sites, photos were obtained of 75 tagged coral colonies and at 1 m intervals along four fixed 50 m long transects. The objectives of the Baseline Phase were to: > Describe natural temporal and spatial changes in coral health indicators; and > Determine the statistical power of the monitoring design to detect specified levels of change to indicators of coral health, were they to occur as a consequence of dredging. Thirteen families of hard coral were identified, no more than ten of which occurred at any one site. Hard coral percentage cover ranged from 13% to 23% at most sites, apart from Northeast Wickham Point, which had approximately 2% cover in the three surveys (Cardno 2012a). Hard coral cover was generally constant throughout the Baseline Phase, whereas overall benthic assemblages differed among Baseline Phase surveys due to varying proportions of silt, sand and turf algae. Temporal differences were similar between Impact and Control sites. The most common families of hard corals were Dendrophylliidae, Faviidae, Pectiniidae and Poritidae. The most common morphology was encrusting and foliose, although massive and submassive forms were also present. At all sites, the majority of the substratum recorded was bare (between 25% and 60%, depending on the site and the survey) and was comprised of a sand or silt veneer covering or infilling consolidated material Season One Dredging Surveys (D1 to D4) and Dry Season Hiatus (D5 and D6) Season One dredging commenced on 27 August 2012 and continued through the wet season to 30 April Dredging surveys D1 to D4 occurred during this period and two surveys, D5 and D6, occurred during the dry season dredging hiatus from 1 May 2013 to 22 October There were no Level 1 turbidity exceedances (intensity, duration or frequency triggers) at Channel Island during Season One dredging or during the hiatus. Variability in the turbidity measured at Channel Island during Season One was generally similar to the variability observed during Environmental Impact Assessment monitoring in the 2010/2011 wet season. In contrast, turbidity at the Informative Impact sites, Northeast Wickham Point and South Shell Island, located close to the dredging operation within the EA of Darwin Harbour, at times showed elevated turbidity levels due to the dredging activities. During Season One dredging, site-wide mortality (measured from tagged coral colonies) increased at all monitoring sites, but at varying rates. Increasing partial mortality is an expected result of the tagged coral methodology, as total or partial mortality of a finite number of individual corals is inevitable, although the level of mortality is dependent on the length of monitoring. The spatial distribution of the range in mortality rates did not appear to relate to distance from dredging operations. That is, although the EA site South Shell Island has shown the highest rate of mortality among the five primary monitoring sites, the rate was similar at Weed Reef 2 and Channel Island, which are both over 7 km from EA dredging operations. The second EA site, Northeast Wickham Point, had a lower mortality rate than South Shell Island, Weed Reef 2 and Channel Island, and was comparable to the rate occurring at Weed Reef 1. Although sediment can potentially be removed by corals, as a precautionary measure, perceived mortality such as sediment on coral has been included in the estimates of mortality. As such, up to half of the measured mortality recorded was made up of sediment covering the tagged coral. The amount of sediment on coral has varied substantially at most sites between surveys making it difficult to determine if sediment cover was showing an overall increasing trend. The average cover of sand and silt along transects at Weed Reef 1 and 2 decreased between the Baseline and Dredging Phases, while no significant change occurred at Channel Island. The difference between Phases was due to a gradual decrease in sand and silt cover from the Baseline Phase through to D3 (February 2013), which then started to increase again from survey D4 (April 2013). Prepared for INPEX Cardno 2

12 In D3 (February 2013), there was substantial thermal bleaching of corals at Weed Reef 2, while a small increase in bleaching was also recorded at Weed Reef 1 and Channel Island. Given that there was no evidence that turbidity from dredging encroached upon Channel Island or Weed Reef 1 and 2, the increase in coral bleaching observed was assessed to be due to seasonally high water temperatures, which reached 32 º C. Further inspection of the bleaching data indicated that the localised bleaching at Weed Reef 2 occurred primarily in colonies of Alveopora sp., which have greater representation at Weed Reef 2 than at the other sites. This suggests that the bleaching response of this species may be more sensitive to increases in water temperature compared with other species of corals. Relative to the Baseline Phase, there was very little change in percentage cover of hard coral at most of the monitoring sites. However, at Weed Reef 2, there was a gradual (non-significant) but continual decline in hard coral cover from D2 (December 2012) to D6 (August 2013). This reduction in cover was most likely due to tentacle retraction of Alveopora spp., which bleached during warm water temperatures during the 2012/2013 wet season, and the subsequent substantial mortality of these species following bleaching. Analyses of hard coral families indicate a general temporal consistency in community assemblages and growth forms at all of the monitoring sites. This is important as the effects of dredging may not reduce the density of coral as a whole, but may cause changes in the dominance of certain coral families or growth forms due to varying sensitivity to the potential effects of dredging (i.e. turbidity and sedimentation). 1.4 Objectives There are two components to the Coral Monitoring Program: reactive and informative monitoring. Reactive monitoring is carried out at Channel Island, Weed Reef 1 and Weed Reef 2. The objective of the reactive component of the Coral Monitoring Program is to: > Protect Channel Island and Weed Reef coral communities by monitoring and implementing appropriate adaptive and/or contingency management measures for limiting potential impacts on corals as a result of dredging (where required). To allow for rapid assessment of potential impacts on the Channel Island and Weed Reef reactive monitoring sites, a series of water quality and coral health triggers are assessed (Table 1-1). The complete process from monitoring corals and assessing data for trigger exceedances to implementing management responses is described in the coral monitoring Trigger Action Response Plan (TARP). The TARP defines the water quality (turbidity) and coral health (net mortality and bleaching) triggers and describes the management response(s) required in the event of an exceedance in accordance with escalating risk to coral communities at Channel Island and Weed Reef. To improve understanding of the potential impacts of dredging on corals, informative monitoring is carried out on a routine basis at Northeast Wickham Point and South Shell Island, which are located within the EA of Darwin Harbour (Figure 2-1). Data collected on informative indicators will be used for interpretative purposes and may support management decisions using a multiple lines of evidence approach, particularly if reactive triggers are exceeded and require a management response. The objectives of the informative component of the Coral Monitoring Program are to: > Detect potential changes in coral health; and > Infer whether any potential changes are a result of dredging and spoil disposal activities. The reactive and informative programs both involve regular sampling of individually tagged coral colonies and repeat surveys of permanent transects to determine trends in the condition of the coral communities. These data are collected in parallel with informative water quality parameters. The suite of coral health and supporting water quality data being monitored will help determine whether or not dredging is having an impact on coral communities. Tagged coral colonies are used as an indicator of potential changes to coral health at a site and are useful to obtain very detailed information about changes in coral health and changes in mortality. Although tagged coral colonies are representative of the local coral community, they cannot provide suitable information about site-wide effects on coral health, as they only represent a small proportion of the total coral cover and the methodology does not take into account growth or recruitment. Additionally, mortality will generally only ever increase for a constrained selection of individual corals, as partial or total mortality of corals is inevitable given a long enough monitoring period. Given that tagged colonies provide Prepared for INPEX Cardno 3

13 an estimate of mortality for the individual corals selected only, and not for the entire coral community or reef (i.e. the change in the total number/cover of corals), coral health needs to be assessed also on a site-wide level. This is done using permanent transects. If the rate of mortality recorded for the tagged coral colonies is an unnaturally high rate (i.e. in excess of the replacement rate through recruitment and growth), coral mortality will be recorded in the fixed transects as a reduction in coral cover. This report details data collected during the twelfth routine coral health survey (D12; sampling dates 8 May 2014 to 10 May 2014) since the commencement of dredging on 27 August 2012 and covers the intervening period since Dredging survey 11 (D11) was completed (i.e. from 10 April 2014). This report is the sixth survey following recommencement of Project dredging for Season Two on 23 October Results are presented for reactive coral sites only (i.e. Channel Island, Weed Reef 1 and Weed Reef 2). East Arm sites South Shell Island and Northeast Wickham Point were not required to be surveyed in D12 in accordance with the monthly coral monitoring requirements of the EA DSDMP (INPEX 2013). Prepared for INPEX Cardno 4

14 Table 1-1 Channel Island coral trigger values (INPEX 2013, 2014a) Components Normal Situation Level 1 Trigger Level 2 Trigger Level 3 Trigger Intensity (95%ile) Daily Average Turbidity Duration (90%ile) Frequency (90%ile) Coral Bleaching Coral Mortality Coral Mortality Channel Island Trigger value (Wet Season) (1 November to 31 April) Trigger value (Dry Season) (1 May to 31 October) Not triggered >44 NTU >26 NTU over 7 consecutive days Not triggered >21 NTU >15 NTU over 5 consecutive days >26 NTU > 3 days per 7-day rolling period >15 NTU > 3 days per 7-day rolling period >20% gross coral bleaching Recorded mean partial mortality greater than upper 95% confidence limit of the predicted mortality For two consecutive surveys recorded mean partial mortality greater than upper 95% confidence limit of the predicted mortality Weed Reef 1 and Weed Reef 2 Trigger value (Wet Season) (1 November to 31 April) Trigger value (Dry Season) (1 May to 31 October) Not triggered >65 NTU >46 NTU over 6 consecutive days Not triggered >14 NTU >11 NTU over 4 consecutive days >46 NTU > 3 days per 7-day rolling period >11 NTU > 3 days per 7-day rolling period >20% gross coral bleaching Recorded mean partial mortality greater than upper 95% confidence limit of the predicted mortality For two consecutive surveys recorded mean partial mortality greater than upper 95% confidence limit of the predicted mortality N.B. Trigger levels 1 to 3 are associated with different management responses and the extent of intervention increases with the trigger level. Prepared for INPEX Cardno 5

15 2 Methodology 2.1 Overview To meet the objectives of the Coral Monitoring Program, the condition of corals is monitored using: > Tagged coral colonies; and > Fixed (permanent) transects. In each survey, divers take photographs of the tagged coral colonies and of the seabed along transects at the monitoring sites to assess and categorise the health (mortality, bleaching and sediment accumulation) of tagged corals and to describe the benthic composition along the transects (including family and growth form of corals and the number of coral recruits). A full description of the methodology is given in the Coral Monitoring Program Baseline Report (Cardno 2012a) and the Coral Monitoring Program Method Statement (Cardno 2012b). 2.2 Vessels, Diving, Safety and Environmental Management Field work was carried out from the diving vessel DSV Joshsarelle. Work was completed in accordance with the Project Health, Safety and Environment Plan. Diving was conducted using a combination of Self Contained Underwater Breathing Apparatus (SCUBA) (Australian Diver Accreditation Scheme (ADAS) Level AS ) and Surface Supply Breathing Apparatus (SSBA) (ADAS Level AS ) in accordance with Australia/New Zealand Standard Occupational Diving Operations Part 1: Standard Operational Practice (ASNZS :2007). Data were collected by scientific divers and site maintenance was completed by commercial divers from Neptune Diving Services. 2.3 Sites, Timing and Frequency of Surveys The locations of coral monitoring sites are given in Figure 2-1. A full description of the physical characteristics and coral cover at each site is provided in the Coral Monitoring Program Baseline Report (Cardno 2012a). Dredging survey 12 (D12) sampling was completed on neap tides from 8 May 2014 to 10 May During D12, sampling was completed at Channel Island, Weed Reef 1 and Weed Reef 2. The timing of coral surveys to date is presented in Table 2-1. A summary of data collected in D12 is given in Table 2-2. Prepared for INPEX Cardno 6

16 Lee Point Coral Monitoring Sites Charles Point 6!( 7!( Mandorah East Point Darwin ICHTHYS NEARSHORE ENVIRONMENTAL MONITORING Legend!( Coral Monitoring Site Gas Export Pipeline Dredging Footprint (Inset Map) Jetty 2A SP1 - Module Offloading Facility SP2 - Jetty Pocket Woods Inlet 4 5!(!( Wickham Point 2 3!(!( Blaydin Point SP3 - Jetty Pocket Berthing Area SP4 - Approach Channel SP5 - Walker Shoal Removal Plant Pad Community Filter feeders 1!( Hard coral Macroalgae Mixed reef Reef mixed communities <10% cover Sand communities Seagrass Mangroves Figure 2-1 Location of coral monitoring sites ID Name 1 Channel Island 2 Northeast Wickham Point 3 South Shell Island 4 Weed Reef 1 5 Weed Reef 2 6 Charles Point 7 Mandorah!( 2 FIGURE 1!( 3 Scale at A3 1:200,000 Kilometres Map Produced by Cardno NSW/ACT Pty Ltd (2812) Date: Coordinate System: GDA 1994 MGA Zone 52 Project: EL / Map: G_S041_001_CoralMonitoringSites.mxd 02

17 Table 2-1 Field sampling dates for coral monitoring in the Baseline and Dredging Phases Period Survey Sampling Dates Sites Surveyed Before Dredging During Season One Dredging Dry Season Dredging Hiatus During Season Two Dredging B1 16 June 2012 to 18 July 2012 CHI, NEW, SSI, WR1, WR2, CHP, MAN B2 27 July 2012 to 30 July 2012 CHI, NEW, SSI, WR1, WR2, CHP, MAN B3 11 August 2012 to 14 August 2012 CHI, NEW, SSI, WR1, WR2, CHP, MAN D1 22 October 2012 to 26 October 2012 CHI, WR1, WR2 D2 5 December 2012 to 9 December 2012 CHI, WR1, WR2, CHP, MAN D3 17 February 2013 to 22 February 2013 CHI, NEW, SSI, WR1, WR2 D4 17 April 2013 to 21 April 2013 CHI, NEW, SSI, WR1, WR2 D5 16 June 2013 to 20 June 2013 CHI, NEW, SSI, WR1, WR2 D6 14 August 2013 to 19 August 2013 CHI, NEW, SSI, WR1, WR2, CHP, MAN D7 26 October 2013 to 30 October 2013 CHI, NEW, SSI, WR1, WR2 D8 9 December 2013 to 14 December 2013 CHI, NEW, SSI, WR1, WR2 D9 23 February 2014 to 28 February 2014 CHI, NEW, SSI, WR1, WR2 D10 8 March 2014 to 14 March 2014 CHI, NEW, SSI, WR1, WR2, CHP, MAN D11 7 April 2014 to 10 April 2014 CHI, NEW, SSI, WR1, WR2 Relevant Report Cardno 2012a Cardno 2012a Cardno 2012a Cardno 2012c Cardno 2013b Cardno 2013c Cardno 2013d Cardno 2013e Cardno 2013f Cardno 2014a Cardno 2014b Cardno 2014c Cardno 2014d Cardno 2014e D12 8 May 2014 to 10 May 2014 CHI, WR1, WR2 This Report Table 2-2 Summary of the number of tagged colonies included in the mortality estimate (max = 75) and fixed transects sampled (max = 4) during D12 Site Name Tagged corals Dredging Survey D12 8 May 2014 to 10 May 2014 Channel Island 62 4 Weed Reef Transects Weed Reef Prepared for INPEX Cardno 8

18 2.4 Tagged Colonies Mortality and Bleaching Seventy-five coral colonies were tagged at each site (Appendix A-1). The number and type of tagged colonies are summarised in the Coral Monitoring Program Baseline Report, along with details of the equipment and methods used by divers to photograph each tagged coral (Cardno 2012a). The health of tagged corals was assessed by assigning several health categories, including healthy coral, dead coral, turf algae and sediment (Table 2-3). The images were assessed using Coral Point Count with Excel extensions (CPCe) (Kohler and Gill 2006), which involved visually assigning the appropriate category to 100 points (crosshairs), distributed on a 10 x 10 stratified random grid on the photos. It is noteworthy that turf algae cover on corals is considered to be actual mortality, as coral tissue beneath turf algae will rarely, if ever, recover, whereas sediment on corals is considered to be only potential mortality as the underlying tissue may be either alive or dead. When sediment is covering living coral tissue, the corals can actively clear the sediment away, revealing living tissue in subsequent surveys. However, sediment also covers nonliving coral tissue, including when the sediment remains unmoved for prolonged periods, resulting in death of the underlying tissue and also where sediment covers coral tissue that has died of other causes (e.g. turf algae). In this case, the coral s active sediment clearance will not be occurring. Given the uncertainty of the health of coral covered by sediment, in this program, mortality is assumed as a precautionary measure (i.e. worst-case mortality scenario). Table 2-3 Coral health categories for tagged coral CPCe analysis Main Category Actual mortality Potential or perceived mortality Live coral Sub-category Encrusting algae Turf algae Immobile fauna Dead coral (e.g. exposed coral skeleton with no live tissue) Missing section (e.g. broken segments) Macroalgae Mobile fauna Sediment Healthy coral Pale bleached White bleached Coral Mortality Trigger Assessment Using the observed rate of coral mortality (i.e. the site-wide trend) in tagged corals at the reactive sites (Channel Island, Weed Reef 1 and Weed Reef 2) during the ten surveys preceding commencement of Season Two dredging in EA (i.e. Baseline Phase surveys B1 to B3, and Dredging Phase surveys D1 to D7, undertaken from June 2012 to October 2013), it is possible to forecast levels of mortality in Season Two dredging. In this prediction, the response variable (the sum of the partial mortalities of tagged coral colonies) is based on the proportions (in each colony) of dead coral, measured as the number of points falling on dead coral (perceived and actual) over the total number of points falling on each colony. A Level 2 coral mortality trigger exceedance would occur if mortality during any survey in Season Two dredging was above the upper 95% confidence limit of the predicted mortality for Channel Island, Weed Reef 1 or Weed Reef 2. The number of missing observations (i.e. partial mortalities of individual colonies) during any survey influences the ability to forecast mortality. The model presented here does not attempt to adjust model predictions for the possible presence of mortality where there are missing observations. A model that provides a probabilistic approach for adjusting predicted site-specific mortality for potential mortality (i.e. adjusts for missing observations) will only be used if empirical evidence suggests that the mortality of missing colonies (as measured in these colonies prior to them becoming missing, that is, from measures of partial mortality in the preceding survey) is biased toward above average mortality. Prepared for INPEX Cardno 9

19 Site-specific time series model: not accounting for missing observations Using observed proportions of coral mortality (total dead/total coral count), a generalized linear mixed model (GLMM) was fitted for interpolating and forecasting global coral mortality for each site. The GLMM is an enhanced logistic regression model that considers the probability of coral mortality as a function of time. Below are several notable features of the model: 1. Each colony has its own intercept parameter as accomplished by a random effect on the intercept for each colony. By including a colony random effect, each colony is allowed to have its own mean mortality value (its own model intercept), while still informing overall site-level trends; 2. The temporal trend component is calculated using the dates of the observations (as opposed to survey number). The result is that the model accounts for irregularly timed samples (i.e. the model handles surveys separated by 70 days differently than surveys separated by 30 days); 3. Similar to the temporal trend component, the model accounts for temporal autocorrelation (the time series component) within each colony. It is assumed that observations across colonies are independent, while observations within a colony are correlated according to a continuous first order autoregressive (CAR(1)) process. This type of correlation structure is a generalisation of an AR(1) structure, but properly accounts for irregularly timed samples. That is, the correlation between adjacent observations within a colony depends on the timing between the observations; 4. Missing observations that disrupt the time series component of other models are handled effectively in this framework. Missing values for each colony can be predicted effectively using this model and standard errors are adjusted accordingly when observations are indeed missing; and 5. The model results in predictions at the individual colony and at the entire site level. Along with these predictions, prediction intervals are calculated and available and can be used to specify trigger values for future observations, given the information content in the data along with the model itself. Mathematically, the model is ( ) where, is the probability of coral mortality at location, on colony, at time (measured in days); is the global intercept parameter for location ; is the global temporal slope parameter for location ; ( ) is the random effect for individual colony, allowing that colony s predicted intercept to deviate from the global intercept for the site ( ); and is a normally distributed, mean 0 error term behaving according to a CAR(1) time series structure, as described above. That is, the correlation of two observations on the same colony depends on the time between them (say and ), i.e., cor( ). The output from the model will include the temporal trend in the site-wide tagged coral community and its prediction to the next coral survey Coral Bleaching Trigger Assessment The coral bleaching trigger assessment is based on the gross coral bleaching (bleaching above the Baseline Phase amount) at Channel Island, Weed Reef 1 and Weed Reef 2. In a given survey, a Level 2 trigger exceedance would occur if gross bleaching for any reactive site was greater than 20%. Gross coral bleaching will be estimated by summing the percentage (partial) bleaching of each colony and dividing by the number of colonies that were successfully analysed. For any site, this value will be expressed as a percentage relative to the Baseline Phase. > B Gross (gross bleaching per colony) = (B Survey B Baseline ); and > B Av (average gross bleaching) = B Gross / N, with N being the number of corals measured at the site. Prepared for INPEX Cardno 10

20 Assumptions and Limitations of the Coral Trigger Assessment Following a review of the Coral Monitoring Program at the end of Season One dredging (see Section 1.2), the trigger assessment protocol has been redesigned to provide an indicator of possible changes to tagged coral mortality rates in the coral communities at Channel Island, Weed Reef 1 and Weed Reef 2. Due to the short time span of the Baseline Phase surveys prior to dredging, there were inadequate data to reliably model a natural rate of coral mortality in the absence of potential dredging effects. Instead, the coral mortality was modelled based on the rates recorded during the ten surveys preceding commencement of Season Two dredging in EA (i.e. Baseline Phase surveys B1 to B3, and Dredging Phase surveys D1 to D7). This method assumes that Season One dredging did not influence the mortality of tagged corals at Channel Island and Weed Reef. Although this assumption is supported by the lack of a significant change in the coral community (as measured using fixed transects) and water quality data during Season One dredging, there remains the possibility that dredging could have had a small, but undetectable, effect on corals at these sites, which has been included in the GLMM predictions. A further limitation of the rate-based trend analysis relates to the lack of control or reference sites that are known to not be affected by dredging. This means there is the potential for a natural increase in the mortality rate, which in the absence of controls cannot be clearly distinguished from a dredging-related increase in mortality. Another main limitation of the tagged coral method arises from the fact that the substratum the corals attach to at most of the sites is unstable and, as the Coral Monitoring Program progresses, it is inevitable that some corals will naturally dislodge, roll or overturn from physical processes unrelated to dredging. Where corals are unable to be located (i.e. where the tag is found but the coral is missing), they are not included in the calculation of site-wide mortality, as to include them would mean the corals would need to be assumed dead based on incomplete information. The indeterminate status of these corals is due to one of the following scenarios: > The coral is healthy, upright and intact, but outside of the search range of the diver or difficult to locate due to poor visibility; or > The coral is partially or completely overturned, broken, buried or dead, making it difficult to locate or identify the coral. As the monitoring sites do not lie within or on the edge of the dredging footprint, corals that dislodge are considered to have done so because of natural processes rather than as a result of direct impacts from the dredge and, hence, are excluded from calculations of mortality. In addition, completely overturned corals (which are located and positively identified by divers) are not used in the mortality calculation as the available surface of the colony used to measure mortality is not comparable to the surface of the same colony analysed during the Baseline Phase. For partially overturned corals, analysis of coral health is generally undertaken on the remaining visible surface of the corals. Although this results in much less surface area being assessed in some cases, some health assessment is still possible and better than the alternative of excluding these corals entirely from the assessment process, which would reduce the statistical power. The disadvantage of including these corals is that the disturbance of dislodgement can lead to mortality in some instances. This mortality is obviously not an effect of dredging. In some instances, however, partially overturned corals are excluded from mortality assessments if the remaining visible surface area is too small for a reliable health assessment to be carried out. It is also possible that missing corals could in fact still be present in their original position but have become grown over by turf algae or buried by sediment. Although dead or buried corals are included where possible in the calculations of mortality, as it can be difficult to differentiate between missing or dead corals in some cases, this adds uncertainty to the mortality calculations. Discrimination between missing corals (i.e. those that have dislodged or rolled) and dead corals relies on a combination of field-based diver observations and office-based inspection of photographs for evidence of the pre-existing coral. Where a colony is missing, a photograph is taken of its position so it can be inspected for partially unburied or living fragments of a coral. In addition, data from previous surveys are used to support conclusions about the fate of corals (i.e. were corals recorded as partially buried or dead in previous surveys). Any coral that had been recorded as entirely dead or buried (i.e. 100% mortality) will continue to be included as part of the mortality trigger assessment, irrespective of whether or not it can be subsequently located. There are also limitations that arise from the ability of the assessment to incorporate growth of corals into the mortality calculation. As growth of tagged colonies occurs, the assessment area expands to include new Prepared for INPEX Cardno 11

21 growth. However, as the assessment of mortality includes the original extent of the coral colony (i.e. any dead area that was previously part of the colony), mortality will still be recorded even if the actual surface area has increased (since the Baseline Phase) to a size that is greater than the portion of the colony that has died Sediment Accumulation At each site, divers measured the depth of sediment on two 20 cm x 20 cm pavers previously placed next to two colonies from each of the families Faviidae, Dendrophylliidae, Pectiniidae and Poritidae, where colonies from these families had been tagged. Pavers were cleaned of sediment after each measurement Coral Colour and Area Colour and surface area analyses were undertaken from the photographed colonies as described by Cardno (2012a), with the exception of Goniopora spp. and Alveopora spp. colonies. The mean greyscale brightness (on the scale 0 = black, 255 = white) of about 1 cm 2 (4,675 pixels) of each colony was measured using the lasso tool in Adobe Photoshop CS3. The greyscale brightness was assigned to one of six colour brightness classes, as follows in order from darkest to brightest: 6 = 0 to 77.5; 5 = 77.5 to 122.5; 4 = to 170; 3 = 170 to 205; 2 = 205 to 228; and 1 = 228 to 255. Mean light and dark brightness values at each site were assessed graphically to evaluate whether colour brightness values were trending. Coral area analysis was not done in D8 to D11 due to unsuitability of photographs for measurements. The images were unsuitable as a wide-angle conversion lens was used to obtain images in low-visibility conditions. The wide-angle lens alters the image dimensions and does not allow for direct comparison of coral area with previous surveys. 2.5 Fixed Transects Percentage Composition At each site, benthic cover was estimated from sets of 50 photoquadrats of the seabed collected at approximately 1 m intervals along four 50 m permanent, lead-line transects. The photographs were taken of the same side of each transect and in the same direction as was done for the Baseline Phase surveys. The equipment used and dimensions of photoquadrats were the same as described in the Coral Monitoring Program Baseline Report (Cardno 2012a). For analysis of benthic cover, 40 photoquadrats were chosen randomly from the total of 50 for each transect. The percentage cover of organisms and substratum were quantified based on 64 points (5 mm circles) overlaid in an 8 x 8 stratified random grid on the photoquadrats using CPCe. Corals were identified to family level and their growth form assigned based on Veron (1986, 2000) and the Coral Finder (Kelley 2009). The coral form categories used were in accordance with the Australian Institute of Marine Science (AIMS) (2008), as described in the Coral Monitoring Program Baseline Report (Cardno 2012a) Recruitment (Juvenile Corals) Juvenile corals (<20 mm in diameter) were counted within the entire image for the 40 random photoquadrats (used for CPCe) to determine the total number of juveniles present in each transect. 2.6 Data Analysis Multivariate and Univariate Analyses of Variance Benthic assemblages, sediment (sand/silt) cover, hard coral cover and recruitment were all analysed using Repeated Measures Analyses of Variance with PERMANOVA+ software in PRIMER v6. The analyses tested the null hypothesis that there were no significant differences in these variables among Sites and Surveys (Table 2-4 and Table 2-5). Multivariate analyses undertaken for percentage cover of all biota and abiota categories, coral families and coral growth forms were based on a Bray-Curtis dissimilarity matrix. Vectors were superimposed on the non-metric multidimensional scaling ordination (nmds) plots, where significant interactions were found in order to graphically represent the variables that were most correlated with patterns in the multivariate dataset. The length and orientation of a vector shows the strength of the Prepared for INPEX Cardno 12

22 correlation between a variable and the two axes of the nmds plot. Thus, a vector indicates a gradient in the abundance of the variable that it represents. Multiple correlation was used to calculate the vectors. This takes of all the other variables into consideration when calculating the correlation coefficient for each variable. Univariate analyses of variance of percentage transect hard coral cover, total sediment (sand/silt) cover, and recruitment were undertaken based on Euclidean dissimilarity matrices (Anderson et al. 2008) using the same factors described above. The repeated measure term for these analyses was Transect within Sites. Where significant interactions or main factors effects were detected, post-hoc permutational t-tests using PERMANOVA+ were undertaken to identify the levels of factors in which differences occurred. No multiple test corrections were applied to t-test results, consistent with a conservative statistical approach and consistent with the Precautionary Principle. Table 2-4 Explanation of factors used in statistical analyses Component of Variation Survey Site Site x Survey Residual Interpretation Indicates a difference among surveys Indicates a difference among sites Indicates that variability at the scale of surveys differs among sites and vice versa A measure of the variation in the data not explained by the variation attributed to the main factors in the experimental model (i.e. surveys and sites and their associated interactions) Table 2-5 Outcome Terms used in describing the outcomes of statistical analyses Interpretation Redundant term, RED Non-significant, ns Asterisks A term becomes redundant if a lower order interaction including that term is significant Describes the convention by which a statistical comparison is deemed not to be an actual effect (i.e. accept the null hypothesis that there is no effect). Here, the cut-off point was set at p > This differs from the criterion used to decide whether to pool a factor, which is much more conservative These signify the probability (p) of an effect being considered to actually occur. Where: * indicates p 0.05; ** indicates p 0.01; and *** indicates p These specify that the likelihood of an effect occurring by chance alone (and therefore not explained by the factor being considered) would be 5 in 100, 1 in 100, or 1 in 1000, respectively Regression Analysis Regression analysis was used to test if there was a significant relationship between the sediment depth on pavers and sediment cover of colonies. If corals are not able to cope with potential increases in sedimentation rates, they would likely accumulate sediment cover at a rate proportional to increasing sediment depth on the tiles (i.e. a linear relationship, hence no transformation was applied). Regression analysis was undertaken using the Analysis ToolPack add-in for Microsoft Excel. 2.7 Physical Environment Temperature, sedimentation and light are major environmental drivers of coral health. Time series measurements of temperature, water turbidity and underwater light were therefore used to assist the interpretation of changes in monitored coral biological parameters. These environmental data were reported fortnightly for Water Quality monitoring sites at Channel Island, Weed Reef 1 and Weed Reef 2 over the period 11 April 2014 to 10 May 2014 (Cardno 2014f-h). The monitoring stations, established as part of the Water Quality and Subtidal Sedimentation Monitoring Program, were installed at approximately -4 m to -6 m LAT at a height of approximately 1.5 m above the seabed. Prepared for INPEX Cardno 13

23 2.8 Quality Control The Quality Control processes followed in the field and in the office by all Project personnel (i.e. divers, field staff and office staff) in order to complete the scope of work to a consistent and high quality are described in detail in the Coral Monitoring Method Statement (Cardno 2012b). Results of the checks of consistency and accuracy of the CPCe Analysts are given in Appendix F. Prepared for INPEX Cardno 14

24 3 Dredging Operations The dredging program involves a number of dredge vessels, including backhoe dredgers (BHDs), a cutter suction dredger (CSD) and trailing suction hopper dredgers (TSHDs), operating in different areas depending on water depths, bed material characteristics and the amount of material to be removed. The EA dredging campaign is divided into five Separable Portions (SP1 to SP5) that refer to the location within the dredge footprint and duration of specific dredging activities. The SPs are summarised in Table 3-1 and presented in Figure 2-1. Dredging for the GEP commenced on 23 October 2013 with the direct TSHD Queen of the Netherlands operating up until 28 November The Gas Export Pipeline (GEP) Backhoe Dredger (BHD) Simson commenced work on 7 March 2014 and continued operating along the GEP route during the reporting period. On 30 March 2014, a second BHD, Baldur, commenced operation along the GEP route and has been targeting the shoreline crossing since 2 April 2014 during spring high tides due to vessel draft restrictions. EA dredging operations recommenced on 1 November 2013 after being temporarily suspended in the six months since 30 April During the reporting period, the CSD dredger Athena was operating in SP4 loading two TSHDs. Direct TSHD operations were conducted in SP2 and SP3 up to 23 April 2014, at which point the Rotterdam departed Darwin Harbour. As of 11 May 2014, overall EA progress was approximately 92% of the proposed total to be removed during the dredging program. Table 3-1 ID SP1 SP2 SP3 SP4 SP5 East Arm dredge footprint summary Separable Portion Separable Portion 1 Module Offloading Facility (MOF) Separable Portion 2 Jetty Pocket Separable Portion 3 Berth Area Separable Portion 4 Approach Channel, Berth Approach and Turning Area Separable Portion 5 Walker Shoal Prepared for INPEX Cardno 15

25 4 Results 4.1 Tagged Colonies Mortality Trigger Assessment Analysis of the mortality of missing colonies indicated no effect of these data on the measurement of sitewide mortality for D12 (Appendix B-1). Thus, the GLMM model used to calculate the trigger values did not require adjustment for missing values. In D12, the measured mean partial mortality at Channel Island (28.2%) was close to the GLMM predicted mean partial mortality (29.1%), but substantially below the 41.1% trigger value (95% upper confidence limit (UCL) of GLMM predicted mean partial mortality) (Table 4-1, Figure 4-1). The measured mean partial mortality at Weed Reef 1 (13.9%) was below the predicted mean partial mortality (17.4%), and substantially below the 25.3% trigger value (95% UCL) (Table 4-1, Figure 4-1). The measured mean partial mortality at Weed Reef 2 (27.9%) was below the predicted mean partial mortality (37.5%), and substantially below the 49.9% trigger value (95% UCL) (Table 4-1, Figure 4-1). Consequently, the Level 2 triggers for coral mortality were not exceeded during the D12 survey. Table 4-1 Measured mortality, predicted mortality and trigger values for D12 from GLMM trend analysis using data from B1 to D7 Channel Island Weed Reef 1 Weed Reef 2 Measured mean partial mortality 28.2% 13.9% 27.9% Predicted mean partial mortality 29.1% 17.4% 37.5% 95% UCL (trigger value) 41.1% 25.3% 49.9% Prepared for INPEX Cardno 16

26 Coral Monitoring Report Dredging Report Channel Island 60 Weed Reef Partial Mortality (%) B1 B2 B3 D1 D2 D3 D4 Modelled mortality D5 D6 D8 D7 D10 D12 D11 D D3 D1 B2 B3 D2 B1 Weed Reef 2 D4 D5 D6 D8 D7 D12 D10 D9 D11 95% Confidence limit Recorded partial mortality included in model Recorded partial mortality excluded from model B1 B3 B2 D1 D2 D3 D4 D5 D6 D7 D8 D10 D12 D11 D9 0 Date Figure 4-1 GLMM trend analysis of mean coral mortality (red line) and 95% confidence limits (green lines) for surveys B1 to D7 (blue dots). Observed mean partial mortality (red dots) for D8 to D12 was not included in GLMM trend analysis Prepared for INPEX Cardno 17

27 Re-analysis of Mortality Trends with Season Two Dredging Surveys Included The effect of including the Season Two dredging surveys (D8 to D12) in the GLMM differed between the monitoring sites and depended on whether the mortality recorded in D8 to D12 was similar to the predicted trend or substantially above or below the predicted trend (Table 4-2, Figure 4-2). At Channel Island, the predicted partial mortality by the end of Season Two dredging using data up to D7 was 32.1%. This decreased to 31.6% when Season Two dredging surveys D8 to D12 were included in the model, and the 95% UCL decreased from 43.6% to 39.0%. Weed Reef 1 also showed a decrease in the predicted partial mortality by the end of Season Two dredging from 19.0% using data up to D7 to 16.1% using data up to D12, with the 95% UCL decreasing from 27.0% to 20.7%. A similar decrease in the partial mortality predicted for the end of Season Two dredging occurred at Weed Reef 2 from 41.4% to 33.7%, while the 95% UCL decreased from 52.7% to 41.3%. Table 4-2 Site Comparison of the predicted mean coral mortality and associated 95% UCL at the end of Season Two dredging based on GLMM trend analyses using data from B1 to D7 and data from B1 to D12 Surveys Included in the Trend Analysis Predicted Mortality by End of Season Two Dredging (31/07/2014) 95% UCL Channel Island B1 to D7 32.1% 43.6% B1 to D % 39.0% Weed Reef 1 B1 to D7 19.0% 27.0% B1 to D % 20.7% Weed Reef 2 B1 to D7 41.4% 52.7% B1 to D % 41.3% Prepared for INPEX Cardno 18

28 Coral Monitoring Report Dredging Report Channel Island 60 Weed Reef Partial Mortality (%) B1 B3 B2 D6 D5 D3 D4 D1 D2 Date Modelled mortality D8 D7 D10 D12 D9 D D3 D1 B2 D2 B1 B3 Weed Reef 2 D4 D5 D6 D8 D7 D12 D10 D9 D11 95% Confidence limit Recorded partial mortality included in model B2 D1 D3 D4 D5 D6 D8 D7 D10 D12 D9 D11 10 B1 B3 D2 0 Date Figure 4-2 GLMM trend analysis of mean coral mortality (red line) and 95% confidence limits (green lines) for all surveys to date (B1 to D12) Prepared for INPEX Cardno 19

29 Description of Partial Mortality Partial mortality of tagged corals increased at Channel Island from 26.9% in D11 to 28.2% in D12 and at Weed Reef 1 from 13.3% in D11 to 13.9% in D12 (Figure 4-3). The overall increases in partial mortality at Channel Island and Weed Reef 1 in D12 were due primarily to increases in immobile fauna (mainly sponges and hydroids obscuring the coral) of 0.9% and 0.6%, respectively, and from increases in sediment recorded on tagged corals of 0.5% for both sites. It is noted that the increases of immobile fauna and sediment at Channel Island and Weed Reef 1 were accompanied by decreases in turf algae of -0.6% and -0.3%, respectively. A slight overall decrease in partial mortality was recorded at Weed Reef 2 from 28.5% in D11 to 27.9% in D12 due to a decrease in immobile fauna (mainly sponges and hydroids) of -0.9% and to a decrease in sediment recorded on tagged corals of -1.8% which, together, were greater than the recorded increase in turf algae of 1.2% Throughout the monitoring program there has been a general increasing trend in partial mortality of the tagged coral colonies at all sites. The main contributors of partial mortality at all sites have been turf algae and sediment (Figure 4-3, Appendix C-1, Appendix D-1). At Channel Island, sediment on tagged corals increased from a minimum of 4.3% during B2 to 12.8% during D10, where it has remained at a similar level (D12: 11.8%). At Weed Reef 1 and Weed Reef 2, sediment on tagged corals has been more variable. At Weed Reef 1, sediment on tagged corals increased from 2.9% in B1 to a peak of 8.8% in D8, however, sediment on corals has since decreased to 6.6% (D12). At Weed Reef 2, sediment on coral increased from 5.9% in B3 to 15.8% in D7 with peaks in D1 (11.5%) and D10 (12.8%) and has since decreased to 8.7% (D12). Turf algae cover increased at Channel Island from 3.3% in survey B1 to a peak of 10.9% in D6, but has since declined to 7.4% in D12 (Figure 4-3). At Weed Reef 1, turf algae cover gradually increased from 2.2% in B1 to a peak in D3 (4.9%). Since D3, turf algae cover has remained at similar levels to D3, and was measured at 3.9% in D12. At Weed Reef 2, turf algae cover has increased from 3.5% in B1 to peak turf algae cover of 11.9% in D6, but has since decline to 6.7% in D12. It is worth noting that the apparent decline in turf algae after D6 and stabilisation of sediment on coral at Channel Island, and declines in turf algae and sediment at Weed Reef 2, are partly due to the natural complete mortality of several entire colonies, which for graphing purposes is displayed as entire colony mortality rather than the specific cause of mortality (i.e. turf algae or sediment on coral). The coral colonies that have reached 100% (i.e. complete) mortality during Dredging Phase surveys (Appendix D-1) showed various signs and symptoms of ill health prior to mortality increasing and, in some cases, these were likely to have contributed to their mortality. At Channel Island, the mortality of one of the corals followed a period of prolonged bleaching (Tag number 03). Prior to the bleaching, there were also borers present on this colony. A second colony at Channel Island died following the spread of tumours (Tag number 08). Tumours have only been recorded on one other colony during the monitoring program (Northeast Wickham Point; Tag number 69), and in this case almost 90% coral mortality has been reported (Appendix D-2). At Weed Reef 2, the mortality of two colonies was preceded by thermal bleaching (Tag numbers 45 and 46) and a further colony showed signs of growth anomalies and potential disease prior to dying in D7 (Tag number 26). Growth anomalies were more prevalent than tumours at the monitoring sites (including South Shell Island and Northeast Wickham Point), with 20 recorded occurrences, primarily present on Turbinaria sp. (18 of 20 occurrences). In the majority of cases, the growth anomalies were present since the Baseline Phase and the mortality associated with the presence of growth anomalies has been variable. Ten corals with growth anomalies had less than a 5% increase in mortality, while complete mortality occurred in one instance. Mortality increased in the rest by between 6% and 60%. Colonisation by turf algae was the ultimate outcome in all of the above cases. Colonisation by turf algae was observed on a further three colonies at Channel Island (Tag numbers 10, 61 and 64) and one colony each at Weed Reef 1 (Tag number 62) and Weed Reef 2 (Tag number 20). In these cases, however, there were no observable signs and symptoms of disease preceding the turf algae colonisation, although, in a few instances, a small patch of recently exposed skeleton was observed prior to turf algae colonisation. In the majority of these cases, turf algae were already present when monitoring commenced. However, the presence of turf algae is not necessarily a precursor to further ongoing mortality, with between 39% and 71% of the tagged colonies at each site having turf algae present in the Baseline Phase (Appendix D-2), most of which have not shown mortality to the extent of those mentioned above. One colony at Channel Island (Tag number 61) had minimal signs of ill health in D8, although there had been a borer present on this colony since the Baseline Phase. This colony showed 100% mortality in D9. Close examination of adjacent and nearby colonies Prepared for INPEX Cardno 20

30 revealed partial colony death in two of three colonies prior to the death of Colony 61. The decline in health was evidenced by localised bleaching and polyp death followed by turf algae colonisation. As Colony 61 adjoins one of these colonies, there may have been opportunity for disease to have been transferred. Although two of the colonies mentioned above (Channel Island tag numbers 03 and 61) had borers present prior to mortality, this was not necessarily the cause of mortality, as borers are prevalent in a large number of colonies throughout the monitoring sites, particularly on faviids. A test for an association between number of borers and increasing mortality showed no significant relationship (r 2 =0.001, p=0.705) (Appendix D-2), indicating that the presence of borers is not adversely affecting the health of the corals as measured by mortality. Two corals were buried by sediment in D12: one at Channel Island (Tag number 31) and one at Weed Reef 2 (Tag number 10). Channel Island coral colony 31 was recorded as completely buried for the first time in D3 while Weed Reef 2 coral colony 10 has been buried since D9. Interestingly, coral colony 2 at Weed Reef 2 had previously shown almost complete burial in surveys D4 and D9, then had cleared off sediment by the subsequent surveys D5 and D10, suggesting a pattern of accretion and erosion in this area (Appendix D-1) Channel Island Weed Reef Contribution to Partial Mortality (%) B1 Jun '12 B2 Jul '12 B3 Aug '12 D1 Oct '12 D2 Dec '12 D3 Feb '13 D4 Apr '13 D5 Jun '13 D6 Aug '13 D7 Oct '13 D8 Dec '13 D9 Feb '14 D10 Mar '14 D11 Apr '14 D12 May '14 Entire Colony Mortality Dead Survey Weed Reef Number of Colonies Included in Partial Mortality Estimate Missing Section Immobile Fauna Encrusting Algae Turf Algae Macroalgae Mobile Fauna Entire Colony Burial Sediment Number of Colonies (n) B1 Jun '12 B2 Jul '12 B3 Aug '12 D1 Oct '12 D2 Dec '12 D3 Feb '13 D4 Apr '13 D5 Jun '13 D6 Aug '13 D7 Oct '13 D8 Dec '13 D9 Feb '14 D10 Mar '14 D11 Apr '14 D12 May '14 Survey 25 0 Figure 4-3 Contribution (%) of each of the coral mortality categories to the mean mortality per coral colony at the monitoring sites in each survey and the remaining number of the original 75 tagged corals included in the mortality estimate (i.e. photograph obtained and of suitable quality) Prepared for INPEX Cardno 21

31 Sediment Deposition and Clearance Sediment accumulation on pavers was greater in the current survey (D12; May 2014) in comparison with D11 (April 2013) at Channel Island (D11: 2.7 ± 0.5 mm, D12: 8.1 ± 0.8) and Weed Reef 1 (D11: 2.3 ± 0.3 mm, D12: 5.0 ± 0.6) (Figure 4-4, Appendix C-2). Sediment accumulation on pavers was slightly less in D12 than D11 at Weed Reef 2 (D11: 4.6 ± 0.9 mm, D12: 3.4 ± 0.6). The average increases in sediment accumulation at Channel Island and Weed Reef 1 in the current survey corresponded to an average increase in sediment cover on tagged corals, and the average decrease in sediment accumulation at Weed Reef 2 corresponded to a decrease in sediment cover on tagged corals (Figure 4-5). However, comparison among surveys of the trends in accumulation of sediment on pavers and changes in cover of sediment on corals has shown no clear relationship; that is, there appears to be no clear link between increases in paver sediment depth and increased sediment on tagged corals (Figure 4-5). Furthermore, in D12, the percentage cover of sediment on the tagged corals adjacent to the pavers did not correlate with the depth of sediment on tiles (r 2 = 0.15, p = 0.41, Figure 4-5). The amount of sediment cover on tagged corals in D12 for all sites combined varied between different coral growth forms. Sediment cover was similar on encrusting corals (10.0 ± 1.9%) and submassive corals (13.3 ± 3.0%) while less sediment cover was recorded on foliose corals (6.4 ± 1.8%) and massive corals (6.2 ± 1.0%) (Figure 4-6). The cover of sediment for all the growth forms is similar to preceding surveys in which the lower-profile growth forms (i.e. encrusting and submassive) have generally had more sediment than the more raised forms (i.e. foliose and massive corals). 20 Channel Island Weed Reef Sediment Cover on Tagged Corals (%) Sediment Cover on Tagged Coral Paver Sediment Depth Weed Reef Paver Sediment Depth (mm) Figure 4-4 Sediment depth on pavers (mm) from survey D2 to D12 and sediment cover (%) on tagged corals from B1 to D12 Prepared for INPEX Cardno 22

32 Sediment Cover on Coral (%) Channel Island Weed Reef 1 Weed Reef 2 Figure Sediment Depth on Pavers (mm) Relationship between sediment cover on corals (%) and sediment depth on adjacent pavers (mm) during D12 Coral Sediment Cover (%) Encrusting Foliose Massive Submassive 0 Figure 4-6 Survey Average (±SE) cover (%) of sediment on different growth forms during B1 to D12 for all sites combined Prepared for INPEX Cardno 23

33 4.1.2 Bleaching Temporal Trends in Coral Bleaching In D12, bleaching at Channel Island and Weed Reef 1 increased slightly compared with the preceding survey (D11), whereas there was a decrease, albeit very small (i.e. 0.1%), in the mean bleaching at Weed Reef 2 (Figure 4-7). The increase in bleaching at Channel Island (from 1.4 ± 0.7% in D11 to 2.6 ± 1.5% in D12) and Weed Reef 1 (from 1.4 ± 0.8% in D11 to 2.6 ± 1.5% in D12) was primarily due to an increase in bleaching of coral colonies of the Goniopora genus. The increase was due to two colonies at Channel Island that showed bleaching in previous surveys and one colony at Weed Reef 1 that was 99% bleached in D12 without showing any bleaching in previous surveys. 15 Channel Island Weed Reef 1 Weed Reef 2 Mean Bleaching (%) B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 Survey D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Figure 4-7 Mean (± SE) partial bleaching (%) of tagged corals at each monitoring site for surveys B1 to D Surveys plotted on a continuous timeline based on number of days since the commencement of monitoring Trigger Assessment In D12, the mean partial bleaching was below levels recorded during the Baseline Phase surveys at all trigger sites except Weed Reef 2 (see Section ), resulting in negative gross bleaching at Channel Island and Weed Reef 1 and positive gross bleaching at Weed Reef 2. The mean gross partial bleaching values were -5.1%, -2.6% and 0.9% at Channel Island, Weed Reef 1 and Weed Reef 2, respectively (Table 4-3, Figure 4-8), which are well below the gross bleaching trigger value of 20%. Table 4-3 Recorded gross bleaching and trigger values for D12 Channel Island Weed Reef 1 Weed Reef 2 Gross partial bleaching (mean) -5.1% -2.6% 0.9% Trigger value 20.0% 20.0% 20.0% Prepared for INPEX Cardno 24

34 Coral Monitoring Report Dredging Report Channel Island Weed Reef 1 Weed Reef 2 Gross Bleaching (%) D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 Survey D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Figure 4-8 Gross (mean ± SE) coral bleaching for surveys D1 to D Surveys plotted on a continuous timeline based on the number of days since the commencement of monitoring Coral Colour The assessment of coral tissue brightness indicated that the majority of corals showed no change in the brightness of either the darkest or lightest tissue areas (Table 4-4). In survey D12, no coral colonies showed a change in brightness by two coral classes or more. However, over a longer timescale, between the Baseline Phase survey B3 and D12, six coral colonies showed a change in brightness by two coral classes or more. Two coral colonies showed an increase in brightness (lightening) and four colonies showed a decrease in brightness (darkening) of coral tissue (Table 4-4). When the changes in coral tissue brightness were assessed at a site-wide level (rather than individual colonies), there was minimal change in the average colour class in terms of either the lightest or darkest coral tissue (Figure 4-9). The dark coral tissue remained consistently close to coral brightness class 5, while the light coral tissue remained consistently close to coral class 4, with no surveys or sites showing a large deviation in the average coral class Colony Surface Area In D12, the change in surface area of corals since B3 (August 2012) ranged between -75% and 444%, with 14 out of 17 colonies showing an increase in measured area (Figure 4-10, Appendix C-3). In general, Faviidae and Pectiniidae were the coral families that showed more marked increases in area, whereas Dendrophyllidae corals showed smaller increases. The greatest increase in measured area was for colonies 07 and 66 at Weed Reef 2 (444% and 325%, respectively). The greatest increase at Channel Island was 83% (Colony 53) and the greatest increase at Weed Reef 1 was 241% (Colony 48). All of these colonies clearly showed visible growth from D7 to D12. The largest decrease in coral area was for Colony 60 at Weed Reef 2 (-75%), which was overgrown by a nearby coral. The second largest decrease (-30%) in coral area was for Colony 20 at Channel Island which showed a reduction in size due to the spread of turf algae. Change in coral surface area throughout the monitoring period has been highly variable. Given the long period of monitoring, a number of corals are now showing measured increases in area, which is generally corresponding with visually apparent growth. As noted previously, short-term variability among surveys is still apparent and is likely to be mostly related to slight differences in the angle and elevation at which the photos were taken rather than actual changes in the surface area of the colonies, but the substantial growth, or in some cases mortality, that has occurred is now apparent despite this variability. Prepared for INPEX Cardno 25

35 Table 4-4 Number of Classes Number of corals showing an increase in coral brightness (positive coral class change), decrease in coral brightness (negative coral class change), and no change in coral brightness between surveys D11 and D12 and between surveys B3 and D12 Change in Coral Class from Surveys D11 to D12 Change in Coral Class from Surveys B3 to D Dark Tissue Brightness CHI WR WR Total Light Tissue Brightness CHI WR WR Total a) Dark Coral Tissue 5 Coral Brightness Class Lighter- Darker b) Light Coral Tissue Channel Island Weed Reef 1 Weed Reef Survey Figure 4-9 Mean (± SE) coral brightness class for surveys B1 to D12 for: a) darkest coral tissue; and b) lightest coral tissue Prepared for INPEX Cardno 26

36 500% Channel Island 500% Weed Reef 1 400% % % 300% 200% 200% 100% 100% % Change in Coral Area 0% -100% B3 to D1 B3 to D2 B3 to D3 B3 to D4 B3 to D5 Survey B3 to D6 B3 to D7 B3 to D12 0% -100% 500% 400% Weed Reef % 200% 100% 0% -100% B3 to D1 B3 to D2 B3 to D3 B3 to B3 to D4 D5 Survey B3 to D6 B3 to D7 B3 to D12 Figure 4-10 Proportional change in coral area (%) during D1 to D12 relative to area recorded in B3 Coral area analysis was not done in D8 to D11 due to unsuitability of photographs for measurements. The key of each graph identifies the unique colony numbers, the families and growth forms of which are provided in Appendix A-1 Prepared for INPEX Cardno 27

37 4.2 Fixed Transects Composition of Sites (Multivariate Analyses) In D12, the main benthic categories at monitoring sites were sand and silt, turf algae and hard coral (Figure 4-11). At Channel Island, there was slightly greater cover of sand and silt and less cover of turf algae in surveys D6 and D8 to D12 than in previous surveys. At Weed Reef 1 there was an initial decrease in sand and silt from B1 through to D2 that corresponded to an increase in turf algae cover. This was followed by a slight increase in cover of sand and silt up to D4. From D4 to D11 the cover of sand and silt and turf algae has been stable; however, in D12, sand and silt decreased by 8.1% and turf algae increased by 8.8% compared to D11. At Weed Reef 2, there was greater cover of turf algae and less sand and silt in surveys B3 to D5 than for the other surveys. Although at Weed Reef 2 there was an apparent increase in sand and silt and decrease in turf algae between D9 and D11, sand and silt decreased by 9.9% and turf algae increased by 10.7% in D12. Overall, there has also been more variability between surveys in the cover of turf algae and in the cover of sand and silt at Weed Reef 1 and 2 compared to Channel Island (Figure 4-11). Hard coral cover has not changed through time at Channel Island or Weed Reef 1 but at Weed Reef 2 it showed a slight decrease in D3, and again in D6, after a bleaching event and has remained at a similar level. There were significant differences in the structure of the benthic assemblages between surveys and among sites as indicated by the significant interaction between Survey and Site (p 0.05); that is, the differences among sites were inconsistent between surveys and vice versa (Appendix B-2, Table 4-5, Figure 4-12a). Pairwise analysis comparing surveys at Channel Island indicated that benthic assemblages differed between some of the earlier and later surveys. There were, however, no differences between assemblages in the latest survey (D12) and all the previous surveys (Appendix B-2). Similarly, at Weed Reef 1 and Weed Reef 2 benthic assemblages differed between some of the earlier and later surveys but this was mainly due to assemblages in D2 at Weed Reef 1 and B3 at Weed Reef 2 which were different from assemblages on most of the other surveys (Appendix B-2, Figure 4-12a). Pairwise analysis comparing sites on each survey indicated that benthic assemblages differed between Weed Reef 1 and Weed Reef 2 in most of the cases (Appendix B-2). Importantly, the differences between surveys at these sites showed no clear patterns or trends indicative of an overall shift in the structure of the benthic assemblages. Furthermore, there were no patterns or trends indicating that any of the sites were becoming more or less similar through time, which would be indicative of a long-term shift in the composition of benthic assemblages (Figure 4-12c, Appendix B-4). The dissimilarities between surveys within each site were driven by differences in sand and silt, and turf algae and Pectiniidae (encrusting growth form) (Appendix B-2, Figure 4-12a). Table 4-5 Summary of multivariate PERMANOVA analyses of entire benthic composition and coral assemblages at the monitoring sites Significant factors indicated by *** p 0.001,** p 0.01,* p 0.05, NS = non-significant result, RED = Redundant term due to significant interaction Source Survey Site Survey x Site Entire benthic composition RED RED * Coral assemblage (family) RED RED * Coral assemblage (growth form) NS ** NS Statistical analysis of coral families, irrespective of their growth form, detected a significant interaction between Survey and Site (p < 0.05; Table 4-5, Figure 4-12b, Appendix B-3) indicating that differences among sites were inconsistent between surveys and vice versa. Pairwise analysis indicated that the only significant differences between surveys at each site were between B1 and D9 at Weed Reef 1. Despite a distinct grouping in the MDS plot (Figure 4-12b), pairwise analysis comparing sites during each survey indicated that the only significant differences were between Channel Island and Weed Reef 2 during survey D2 and D12 (Appendix B-3). These differences were mainly driven by Pectiniidae, Poritidae and Dendrophyillidae (Figure 4-12b). Analysis of coral growth form, irrespective of family, detected a significant difference between Sites (p < 0.01, Table 4-5, Figure 4-12c, Appendix B-4). Pairwise analysis comparing sites found that this was Prepared for INPEX Cardno 28

38 due to Channel Island being different from Weed Reef 1 and 2 (Appendix B-4). Encrusting and massive growth forms were the main drivers of these differences (Figure 4-12c). 100% Channel Island Weed Reef 1 100% 80% 80% Cover (%) 60% 40% 60% 40% 20% 20% 0% 0% Hard Coral Non-scleractinian Coral 100% SurveySoft Coral Ascidian 80% Sponge Weed Reef 2 Hard Coral Non-scleractinian Coral Soft Coral Other fauna Algae Rock&Rubble Cover (%) 60% 40% Ascidian Sponge Sand & Silt 20% Other fauna Algae 0% Rock&Rubble Sand & Silt 100% 20% 40% 60% 80% 0% Survey Figure 4-11 Contribution (%) of main categories to the composition of benthic assemblage along transects at the monitoring sites Prepared for INPEX Cardno 29

39 a) All categories b) Coral families c) Coral growth forms Figure 4-12 nmds ordinations based on: a) the entire benthic composition; b) coral families; and c) coral growth forms at the monitoring sites Prepared for INPEX Cardno 30

40 4.2.2 Percentage Cover Analyses (Univariate Analyses) Statistical analysis of hard coral did not detect any significant difference among surveys (p = 0.07) or sites (p= 0.99; Table 4-6, Figure 4-13a, Appendix B-5). The only evident pattern, although not statistically significant, is a slow decline in hard coral cover at Weed Reef 2 between D2 and D7 that corresponded to a bleaching event (Figure 4-13a). Statistical analysis of sand and silt cover showed significant differences among surveys independent of site (p < 0.001; Table 4-6, Appendix B-6). Pairwise analysis comparing surveys suggested that sand and silt cover was consistently greater during D6 to D11 (range: 53.0% to 60.3%) compared with earlier surveys B2 to D3 (range: 45.6% to 48.9%), with an apparent increase occurring between D3 and D6 (Appendix B-6). The cover of sand and silt in D12 was significantly lower compared to D11 but similar compared to D10 (Appendix B-6). Moreover, although the interaction between Survey and Site is not statistically significant, inspection of the trends for sand and silt cover at each site individually suggests that increases are more apparent at Channel Island, while Weed Reef 1 has shown minimal overall increase and Weed Reef 2 has shown periods of increased sand and silt cover in D6, D7 and D11 (Figure 4-13b). There were significant differences in the cover of turf algae between surveys (p < 0.01) and among sites (p < 0.01; that is, differences between sites were consistent among surveys and vice versa (Table 4-6)). Pairwise analysis comparing surveys found that the majority of significant differences detected were due to greater turf algae cover in D3 (32.9%) than during surveys D6 to D11 (range: 23.3% to 19.7%) (Appendix B-7). Turf algae cover in D12 was significantly higher than in D11 (7.6% increase) and was similar to all the previous surveys (Appendix B-7). Inspection of the trends at each site individually (Figure 4-13c) indicates that the trends in turf algae are the opposite of those apparent for the cover of sand and silt (Figure 4-13b). That is, there is a distinct relationship between the cover of sand and silt and turf algae, with surveys with greater sand and silt having less turf algae and vice versa (Pearson s correlation coefficient: r 2 = 0.74) (Appendix B-7). Pairwise analysis between sites irrespective of survey indicates that Weed Reef 2 had less cover of turf algae compared to Weed Reef 1 (Appendix B-7). Table 4-6 Summary of univariate PERMANOVA analyses of hard coral, sand and silt, and turf algae at the monitoring sites Significant factors indicated by *** p 0.001, ** p 0.01, NS = non-significant result Source Survey Site Survey x Site Hard coral cover NS NS NS Sand and silt *** NS NS Turf algae ** ** NS Prepared for INPEX Cardno 31

41 a) Hard coral 25 CHI WR1 WR b) Sand/silt Cover (%) c) Turf algae Survey Figure 4-13 Mean (±SE) hard coral, sand/silt and turf algae cover for all sites Prepared for INPEX Cardno 32

42 4.2.3 Recruitment (Juvenile Corals) A significant difference in the number of coral recruits among surveys (p < 0.05) was detected (Table 4-7, Appendix B-8, Appendix C-5), but there was no significant interaction between Survey and Site (p = 0.29); that is, differences between surveys were consistent among sites. Pairwise analysis between surveys indicated that the majority of the differences were due to significantly greater numbers of coral recruits recorded during the surveys B1 and D5 than some of the other surveys (Appendix B-8). The number of coral recruits in D12 was, however, similar to all the previous Baseline and Dredging Phase surveys (Appendix B-8). As confirmed by the non-significant interaction between Survey and Site, temporal patterns at the three monitoring sites were similar (Figure 4-14). Nonetheless, there was some variability among sites, although not statistically significant, during the monitoring period. Coral recruits were, for instance, more abundant at Channel Island than at Weed Reef 1 and Weed Reef 2 during D2 and more abundant at Weed Reef 2 than Channel Island and Weed Reef 1 during D4 and D9 (Figure 4-14). Table 4-7 Summary of univariate PERMANOVA analyses of the number of coral recruits for all sites Significant factors indicated by * p 0.05, NS = non-significant result Source Survey Site Survey x Site Number of coral recruits * NS NS 30 Channel Island Weed Reef 1 Weed Reef 2 25 Number of Recruits Survey Figure 4-14 Mean (±SE) number of coral recruits at all sites Prepared for INPEX Cardno 33

43 4.3 Comparative Analysis of Tagged Corals and Fixed Transects Sediment on Tagged Corals and Transects 1 It is necessary to compare changes in sediment cover on tagged corals with changes in sediment at a sitewide level to determine if sediment on tagged corals is representative of a site-wide trend or natural variability (Figure 4-15, see Section 4.2.2). While sediment on tagged corals has increased at Channel Island since the Baseline Phase, it has been more variable at Weed Reef 1 and Weed Reef 2 and in this survey (D12) was measured at a level that is comparable to the Baseline Phase (Figure 4-15, see Section 4.1.1). Similarly, sediment cover along transects at Channel Island has shown a pattern of increase through time but not so at the two Weed Reef sites (Figure 4-15, see Section 4.2.1). Sediment on tagged coral at Channel Island increased from an average of 5.3% during the Baseline Phase surveys to 11.8% in D12, with a distinct increase between D6 and D8, while transects at Channel Island showed a similar increase in sediment between D5 and D9, which has subsequently not increased further to D12. In contrast, sediment on tagged corals at Weed Reef 1 increased from an average 4.4% during the Baseline Phase surveys to 8.7% by D8 and decreased to 6.6% in D12, whereas sediment cover in transects at this site decreased from B1 to D2 before increasing through to D11 and decreasing again in D12. At Weed Reef 2, sediment on tagged corals showed minimal increase from B1 to D5, excluding a temporary spike in sediment on coral in D1, before increasing from 7.0% in D5 to 15.8% in D7. Since D7, sediment on coral has declined to 8.7% in the current survey (D12). Although no overall increase in sediment along transects has occurred at Weed Reef 2, there is an overall similarity in the pattern of increases and decreases to sediment on tagged corals, particularly an increase in sediment cover between D4 and D7, followed by a decrease through to D Mortality of Tagged Corals and Transect Coral Cover As a consequence of mortality being measured using a permanently tagged subset of corals, partial mortality of tagged corals has generally been increasing during the monitoring period at all sites but at differing rates, as expected (Figure 4-16, see Section 4.1.1). For this reason, it is necessary to interpret coral mortality on a site-wide level in addition to that observed for individual tagged coral colonies. If the mortality of individual corals was occurring at a rate that was greater than replacement by growth and recruitment, it would be expected that site-wide coral cover would also be declining. There has been no apparent change in coral cover at any of the monitoring sites, with the exception of Weed Reef 2, which showed a decreasing trend up until D7 that is considered to be primarily due to coral mortality following the thermal bleaching event recorded in D3 (February 2013) (Figure 4-16, see Section 4.2.2). Since D7, coral at Weed Reef 2 has remained relatively stable. 1 Sediment on transects is the percentage cover of sediment along the transect strip (i.e. the cover of sediment on the substratum and overlaying biota). For transect photos, each point analysed is assigned as sediment if sand or silt occupies more than 50% of the 5 mm circle that constitutes the point. Prepared for INPEX Cardno 34

44 30 Channel Island Weed Reef Sediment Cover on Corals (%) 10 0 Sediment on Tagged Corals Total Sediment on Transects Weed Reef Total Sediment Cover on Transects (%) Figure 4-15 Sediment on tagged corals and total sediment cover along transects from B1 to D12 Prepared for INPEX Cardno 35

45 40 Channel Island Weed Reef Tagged Coral Partial Mortality (%) 10 0 Tagged Coral Partial Mortality Transect Coral Cover Weed Reef Transect Coral Cover (%) Figure 4-16 Partial mortality of tagged corals and coral cover along transects from B1 to D12 Prepared for INPEX Cardno 36

46 4.4 Physical Environment The winds during the current reporting period (D11 to D12; 11 April 2014 to 10 May 2014) were typically moderate to fresh (in the range of 15 to 30 km/hr), with short periods of increased wind speeds of up to 35 km/hr and gusts of up to 60 km/hr on 19 April 2014 and 50 km/h on 5 May 2014 associated with Tropical Cyclone Gillian (Cardno 2014f-h). Cumulative rainfall from 1 April 2014 to 30 April 2014 was mm, which is more than the April monthly average of mm. Cumulative rainfall from 1 May 2014 to 11 May 2014 was 3.2 mm, which is less than the May monthly average of 21.3 mm (Cardno 2014h). Modelled significant wave height during the current reporting period was generally less than 0.5 m and there were short periods of moderately elevated wave height in the range of 0.5 m to 0.8 m from 19 April 2014 to 21 April 2014 and from 4 May 2014 to 6 May 2014 as a result of the elevated winds during this period. Mean daily-averaged water temperature in the current reporting period varied between 30.5 º C and 30.6 º C among the monitoring sites (D11 to D12) (Table 4-8, Appendix E-1). The water temperature decreased through the reporting period from a maximum daily-averaged temperature of 31.2 º C at the beginning of the monitoring period (11 April 2014) to a minimum of 29.5 º C by the end of the period (10 May 2014) (Table 4-8, Figure 4-17, Appendix E-1). The National Oceanic and Atmospheric Administration's (NOAA) Coral Reef Watch Program (CRW) monitors sea surface temperature (SST) in near-real-time by satellite to assess the risk of thermal coral bleaching (NOAA Coral Reef Watch 2014). This risk is based on degree heating weeks, a cumulative measurement of the intensity and duration of thermal stress expressed in the unit C-weeks. The risk of bleaching is then assessed and categorised into one of five bleaching alert levels: No Stress, Watch, Warning, Alert Level 1 and Alert Level 2. The alert level in the waters around Darwin was at the alert level Watch from the beginning of the monitoring period (11 April 2014) until 7 May 2014 and then decreased to the alert level No Stress. The mean daily-averaged turbidity (based on 15-minute logged data) for the current reporting period (D11 to D12) was greatest for Weed Reef 1 (7.1 NTU). Channel Island and Weed Reef 2 had lower turbidity values at 6.2 NTU and 6.0 NTU, respectively (Table 4-8, Appendix E-2). The maximum daily-averaged turbidity for any of the monitoring sites was 15.2 NTU, which was recorded at Channel Island. Weed Reef 1 and Weed Reef 2 recorded similar maximum daily-averaged turbidity values of 13.5 NTU and 12.7 NTU, respectively. Overall, turbidity was generally lower in the current period than the preceding period, in terms of both mean and maximum daily averages (Table 4-8). Photosynthetically active radiation (PAR) was variable among the monitoring sites. The greatest PAR peak flux recorded in this reporting period was at Weed Reef 2 (mean of µmol/m 2 /s), followed by Weed Reef 1 and Channel Island (means of mol/m 2 /day and 88.0 µmol/m 2 /s, respectively) (Table 4-8, Appendix E-3). PAR daily dose levels (depth adjusted to -3 m LAT) closely resembled the trends in PAR peak flux, with a maximum daily dose of 5.58 mol/m 2 /day at Weed Reef 2, while Weed Reef 1 and Channel Island recorded maximum daily doses of 3.99 mol/m 2 /day and 2.10 mol/m 2 /day, respectively (Table 4-8, Appendix E-4). Prepared for INPEX Cardno 37

47 Table 4-8 Summary of daily-averaged turbidity, daily-averaged water temperature, PAR peak flux and PAR daily dose (mean, minimum and maximum) at monitoring stations during the current reporting period, previous reporting period and the period corresponding to the current reporting period from the previous year Site Current Period Mean Min Max Previous Period Mean Min Max Corresponding Period Mean Min Max Water Temperature ( 0 C) CHI 11/04/14 05/05/ /03/14 10/04/ /04/13 10/05/ WR1 11/04/14 08/05/ /03/14 10/04/ /04/13 10/05/ WR2 11/04/14 07/05/ /03/14 10/04/ /04/13 10/05/ Turbidity (NTU) CHI 11/04/14 27/04/ /03/14 10/04/ /04/13 10/05/ WR1 11/04/14 08/05/ /03/14 10/04/ /04/13 10/05/ WR2 11/04/14 07/05/ /03/14 10/04/ /04/13 10/05/ PAR Peak Flux (µmol/ m 2 /s, adjusted to -3m LAT) CHI 11/04/14 27/04/ /03/14 10/04/ /04/13 10/05/ WR1 11/04/14 08/05/ /03/14 10/04/ /04/13 10/05/ WR2 11/04/14 07/05/ /03/14 10/04/ /04/13 10/05/ PAR Daily Dose (mol/m 2 /d, adjusted to -3m LAT) CHI 11/04/14 27/04/ /03/14 10/04/ /04/13 10/05/ WR1 11/04/14 08/05/ /03/14 10/04/ /04/13 10/05/ WR2 11/04/14 07/05/ /03/14 10/04/ /04/13 10/05/ Prepared for INPEX Cardno 38

48 CHI WR1 WR2 CHI WR1 WR2 CHI WR1 WR2 Figure 4-17 Temperature, turbidity and PAR daily dose data (adjusted to -3 m LAT) from all coral monitoring sites between 1 August 2012 and 10 May 2014 Prepared for INPEX Cardno 39

49 5 Discussion 5.1 Trigger Assessment Level 1 Water Quality During the reporting period (11 April 2014 to 10 May 2014), Weed Reef 1 experienced dry season Level 1 turbidity Duration and Frequency trigger exceedances when daily average turbidity marginally exceeded 11.0 NTU for four consecutive days on 4 May 2014 (Cardno 2014h). The INPEX Exceedance Attributability and Implementation Report considered these trigger exceedances to be primary attributable to natural causes (INPEX 2014b). There were no wet or dry season trigger exceedances of turbidity threshold at Channel Island or Weed Reef 2 (Cardno 2014f-h) Coral Mortality The coral mortality trigger assessment is based on the rate of coral mortality observed in tagged corals at the reactive sites (Channel Island, Weed Reef 1 and Weed Reef 2) during the ten surveys preceding commencement of Season Two dredging in EA. The Level 2 trigger for the EA and GEP DSDMPs would be exceeded if the measured mortality at any of the reactive sites was above the upper 95% confidence limit of the predicted mortality (INPEX 2013, 2014a). In D12, no Level 2 coral mortality triggers occurred, as the measured mortality at Channel Island, Weed Reef 1 and Weed Reef 2 was below the upper 95% confidence interval of the predicted mortality Coral Bleaching The coral bleaching trigger assessment is based on the gross coral bleaching (bleaching above the Baseline Phase amount) at each of the reactive sites. In a given survey, the Level 2 trigger would be exceeded if gross bleaching for any reactive site was greater than 20%. In D12, no Level 2 coral bleaching triggers occurred. Measured bleaching at Channel Island and Weed Reef 1 was below the average bleaching of the Baseline Phase surveys, while Weed Reef 2 was slightly above. Gross bleaching since the Baseline Phase was -5.1%, -2.6% and 0.9% at Channel Island, Weed Reef 1 and Weed Reef 2, respectively. 5.2 Interpretation and Attributability As well as the key indicators of coral health triggers (tagged coral mortality and bleaching), additional informative coral metrics (i.e. coral colour, community composition, hard coral cover, cover of silt/sand and coral recruitment) and water quality parameters (i.e. turbidity, water temperature, light and sediment accumulation) were measured to provide a multiple lines of evidence assessment of the health of corals at the monitoring sites Comparison of Mortality Rates Tagged coral mortality between D11 and D12 increased slightly at Channel Island (1.3%) and Weed Reef 1 (0.6%), while mortality decreased at Weed Reef 2 (-0.6%). The measured mortality at Channel Island in D8 to D12 was very similar to the predicted mortality, suggesting that coral mortality has been increasing in a manner consistent with the GLMM predictions. At Weed Reef 1 the measured mortality was above that predicted in D8 but below the predicted values in D9 to D12, which is consistent with the large degree of variability in the measured mortality at this site. Mortality levels at Channel Island and Weed Reef 1 are within the 95% confidence limits of predictions, indicating that measurements are within the uncertainty associated with the model as a consequence of the inherent variability in coral mortality. At Weed Reef 2, the measured mortality in D8 to D12 was less than the predicted mortality and mortality in D12 was below the lower 95% confidence limit of predictions. After the effects of the thermal bleaching event in Season One dredging, there has been effectively no additional mortality, on average, at Weed Reef 2 since D6. As the mortality at the reactive sites in D8 to D12 was lower than had been predicted for almost all surveys, inclusion of these surveys in the (GLMM) trend analysis has the effect of reducing the predicted mortality for Prepared for INPEX Cardno 40

50 the end of the dredging period (31 July 2014) at all of the reactive sites. At Channel Island, despite the measured mortality being similar to predicted mortality, the inclusion of measured D8 to D12 data causes predicted mortality to decrease from 32.1% to 31.6% by the end of the dredging period. The forecasted mortality decreases at Weed Reef 1 from 19.0% to 16.1% and at Weed Reef 2 from 41.4% to 33.7%. Overall, the reduction of the upper 95% confidence intervals was between 4.6% and 11.4% at the trigger sites by 31 July It is not considered necessary at this stage to adjust the trigger predictions for subsequent surveys Coral Mortality Increasing mortality is an expected result of the tagged coral methodology, as total or partial mortality of a finite number of individual corals is inevitable, although the amount of mortality recorded is dependent on the length of the monitoring program and the natural rate of mortality. The naturally low cover of corals at the monitoring sites (approximately 15% or less during the Baseline Phase), as well as the generally small size of the corals (potentially indicative of young coral colonies), suggests that Darwin Harbour coral species may experience a naturally high rate of mortality. However, it is also possible that some species are slow growing, due to the marginal conditions for coral, which may inhibit rapid colony growth. The results of the tagged coral monitoring suggest that both strategies exist in Darwin Harbour for different coral species. In fact, some types of coral have grown rapidly during the monitoring program, such as Mycedium sp., which has also shown some rapid mortality during the monitoring program. Conversely, other types of coral, such as faviids, appear to be less susceptible to mortality, and in many cases have also shown no visibly apparent increase in size. A primary component of the partial mortality of tagged corals is the growth of turf algae. Given the general increase in mortality of tagged corals in the monitoring program, the presence of turf algae upon colonies is expected to increase as turf algae colonises the dead parts of coral. There has been less variability in the presence of turf algae upon colonies between surveys compared with sediment, as would be expected as areas colonised by turf algae generally do not recover, and in the rare instances where they do, they rely on the coral colony regrowing over the dead area. Although there have been some decreases in turf algae cover at some sites, these often coincide with increases in sediment, and vice versa. This suggests that some of the previous areas of turf algae may have been covered by sediment, increasing the measured levels of sediment on corals. Coral growth, along with recruitment, should theoretically compensate for mortality of other coral colonies, so that at a site level there is no long-term net reduction in coral cover. The analyses of coral surface area undertaken do not allow for an accurate estimate of coral growth. The analysis of coral surface area has previously been reported as being unreliable as an indicator of growth due to the limitations of the current techniques (Veal et al. 2010). However, given the extended period of time over which the measurements have occurred, the results are proving to be better aligned with the observed changes of each of the measured corals. This is likely to be due to many of the corals having changed to such an extent that previous measurement errors are less than the growth (or reduction in size) that has actually occurred. However, as only a small number of corals from each site are being measured, due to limited numbers of corals suitable for measurement (small, flat and on level substrate), it is difficult to say anything meaningful about site-wide changes in coral colony size. Given that the effects of dredging are most likely to be dispersed widely at a given site, it would be expected that any effects measured (i.e. reductions in coral cover) would be site-wide. Therefore, in instances of isolated complete mortality of a few individual colonies, assessment of potential causes for the individual mortalities is informative. At Channel Island, the mortality of one of the corals followed a period of prolonged bleaching and another died following the spread of disease (tumours). At Weed Reef 2, the mortality of two colonies was preceded by thermal bleaching and one colony showed signs of tumours/disease prior to complete mortality. Colonisation by turf algae was the ultimate outcome in all of the above cases, which reflects the fact that turf algae is often a rapid coloniser following stress or other causes of tissue mortality (McCook et al. 2001). Colonisation by turf algae was observed on a further three colonies at Channel Island and one colony at Weed Reef 1 and Weed Reef 2. In most of these cases, however, there were no observable symptoms preceding the turf algae colonisation as it was already present at the commencement of monitoring (Baseline Phase). One colony at Channel Island (Colony 61) went from no observable signs of ill health and no turf algae in D8 to complete mortality in D9. The decline in health was evidenced by Prepared for INPEX Cardno 41

51 localised bleaching and polyp death followed by turf algae colonisation. Close examination of adjacent and nearby colonies revealed partial colony death in two of three colonies prior to the death of Colony 61. As Colony 61 adjoins one of these colonies, there may have been opportunity for disease to have been transferred. It is unclear whether turf algae itself can cause mortality of coral tissue as there is limited documented evidence of this occurring (McCook et al. 2001). However, once turf algae is established on a coral, it is likely to cause stress and tissue mortality and expand across the colony. One mechanism for this spread is enhanced microbial activity at the coral algal interface due to chemical compounds released by the algae (Smith et al. 2006). While tagged colony monitoring provides highly accurate data (repeated measures of the same corals) on the colonies selected for monitoring, it has some limitations in terms of its ability to represent site-wide effects that may be occurring. The tagged corals are not an accurate indicator of the net site-wide coral cover; that is, whether there is an overall increase or decrease in the number or area of corals. In addition, only the health of hard corals is assessed using this technique, not the entire benthic community as a whole. For this reason, the use of photoquadrat analyses along fixed transects is included in the monitoring program to provide a robust measure of the entire community condition. Several different measures are obtained from this dataset, allowing for the analysis of several components of community health. In the current dredging survey (D12), analysis of entire community composition, which includes both biota (hard coral families and their growth forms, soft corals, algae and sessile invertebrates) and abiota (sand, silt and rock), showed that differences between surveys were dependent on site and vice versa. The differences between surveys were driven by changes in the cover of sand and silt and the cover of turf algae. Further analysis of the cover of sand and silt indicated a decrease in D12 since D11 at all sites. Sand and silt cover in D12 was similar to Baseline Phase levels, suggesting that the difference in sand and silt could be part of a longer-term cycle of increases and decreases. It is also important to note that increases in sand and silt appear to be strongly correlated with decreases in turf algae and vice versa. Although total coral cover for all sites combined did not show a significant decrease through time at any site or for all sites combined, there has been a slight decrease in coral cover at Weed Reef 2. At this site, coral cover decreased from 16.8% in B1 to 11.7% in D6, initially as a result of tentacle retraction of Alveopora spp. due to thermal bleaching in the 2012/2013 wet season, followed by a reduction in coral cover primarily due to substantial mortality of many of the bleached colonies. Although the above analysis of the entire community composition includes coral families and growth forms, the analysis was also done by including only hard coral families (irrespective of growth form), and again by including only growth forms (irrespective of family), which allows a more targeted analysis for potential changes in coral family composition or growth forms. This is important as the effects of dredging may not reduce the density of coral as a whole, but may cause changes in the dominance of certain coral families or growth forms due to varying sensitivity to the potential effects of dredging (i.e. turbidity and sedimentation). Despite analysis of the entire community composition and coral families showing a significant interaction between surveys and sites, there were no trends or patterns of difference indicative of a consistent shift in the composition of benthic assemblages or families through time. The analysis of hard coral growth forms indicated that, although there were small differences in the composition of coral growth forms among sites, there have been no significant shifts in the coral growth form composition at any of the reactive monitoring sites, with the sites generally showing a large degree of consistency through time Sediment on Coral Sediment on coral, which constitutes up to half of the partial mortality of tagged corals recorded, has generally increased through time at Channel Island. At Weed Reef 1 and Weed Reef 2 sediment on coral has been more variable, and in this survey (D12) was measured at a level that is comparable to the Baseline Phase. It is worth noting that the apparent decline in sediment on corals (and turfing algae) at Weed Reef 2 after D7 are partly due to the mortality of several entire colonies. When the sediment on coral and turfing algae of these completely dead colonies is included, it is clear that sediment on coral and turfing algae at Weed Reef 2 has also generally increased through time. Increasing sediment on tagged corals (and turfing algae) through time is expected, similar to the expected increase in mortality, as areas of dead coral become buried with no active removal of sediment occurring. The variability in sediment on coral and turf algae at Weed Reef 1, however, likely reflects the minimal amount of actual mortality at this site. A pattern of shortterm variability though time in sediment on tagged corals has occurred at all of the reactive sites underlying Prepared for INPEX Cardno 42

52 the longer-term trend. Often, this variability in sediment translates to variability in the overall mortality trend, indicating that sediment is settling on living coral, which is then both actively and passively cleared by the coral before subsequent surveys. However, in some instances, an increase in sediment does not translate to an increase in mortality or variability in the overall trend, which is likely the result of other areas of mortality, such as turf algae, becoming covered in sediment. In the current survey (D12), a greater amount of sediment accumulated on pavers at Channel Island and Weed Reef 1 compared with the previous survey. Sediment recorded on tagged corals at Channel Island and Weed Reef 1 also increased in the current survey; however, when all surveys are considered, the deposition of sediment on the settlement pavers at each site does not correlate with sediment cover on the corals. This indicates that the amount of sediment on the corals is not generally a direct result of sediment deposition processes and that the corals can actively remove sediment. As discussed above, sediment on tagged coral colonies is expected to increase through time due to its association with coral mortality but it is also possible that the increased sediment on coral is a temporary event, and that the corals may clear some or all of this increased sediment prior to the next survey. Corals have the ability to remove sediment through the excretion of mucus from individual polyps (Erftemeijer et al. 2012). The rate that corals can clear the sediment is highly variable between species and can be influenced by factors such as sediment load, growth form, stress and the general health of the coral colony (Lirman and Manzello 2009; Erftemeijer et al. 2012). In general, the lower profile encrusting and submassive growth forms at sites have the greatest cover of sediment. It is likely that the amount of sediment cover on the tagged corals is the result of a combination of both physical and biological processes. However, all corals will have a limit in terms of the amount of sediment they can successfully remove and, where sedimentation is above this rate, coral burial is likely to occur. As a consequence of sediment on coral being measured for a permanently tagged subset of corals, partial or complete burial is expected for at least some corals over time. This is particularly the case in Darwin Harbour given the large tidal movements, which also move large amounts of sediment. Additionally, many of the corals at the monitoring sites, including the tagged corals, are growing directly on the unconsolidated seabed and are not elevated on reefs. As such, localised movement of sediment due to currents will inevitably cause some corals to become buried, particularly low-growing encrusting and submassive corals. For example, one coral at Weed Reef 2 (Tag Number 2) underwent complete or almost complete burial then subsequently became unburied on two occasions with no apparent changes to nearby tagged corals, which demonstrates the large degree of localised sediment transport occurring. Although sediment deposition is often temporary, and is subsequently removed by active and passive processes, some corals may become permanently buried, and may suffer mortality as a consequence. However, if there is no overall change to the sedimentary regime, there should be no long-term increase in sediment or decrease in coral cover at a site-wide level. When considered at a site-wide level, analysis of the cover of sand and silt indicated a slight increase in surveys D9 to D11 for all sites combined, regardless of whether South Shell Island and Northeast Wickham Point were included in the analysis (see Cardno 2014e). This trend was followed by a sharp decrease in D12 at Weed Reef 1 and Weed Reef 2 and, to a lesser extent at Channel Island. Although the statistical analysis found a significant difference for all sites combined, inspection of the trends at individual sites suggests that the trends vary among the sites. It is important to note that the significant difference between surveys when all sites are combined could be a result of the increased power at this level, rather than an indicator that the trends are consistent between the sites. Although Channel Island has shown increases consistent with the overall trend, Weed Reef 1 and Weed Reef 2 have shown more temporal variability, with both decreases and increases in sediment through time, and no clear increasing trend. It is also important to note that increases in sand and silt appear to be correlated with decreases in turf algae, rather than with decreases in coral cover. Also of importance is that the increases in cover of sediment observed on tagged corals and pavers in D12 were not reflected at a site-wide level Bleaching In D12, bleaching increased at Channel Island and Weed Reef 1 and decreased slightly at Weed Reef 2 compared with the preceding survey (D11). These increases were due to coral colonies of the Goniopora genus. In particular, the increase at Weed Reef 1 was due to a single colony that bleached almost completely in D12 but showed only minor signs of bleaching in the previous surveys. Bleaching was Prepared for INPEX Cardno 43

53 generally lower than recorded in previous peak bleaching periods between June 2012 and August 2012 (B1 to B3) and between February 2013 and April 2013 (D3 to D4). Coral bleaching is also measured by changes in the brightness of coral colour, with a reduction in coral tissue brightness likely to occur if photosynthetic symbiont density is reduced ( bleaching ). Such reductions occur during thermal bleaching events, but could also be caused by other stressors. The systematic measuring of coral colour could potentially detect a site-wide colour reduction that may not be apparent during the visual CPCe health assessments and, as such, coral colour may be an early indicator of the onset of bleaching. There were no increases in coral colour brightness in the current reporting period that would potentially indicate any bleaching or pre-bleaching characteristics Coral Recruits Coral recruits have previously been found to be more susceptible to sedimentation than established corals (Rogers 1990; Fabricius 2005; Erftemeijer et al. 2012). As such, the number of coral recruits could be influenced by dredging at sites with elevated turbidity. The number of recruits showed minimal change at all sites, relative to D11, although small decreases in the counts of juvenile corals (i.e. corals less than 20 mm in size) during D12 relative to D11were noted at Channel Island and Weed Reef 1, as well as a small increase at Weed Reef 2. Overall, there were apparent peaks in the number of coral recruits recorded in B1 (June 2012) and again in D5 (June 2013), which could be indicative of a seasonal recruitment pulse. It is worth noting that the temporal trend of the number of total recruits including Channel Island, Weed Reef 1 and Weed Reef 2 up to D11 is similar to the trend when including South Shell Island and Northeast Wickham Point (see Cardno 2014e). Only the peak in coral recruits at D1 became less pronounced when South Shell Island and Northeast Wickham Point were excluded from the dataset. This is in agreement with the results of the statistical analysis indicating a similarity among temporal trends at the five monitoring sites (see D11 report; Cardno 2014). The subsequent decline, following each recruitment event, may be due to the naturally high mortality rates experienced by coral recruits (Rogers 1990, Erftemeijer et al. 2012), but may also be a result of the recruits growing larger than the 20 mm size class and hence no longer being recorded as juveniles. It is important to note, however, that there is also the potential for the counts of recruits recorded to be influenced by changes in photograph quality. That is, during times of high turbidity, photograph quality is reduced, and due to the small and sometimes cryptic nature of coral recruits, the ability to identify and hence count recruits is likely to be reduced in poor conditions. Prepared for INPEX Cardno 44

54 6 Conclusions > During the D12 monitoring period (10 April 2014 to 10 May 2014) there were dry season turbidity Duration and Frequency trigger exceedances at Weed Reef 1 (Cardno 2014h). The INPEX Exceedance Attributability and Implementation Report considered these trigger exceedances to be primary attributable to natural causes (INPEX 2014b). There were no wet or dry season trigger exceedances of turbidity threshold at Channel Island or Weed Reef 2 (Cardno 2014f-h). > In D12, no Level 2 coral mortality trigger exceedances occurred, as the measured mortality at Channel Island, Weed Reef 1 and Weed Reef 2 was below the upper 95% confidence limit of the predicted mortality. > In D12, there were no Level 2 coral bleaching trigger exceedances. The measured bleaching at Channel Island and Weed Reef 1 was below the average bleaching of the Baseline Phase surveys, while Weed Reef 2 was slightly above. This resulted in gross bleaching of -5.1%, -2.6% and 0.9% at Channel Island, Weed Reef 1 and Weed Reef 2, respectively. > The measured mortality in D12 at Channel Island was similar to that predicted by the GLMM, while mortality at Weed Reef 1 and Weed Reef 2 was below the predicted mortality. Increase in mortality at Channel Island was mainly due to increases in immobile fauna (mainly sponges and hydroids) and sediment. The less than predicted mortality at Weed Reef 2 in D12 is in part because the predicted rate had been influenced by some corals dying as a consequence of being bleached prior to D3. As all of the bleached corals in Weed Reef 2 had either recovered or died by D7, the effects of bleaching have no longer contributed to the measured rate of mortality after D7. At Weed Reef 1 the measured mortality was above that predicted in D8 but below the predicted values in D9 to D12, which is consistent with the variability, relative to the slight increasing trend, in the measured mortality at this site. > In D12 there was no significant change in site-wide coral cover (along transects) at any site, or for all sites combined. The absence of a clear decrease in coral cover at most of the monitoring sites suggests that the mortality observed in the tagged corals is likely being compensated at a site-wide level by growth and, possibly, albeit to a lesser extent, by recruitment. > Sediment on coral, which constitutes up to half of the partial mortality of tagged corals recorded, has generally increased through time. Increasing sediment on tagged corals through time is expected, similar to the expected increase in mortality, as areas of dead coral become buried with no active removal of sediment occurring. Sediment on coral has shown a large degree of variability though time, which has generally translated to variability in the overall mortality trend. This suggests that the sediment may be settling on living coral, which can then be actively and passively cleared by the coral before subsequent surveys. > The increase in mortality since D11 at Channel Island and Weed Reef 1 was partially associated with increases in sediment on coral. There was also an increase in sediment paver depth since D11, noting there has been no correlation between sediment paver depth and sediment on coral to date. In addition, these increases in sediment were not evident at a community level, as sand and silt cover along transects at all reactive sites was significantly less than D11 and similar to Baseline Phase levels. Prepared for INPEX Cardno 45

55 7 Acknowledgements This report was prepared by Chris Roberts, Andrea Nicastro, David Cummings and Kelley Whitaker. The report was reviewed by Craig Blount. Field work was undertaken by David Cummings, Hamish Maitland, Kade Mills and Dan Pygas. CPCe analysis was undertaken by Kelley Whitaker, Yesmin Chikhani, Greig Campbell, Guy Graham, Rad Nair, Belinda Parkes and Dan Aveling. Prepared for INPEX Cardno 46

56 8 References Anderson, M. J., Gorley, R. N. and Clarke, K. R. (2008). PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods. PRIMER-E: Plymouth, UK. AHPI (2012). Australian Heritage Places Inventory. Web publication Date of access 22 October Australian Institute of Marine Science (2008). Surveys of benthic reef communities using underwater digital photography and counts of juvenile corals. Long-term Monitoring of the Great Barrier Reef Standard Operation Procedure Number 10. BOM (2014). Climate Data Online. Web publication Date of access 29 Apr Cardno (2012a). Coral Monitoring Program Baseline Report. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2012b). Coral Monitoring Method Statement. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2012c). Bimonthly Coral Monitoring Report: Dredging Report 1. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013a). Nearshore Environmental Monitoring Plan (Rev 2). Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013b). Bimonthly Coral Monitoring Report: Dredging Report 2. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013c). Bimonthly Coral Monitoring Report: Dredging Report 3. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013d). Bimonthly Coral Monitoring Report: Dredging Report 4. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013e). Bimonthly Coral Monitoring Report: Dredging Report 5. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2013f). Bimonthly Coral Monitoring Report: Dredging Report 6. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014a). Bimonthly Coral Monitoring Report: Dredging Report 7. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014b). Bimonthly Coral Monitoring Report: Dredging Report 8. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014c). Coral Monitoring Report: Dredging Report 9. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014d). Coral Monitoring Report: Dredging Report 10. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014e). Coral Monitoring Report: Dredging Report 11. Report for INPEX. Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014f). Fortnightly Water Quality Report Weeks 85/86: 7 April 2014 to 21 April Report for INPEX, Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014g). Fortnightly Water Quality Report - Weeks 86/87: 14 April to 27 April Report for INPEX, Cardno (NSW/ACT) Pty Ltd, Sydney. Cardno (2014h). Fortnightly Water Quality Report - Weeks 88/89: 28 April to 11 May Report for INPEX, Cardno (NSW/ACT) Pty Ltd, Sydney. Cortes, J. and Risk, M. J. (1985). A reef under siltation stress: Cahuita, Costa Rica. Bulletin of Marine Science, 36(2), pp Erftemeijer, P. L. A., Riegl, B., Hoeksema, B. W. and Todd, P. A. (2012). Environmental impacts of dredging and other sediment disturbances on corals: a review. Marine Pollution Bulletin, 64, pp Fabricius, K. E. (2005). Effects of terrestrial runoff on the ecology of corals and coral reefs: review and synthesis. Marine Pollution Bulletin, 50, pp INPEX (2011). Ichthys Gas Field Development Project - Draft Environmental Impact Statement. INPEX Operations Australia Pty Ltd. Prepared for INPEX Cardno 47

57 INPEX (2013). Dredging and Spoil Disposal Management Plan East Arm (Rev 4). INPEX Operations Australia Pty Ltd. INPEX (2014a). Dredging and Spoil Disposal Management Plan Gas Export Pipeline (Rev 6). INPEX Operations Australia Pty Ltd. INPEX (2014b). Exceedance Attributability and Implementation Report: Weed Reef 1, 4 May INPEX Operations Australia Pty Ltd Kelley, R. (2009). Indo Pacific Coral Finder, BYO Guides. The Australian Coral Reef Society, Townsville. Kohler, K. E. and Gill, S. M. (2006). Coral Point Count with Excel Extensions (CPCe): A Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Computers and Geosciences, 32, pp Lirman, D. and Manzello, D. (2009). Patterns of resistance and resilience of the stress-tolerant coral Siderastrea radians (Pallas) to suboptimal salinity and sediment burial. Journal of Experimental Marine Biology and Ecology, 369, pp McCook, L. J., Jompa, J. and Diaz-Pulido, G. (2001). Competition between corals and algae on coral reefs: a review of evidence and mechanisms. Coral Reefs, 19, pp NOAA Coral Reef Watch (2014). NOAA/NESDIS Bleaching Alert Area. Web publication Date of access 01 Apr Phillipp, E. and Fabricius, K. E. (2003). Photophysiological stress in scleractinian corals in response to short-term sedimentation. Journal of Experimental Marine Biology and Ecology 287, Rogers, C. S. (1990). Responses of Coral Reefs and Reef Organisms to Sedimentation. Marine Ecology Progress Series, 62, pp Smith, J. E., Shaw, M., Edwards, R. A., Obura, D., Pantos, O., Sala, E., Sandin, S. A., Smriga, S., Hatay, M. and Rohwer, L. (2006). Indirect effects of algae on coral: algae-mediated, microbe-induced coral mortality. Ecology Letters, 9, pp Underwood, A. J. (1992). Beyond BACI: the detection of environmental impacts on populations in the real, but variable, world. Journal of Experimental Marine Biology and Ecology, 161, pp Veal, C. J., Holmes, G., Hoegh-Guldberg, O., Osborn, J. (2010). A comparative study of methods for surface area and threedimensional shape measurement of coral skeletons. Limnology and Oceanography: Methods, 8, pp Veron, J. E. N. (1986). Corals of Australia and the Indo Pacific. Angus and Robertson Ltd, Hong Kong, 644 pp. Veron, J. E. N. (2000). Corals of the World. Australian Institute of Marine Science, Townsville, 489 pp. Prepared for INPEX Cardno 48

58 APPENDIX A TYPE AND FORM OF CORAL COLONIES Prepared for INPEX Cardno 49

59 Appendix A-1 Type and form of all tagged coral colonies Site Tag Family Genus Form Channel Island CHI_01 Faviidae Montastrea Massive CHI_02 Faviidae Favites Submassive CHI_03 Poritidae Goniopora Massive CHI_04 Faviidae Cyphastrea Submassive CHI_05 Pectiniidae Mycedium Foliose CHI_06 Pectiniidae Mycedium Encrusting CHI_07 Pectiniidae Mycedium Encrusting CHI_08 Dendrophylliidae Turbinaria Foliose CHI_09 Poritidae Goniopora Massive CHI_10 Pectiniidae Mycedium Encrusting CHI_11 Faviidae Favites Encrusting CHI_12 Pectiniidae Echinophyllia Submassive CHI_13 Faviidae Favites Submassive CHI_14 Faviidae Favia/Montastrea Massive CHI_15 Pectiniidae Mycedium Foliose CHI_16 Poritidae Goniopora Massive CHI_17 Pectiniidae Mycedium Encrusting CHI_18 Faviidae Favia/Montastrea Massive CHI_19 Poritidae Goniopora Massive CHI_20 Pectiniidae Mycedium Encrusting CHI_21 Faviidae Favites Submassive CHI_22 Faviidae Favites Submassive CHI_23 Faviidae Favia/Montastrea Encrusting CHI_24 Pectiniidae Echinophyllia Encrusting CHI_25 Faviidae Favia/Montastrea Massive CHI_26 Dendrophylliidae Turbinaria Foliose CHI_27 Poritidae Goniopora Massive CHI_28 Pectiniidae Echinophyllia Submassive CHI_29 Faviidae Cyphastrea Submassive CHI_30 Dendrophyllidae Turbinaria Foliose CHI_31 Pectiniidae Mycedium Encrusting CHI_32 Faviidae Favia/Montastrea Encrusting CHI_33 Faviidae Favites Submassive CHI_34 Faviidae Cyphastrea Massive CHI_35 Dendrophylliidae Turbinaria Foliose CHI_36 Faviidae Cyphastrea Submassive CHI_37 Poritidae Goniopora Massive CHI_38 Poritidae Goniopora Massive CHI_39 Faviidae Platygyra Encrusting CHI_40 Dendrophylliidae Turbinaria Foliose CHI_41 Faviidae Montastrea Encrusting CHI_42 Pectiniidae Mycedium Encrusting CHI_43 Pectiniidae Echinophyllia Submassive Prepared for INPEX Cardno 50

60 Site Tag Family Genus Form CHI_44 Faviidae Moseleya Submassive CHI_45 Pectiniidae Mycedium Submassive CHI_46 Faviidae Favia Massive CHI_47 Faviidae Favia/Montastrea Massive CHI_48 Faviidae Favites Submassive CHI_49 Dendrophylliidae Turbinaria Encrusting CHI_50 Faviidae Favia/Montastrea Massive CHI_51 Faviidae Cyphastrea Submassive CHI_52 Pectiniidae Mycedium Foliose CHI_53 Faviidae Favia/Montastrea Encrusting CHI_54 Pectiniidae Mycedium Encrusting CHI_55 Pectiniidae Mycedium Encrusting CHI_56 Pectiniidae Echinophyllia Encrusting CHI_57 Dendrophylliidae Turbinaria Foliose CHI_58 Pectiniidae Echinophyllia Submassive CHI_59 Faviidae Favia Encrusting CHI_60 Faviidae Favia Massive CHI_61 Faviidae Favia Massive CHI_62 Faviidae Favia/Montastrea Encrusting CHI_63 Poritidae Goniopora Massive CHI_64 Pectiniidae Mycedium Encrusting CHI_65 Faviidae Moseleya Submassive CHI_66 Poritidae Goniopora Massive CHI_67 Pectiniidae Mycedium Encrusting CHI_68 Pectiniidae Mycedium Encrusting CHI_69 Pectiniidae Mycedium Encrusting CHI_70 Faviidae Favia/Montastrea Encrusting CHI_71 Faviidae Cyphastrea Submassive CHI_72 Pectiniidae Echinophyllia Submassive CHI_73 Pectiniidae Mycedium Encrusting CHI_74 Faviidae Favia/Montastrea Massive CHI_75 Pectiniidae Echinophyllia Encrusting Charles Point CHP_01 Poritidae Goniopora Massive CHP_02 Dendrophylliidae Turbinaria Foliose CHP_03 Dendrophylliidae Turbinaria Foliose CHP_04 Dendrophylliidae Turbinaria Foliose CHP_05 Faviidae Favia/Montastrea Massive CHP_06 Poritidae Goniopora Massive CHP_07 Faviidae Favia/Montastrea Massive CHP_08 Dendrophylliidae Turbinaria Foliose CHP_09 Faviidae Favia/Montastrea Encrusting CHP_10 Dendrophylliidae Turbinaria Foliose CHP_11 Poritidae Porites Encrusting CHP_12 Dendrophylliidae Turbinaria Encrusting Prepared for INPEX Cardno 51

61 Site Tag Family Genus Form CHP_13 Dendrophylliidae Turbinaria Encrusting CHP_14 Poritidae Goniopora Massive CHP_15 Dendrophylliidae Turbinaria Foliose CHP_16 Faviidae Favites Submassive CHP_17 Faviidae Goniastrea Massive CHP_18 Faviidae Goniastrea Submassive CHP_19 Poritidae Goniopora Massive CHP_20 Dendrophylliidae Turbinaria Foliose CHP_21 Dendrophylliidae Turbinaria Foliose CHP_22 Faviidae Favia/Montastrea Massive CHP_23 Poritidae Goniopora Massive CHP_24 Faviidae Favites Submassive CHP_25 Faviidae Favia/Montastrea Massive CHP_26 Faviidae Moseleya Submassive CHP_27 Faviidae Favia Massive CHP_28 Dendrophylliidae Turbinaria Foliose CHP_29 Poritidae Goniopora Massive CHP_30 Faviidae Favites Massive CHP_31 Faviidae Favia/Montastrea Massive CHP_32 Faviidae Favia Encrusting CHP_33 Poritidae Goniopora Massive CHP_34 Dendrophylliidae Turbinaria Foliose CHP_35 Faviidae Favites Submassive CHP_36 Dendrophylliidae Turbinaria Foliose CHP_37 Poritidae Goniopora Massive CHP_38 Dendrophylliidae Turbinaria Foliose CHP_39 Faviidae Favia Massive CHP_40 Faviidae Favia/Montastrea Massive CHP_41 Dendrophylliidae Turbinaria Foliose CHP_42 Faviidae Favites Submassive CHP_43 Faviidae Favia Massive CHP_44 Poritidae Porites Submassive CHP_45 Dendrophylliidae Turbinaria Foliose CHP_46 Faviidae Montastrea Submassive CHP_47 Oculinidae Galaxea Submassive CHP_48 Faviidae Favia Massive CHP_49 Dendrophylliidae Turbinaria Foliose CHP_50 Poritidae Goniopora Massive CHP_51 Faviidae Goniastrea Submassive CHP_52 Faviidae Favia Massive CHP_53 Dendrophylliidae Turbinaria Foliose CHP_54 Dendrophylliidae Turbinaria Foliose CHP_55 Dendrophylliidae Turbinaria Foliose CHP_56 Faviidae Platygyra Submassive Prepared for INPEX Cardno 52

62 Site Tag Family Genus Form CHP_57 Faviidae Favia/Montastrea Encrusting CHP_58 Dendrophylliidae Turbinaria Foliose CHP_59 Faviidae Barabattoia Massive CHP_60 Dendrophylliidae Turbinaria Foliose CHP_61 Faviidae Favia/Montastrea Massive CHP_62 Faviidae Favites Encrusting CHP_63 Faviidae Favia/Montastrea Massive CHP_64 Faviidae Favia Massive CHP_65 Dendrophylliidae Turbinaria Foliose CHP_66 Faviidae Favia/Montastrea Massive CHP_67 Faviidae Favia/Montastrea Massive CHP_68 Faviidae Platygyra Submassive CHP_69 Poritidae Goniopora Massive CHP_70 Faviidae Montastrea Massive CHP_71 Faviidae Favia/Montastrea Massive CHP_72 Faviidae Montastrea Massive CHP_73 Faviidae Favia/Montastrea Massive CHP_74 Faviidae Favia/Montastrea Massive CHP_75 Faviidae Favia/Montastrea Submassive Mandorah MAN_01 Dendrophylliidae Turbinaria Foliose MAN_02 Dendrophylliidae Turbinaria Foliose MAN_03 Dendrophylliidae Turbinaria Foliose MAN_04 Dendrophylliidae Turbinaria Foliose MAN_05 Faviidae Favia Encrusting MAN_06 Dendrophylliidae Turbinaria Foliose MAN_07 Faviidae Goniastrea Submassive MAN_08 Faviidae Favia Encrusting MAN_09 Dendrophylliidae Turbinaria Foliose MAN_10 Dendrophylliidae Turbinaria Foliose MAN_11 Faviidae Favites Encrusting MAN_12 Faviidae Favia Encrusting MAN_13 Faviidae Favites Submassive MAN_14 Dendrophylliidae Turbinaria Foliose MAN_15 Dendrophylliidae Turbinaria Foliose MAN_16 Dendrophylliidae Turbinaria Foliose MAN_17 Faviidae Favia/Montastrea Encrusting MAN_18 Faviidae Cyphastrea Encrusting MAN_19 Faviidae Montastrea Encrusting MAN_20 Dendrophylliidae Turbinaria Foliose MAN_21 Faviidae Cyphastrea Submassive MAN_22 Faviidae Favia/Montastrea Encrusting MAN_23 Dendrophylliidae Turbinaria Foliose MAN_24 Poritidae Goniopora Massive MAN_25 Faviidae Favia Encrusting Prepared for INPEX Cardno 53

63 Site Tag Family Genus Form MAN_26 Faviidae Favites Submassive MAN_27 Faviidae Favia Encrusting MAN_28 Dendrophylliidae Turbinaria Foliose MAN_29 Dendrophylliidae Turbinaria Foliose MAN_30 Faviidae Favia Encrusting MAN_31 Dendrophylliidae Turbinaria Foliose MAN_32 Dendrophylliidae Turbinaria Foliose MAN_33 Dendrophylliidae Turbinaria Foliose MAN_34 Faviidae Favia/Montastrea Encrusting MAN_35 Dendrophylliidae Turbinaria Foliose MAN_36 Dendrophylliidae Turbinaria Foliose MAN_37 Dendrophylliidae Turbinaria Foliose MAN_38 Faviidae Cyphastrea Massive MAN_39 Faviidae Goniastrea Submassive MAN_40 Dendrophylliidae Turbinaria Foliose MAN_41 Dendrophylliidae Turbinaria Foliose MAN_42 Dendrophylliidae Turbinaria Foliose MAN_43 Dendrophylliidae Turbinaria Foliose MAN_44 Dendrophylliidae Turbinaria Foliose MAN_45 Faviidae Favites Submassive MAN_46 Faviidae Favia Submassive MAN_47 Dendrophylliidae Turbinaria Foliose MAN_48 Faviidae Favia Encrusting MAN_49 Dendrophylliidae Turbinaria Foliose MAN_50 Dendrophylliidae Turbinaria Foliose MAN_51 Dendrophylliidae Turbinaria Foliose MAN_52 Faviidae Favia Massive MAN_53 Dendrophylliidae Turbinaria Foliose MAN_54 Faviidae Favia Encrusting MAN_55 Faviidae Favia/Montastrea Encrusting MAN_56 Faviidae Leptastrea Submassive MAN_57 Faviidae Moseleya Submassive MAN_58 Faviidae Favia Massive MAN_59 Faviidae Favia Encrusting MAN_60 Dendrophylliidae Turbinaria Foliose MAN_61 Dendrophylliidae Turbinaria Foliose MAN_62 Faviidae Favia Massive MAN_63 Dendrophylliidae Turbinaria Foliose MAN_64 Dendrophylliidae Turbinaria Foliose MAN_65 Dendrophylliidae Turbinaria Foliose MAN_66 Faviidae Favia/Montastrea Submassive MAN_67 Dendrophylliidae Turbinaria Foliose MAN_68 Faviidae Moseleya Submassive MAN_69 Faviidae Favia Submassive Prepared for INPEX Cardno 54

64 Site Tag Family Genus Form MAN_70 Faviidae Favia Encrusting MAN_71 Dendrophylliidae Turbinaria Foliose MAN_72 Poritidae Goniopora Massive MAN_73 Faviidae Favites Submassive MAN_74 Dendrophylliidae Turbinaria Foliose MAN_75 Dendrophylliidae Turbinaria Foliose South Shell Island SSI_01 Pectiniidae Mycedium Encrusting SSI_02 Pectiniidae Mycedium Encrusting SSI_03 Pectiniidae Mycedium Encrusting SSI_04 Pectiniidae Mycedium Encrusting SSI_05 Pectiniidae Mycedium Encrusting SSI_06 Pectiniidae Mycedium Foliose SSI_07 Pectiniidae Mycedium Encrusting SSI_08 Pectiniidae Mycedium Foliose SSI_09 Poritidae Goniopora Massive SSI_10 Pectiniidae Mycedium Foliose SSI_11 Poritidae Goniopora Massive SSI_12 Pectiniidae Mycedium Encrusting SSI_13 Poritidae Goniopora Massive SSI_14 Pectiniidae Mycedium Encrusting SSI_15 Pectiniidae Mycedium Encrusting SSI_16 Pectiniidae Mycedium Foliose SSI_17 Pectiniidae Mycedium Encrusting SSI_18 Pectiniidae Mycedium Encrusting SSI_19 Pectiniidae Mycedium Encrusting SSI_20 Pectiniidae Mycedium Foliose SSI_21 Pectiniidae Mycedium Foliose SSI_22 Pectiniidae Mycedium Encrusting SSI_23 Merulinidae Hydnophora Encrusting SSI_24 Pectiniidae Mycedium Foliose SSI_25 Pectiniidae Mycedium Foliose SSI_26 Pectiniidae Mycedium Foliose SSI_27 Pectiniidae Mycedium Foliose SSI_28 Pectiniidae Mycedium Encrusting SSI_29 Pectiniidae Mycedium Encrusting SSI_30 Pectiniidae Mycedium Encrusting SSI_31 Faviidae Favia/Montastrea Massive SSI_32 Pectiniidae Mycedium Encrusting SSI_33 Faviidae Favia/Montastrea Massive SSI_34 Faviidae Favia/Montastrea Massive SSI_35 Pectiniidae Mycedium Foliose SSI_36 Pectiniidae Mycedium Encrusting SSI_37 Pectiniidae Mycedium Encrusting SSI_38 Pectiniidae Mycedium Encrusting Prepared for INPEX Cardno 55

65 Site Tag Family Genus Form SSI_39 Pectiniidae Mycedium Encrusting SSI_40 Faviidae Moseleya Submassive SSI_41 Pectiniidae Mycedium Encrusting SSI_42 Pectiniidae Mycedium Foliose SSI_43 Fungiidae Lithophyllon Foliose SSI_44 Pectiniidae Mycedium Encrusting SSI_45 Pectiniidae Mycedium Encrusting SSI_46 Dendrophylliidae Duncanopsammia Branching SSI_47 Pectiniidae Mycedium Foliose SSI_48 Pectiniidae Mycedium Encrusting SSI_49 Pectiniidae Mycedium Foliose SSI_50 Pectiniidae Mycedium Foliose SSI_51 Pectiniidae Mycedium Foliose SSI_52 Pectiniidae Mycedium Encrusting SSI_53 Pectiniidae Mycedium Encrusting SSI_54 Oculinidae Galaxea Submassive SSI_55 Pectiniidae Mycedium Encrusting SSI_56 Poritidae Goniopora Massive SSI_57 Poritidae Goniopora Massive SSI_58 Poritidae Goniopora Massive SSI_59 Pectiniidae Mycedium Encrusting SSI_60 Pectiniidae Mycedium Encrusting SSI_61 Pectiniidae Mycedium Encrusting SSI_62 Pectiniidae Mycedium Foliose SSI_63 Poritidae Goniopora Massive SSI_64 Poritidae Goniopora Massive SSI_65 Poritidae Goniopora Massive SSI_66 Pectiniidae Mycedium Foliose SSI_67 Poritidae Goniopora Massive SSI_68 Pectiniidae Mycedium Encrusting SSI_69 Pectiniidae Mycedium Encrusting SSI_70 Pectiniidae Mycedium Encrusting SSI_71 Pectiniidae Mycedium Encrusting SSI_72 Pectiniidae Mycedium Encrusting SSI_73 Poritidae Goniopora Massive SSI_74 Faviidae Moseleya Submassive SSI_75 Poritidae Goniopora Massive Weed Reef 1 WR1_01 Faviidae Goniastrea Submassive WR1_02 Faviidae Montastrea Submassive WR1_03 Dendrophylliidae Turbinaria Foliose WR1_04 Pectiniidae Mycedium Encrusting WR1_05 Faviidae Goniastrea Submassive WR1_06 Poritidae Goniopora Massive WR1_07 Faviidae Montastrea Massive Prepared for INPEX Cardno 56

66 Site Tag Family Genus Form WR1_08 Dendrophylliidae Turbinaria Foliose WR1_09 Faviidae Favia Massive WR1_10 Dendrophylliidae Turbinaria Foliose WR1_11 Dendrophylliidae Turbinaria Encrusting WR1_12 Faviidae Favites Submassive WR1_13 Dendrophylliidae Turbinaria Foliose WR1_14 Faviidae Favia/Montastrea Massive WR1_15 Pectiniidae Mycedium Encrusting WR1_16 Faviidae Favia Submassive WR1_17 Pectiniidae Mycedium Encrusting WR1_18 Faviidae Favites Submassive WR1_19 Faviidae Favia Massive WR1_20 Pectiniidae Mycedium Encrusting WR1_21 Faviidae Montastrea Submassive WR1_22 Agariciidae Pachyseris Foliose WR1_23 Pectiniidae Mycedium Encrusting WR1_24 Merulinidae Hydnophora Submassive WR1_25 Faviidae Moseleya Submassive WR1_26 Merulinidae Merulina Submassive WR1_27 Faviidae Favites Submassive WR1_28 Faviidae Platygyra Encrusting WR1_29 Faviidae Favia/Montastrea Massive WR1_30 Poritidae Goniopora Massive WR1_31 Pectiniidae Mycedium Encrusting WR1_32 Poritidae Goniopora Massive WR1_33 Fungiidae Lithophyllon Encrusting WR1_34 Faviidae Favia/Montastrea Submassive WR1_35 Faviidae Goniastrea Massive WR1_36 Faviidae Favia Massive WR1_37 Dendrophylliidae Turbinaria Foliose WR1_38 Pectiniidae Echinophyllia Encrusting WR1_39 Faviidae Favites Massive WR1_40 Dendrophylliidae Turbinaria Foliose WR1_41 Pectiniidae Echinophyllia Encrusting WR1_42 Pectiniidae Echinophyllia Encrusting WR1_43 Pectiniidae Mycedium Encrusting WR1_44 Faviidae Favites Submassive WR1_45 Pectiniidae Oxypora Encrusting WR1_46 Dendrophylliidae Turbinaria Foliose WR1_47 Faviidae Favites Submassive WR1_48 Pectiniidae Mycedium Encrusting WR1_49 Poritidae Goniopora Massive WR1_50 Faviidae Favia Encrusting WR1_51 Dendrophylliidae Turbinaria Foliose Prepared for INPEX Cardno 57

67 Site Tag Family Genus Form WR1_52 Pectiniidae Echinophyllia Encrusting WR1_53 Dendrophylliidae Turbinaria Foliose WR1_54 Poritidae Goniopora Massive WR1_55 Faviidae Favites Submassive WR1_56 Dendrophylliidae Turbinaria Foliose WR1_57 Faviidae Favites Massive WR1_58 Pectiniidae Mycedium Encrusting WR1_59 Mussidae Lobophyllia Massive WR1_60 Dendrophylliidae Turbinaria Foliose WR1_61 Fungiidae Lithophyllon Encrusting WR1_62 Pectiniidae Mycedium Encrusting WR1_63 Pectiniidae Mycedium Encrusting WR1_64 Faviidae Favia Submassive WR1_65 Pectiniidae Mycedium Encrusting WR1_66 Pectiniidae Echinophyllia Encrusting WR1_67 Pectiniidae Mycedium Encrusting WR1_68 Faviidae Montastrea Massive WR1_69 Dendrophylliidae Turbinaria Foliose WR1_70 Faviidae Favia Massive WR1_71 Dendrophylliidae Turbinaria Foliose WR1_72 Pectiniidae Mycedium Foliose WR1_73 Faviidae Favia/Montastrea Encrusting WR1_74 Pectiniidae Mycedium Encrusting WR1_75 Faviidae Favites Submassive Weed Reef 2 WR2_01 Faviidae Montastrea Massive WR2_02 Poritidae Porites Massive WR2_03 Faviidae Favites Encrusting WR2_04 Faviidae Montastrea Encrusting WR2_05 Faviidae Favia Encrusting WR2_06 Poritidae Goniopora Massive WR2_07 Faviidae Moseleya Submassive WR2_08 Pectiniidae Mycedium Foliose WR2_09 Faviidae Montastrea Submassive WR2_10 Pectiniidae Mycedium Encrusting WR2_11 Dendrophylliidae Turbinaria Foliose WR2_12 Faviidae Favia Massive WR2_13 Poritidae Goniopora Massive WR2_14 Faviidae Favia/Montastrea Encrusting WR2_15 Pectiniidae Mycedium Encrusting WR2_16 Pectiniidae Echinophyllia Encrusting WR2_17 Faviidae Favia/Montastrea Encrusting WR2_18 Faviidae Montastrea Encrusting WR2_19 Pectiniidae Echinophyllia Encrusting WR2_20 Pectiniidae Mycedium Encrusting Prepared for INPEX Cardno 58

68 Site Tag Family Genus Form WR2_21 Dendrophylliidae Turbinaria Foliose WR2_22 Pectiniidae Mycedium Encrusting WR2_23 Pectiniidae Mycedium Encrusting WR2_24 Poritidae Goniopora Massive WR2_25 Faviidae Favia Encrusting WR2_26 Dendrophylliidae Turbinaria Foliose WR2_27 Fungiidae Lithophyllon Encrusting WR2_28 Poritidae Goniopora Massive WR2_29 Pectiniidae Mycedium Encrusting WR2_30 Dendrophylliidae Turbinaria Foliose WR2_31 Poritidae Alveopora Massive WR2_32 Dendrophylliidae Turbinaria Foliose WR2_33 Dendrophylliidae Turbinaria Foliose WR2_34 Faviidae Cyphastrea Massive WR2_35 Poritidae Alveopora Massive WR2_36 Poritidae Alveopora Massive WR2_37 Poritidae Alveopora Massive WR2_38 Faviidae Favites Encrusting WR2_39 Poritidae Goniopora Massive WR2_40 Pectiniidae Mycedium Encrusting WR2_41 Dendrophylliidae Turbinaria Foliose WR2_42 Poritidae Alveopora Massive WR2_43 Dendrophylliidae Turbinaria Foliose WR2_44 Faviidae Favites Submassive WR2_45 Poritidae Alveopora Massive WR2_46 Poritidae Alveopora Massive WR2_47 Pectiniidae Mycedium Encrusting WR2_48 Faviidae Montastrea Massive WR2_49 Faviidae Montastrea Massive WR2_50 Faviidae Favia Massive WR2_51 Poritidae Alveopora Massive WR2_52 Faviidae Favia/Montastrea Massive WR2_53 Poritidae Goniopora Massive WR2_54 Dendrophylliidae Turbinaria Foliose WR2_55 Fungiidae Lithophyllon Encrusting WR2_56 Pectiniidae Echinophyllia Submassive WR2_57 Dendrophylliidae Turbinaria Foliose WR2_58 Faviidae Favites Encrusting WR2_59 Siderastreidae Psammocora Submassive WR2_60 Dendrophylliidae Turbinaria Encrusting WR2_61 Pectiniidae Mycedium Encrusting WR2_62 Faviidae Favia Submassive WR2_63 Dendrophylliidae Turbinaria Foliose WR2_64 Pectiniidae Mycedium Encrusting Prepared for INPEX Cardno 59

69 Site Tag Family Genus Form WR2_65 Poritidae Goniopora Massive WR2_66 Pectiniidae Mycedium Encrusting WR2_67 Dendrophylliidae Turbinaria Encrusting WR2_68 Faviidae Favia Submassive WR2_69 Faviidae Favia/Montastrea Massive WR2_70 Faviidae Favites Submassive WR2_71 Pectiniidae Mycedium Encrusting WR2_72 Fungiidae Lithophyllon Encrusting WR2_73 Dendrophylliidae Turbinaria Foliose WR2_74 Faviidae Montastrea Massive WR2_75 Faviidae Favites Submassive Northeast Wickham Point NEW_01 Faviidae Cyphastrea Massive NEW_02 Dendrophylliidae Turbinaria Foliose NEW_03 Faviidae Moseleya Submassive NEW_04 Faviidae Favites Submassive NEW_05 Faviidae Favia Massive NEW_06 Dendrophylliidae Turbinaria Foliose NEW_07 Faviidae Favia Massive NEW_08 Faviidae Cyphastrea Submassive NEW_09 Faviidae Favia/Montastrea Massive NEW_10 Faviidae Cyphastrea Submassive NEW_11 Poritidae Goniopora Massive NEW_12 Faviidae Favia/Montastrea Massive NEW_13 Faviidae Favites Encrusting NEW_14 Faviidae Cyphastrea Submassive NEW_15 Faviidae Montastrea Massive NEW_16 Faviidae Moseleya Submassive NEW_17 Faviidae Favia/Montastrea Submassive NEW_18 Faviidae Favites Submassive NEW_19 Faviidae Favia Massive NEW_20 Pectiniidae Mycedium Encrusting NEW_21 Faviidae Montastrea Massive NEW_22 Faviidae Moseleya Submassive NEW_23 Faviidae Favia/Montastrea Massive NEW_24 Faviidae Favia/Montastrea Massive NEW_25 Pectiniidae Mycedium Encrusting NEW_26 Dendrophylliidae Turbinaria Foliose NEW_27 Faviidae Moseleya Submassive NEW_28 Faviidae Favia Encrusting NEW_29 Faviidae Favia/Montastrea Massive NEW_30 Faviidae Favites Encrusting NEW_31 Pectiniidae Mycedium Foliose NEW_32 Faviidae Favia/Montastrea Massive NEW_33 Faviidae Favia/Montastrea Encrusting Prepared for INPEX Cardno 60

70 Site Tag Family Genus Form NEW_34 Faviidae Favia/Montastrea Massive NEW_35 Pectiniidae Mycedium Encrusting NEW_36 Pectiniidae Mycedium Encrusting NEW_37 Pectiniidae Mycedium Encrusting NEW_38 Pectiniidae Mycedium Encrusting NEW_39 Faviidae Favia/Montastrea Encrusting NEW_40 Faviidae Favia/Montastrea Massive NEW_41 Faviidae Moseleya Submassive NEW_42 Poritidae Porites Encrusting NEW_43 Fungiidae Lithophyllon Encrusting NEW_44 Faviidae Favia Massive NEW_45 Faviidae Goniastrea Submassive NEW_46 Faviidae Favites Encrusting NEW_47 Dendrophylliidae Turbinaria Foliose NEW_48 Dendrophylliidae Turbinaria Foliose NEW_49 Faviidae Moseleya Submassive NEW_50 Faviidae Favia/Montastrea Massive NEW_51 Dendrophylliidae Turbinaria Foliose NEW_52 Faviidae Favia/Montastrea Encrusting NEW_53 Faviidae Favia/Montastrea Massive NEW_54 Dendrophylliidae Turbinaria Foliose NEW_55 Pectiniidae Mycedium Encrusting NEW_56 Dendrophylliidae Turbinaria Foliose NEW_57 Pectiniidae Mycedium Encrusting NEW_58 Poritidae Goniopora Massive NEW_59 Poritidae Goniopora Massive NEW_60 Poritidae Goniopora Massive NEW_61 Dendrophylliidae Turbinaria Encrusting NEW_62 Poritidae Goniopora Massive NEW_63 Dendrophylliidae Turbinaria Foliose NEW_64 Faviidae Favia Massive NEW_65 Pectiniidae Mycedium Encrusting NEW_66 Faviidae Favia/Montastrea Massive NEW_67 Poritidae Goniopora Massive NEW_68 Faviidae Montastrea Massive NEW_69 Dendrophylliidae Turbinaria Foliose NEW_70 Faviidae Favia Massive NEW_71 Faviidae Cyphastrea Submassive NEW_72 Faviidae Favia Massive NEW_73 Dendrophylliidae Turbinaria Foliose NEW_74 Faviidae Cyphastrea Submassive NEW_75 Pectiniidae Mycedium Encrusting Prepared for INPEX Cardno 61

71 APPENDIX B STATISTICAL ANALYSES Prepared for INPEX Cardno 62

72 Appendix B-1 Background Comparison between sampled and non-sampled missing colonies Using the observed rate of coral mortality (i.e. the site-wide trend) in tagged corals at the reactive sites (Channel Island, Weed Reef 1 and Weed Reef 2) during the ten surveys preceding commencement of Season Two dredging in EA (i.e. Baseline Phase surveys B1 to B3, and Dredging Phase surveys D1 to D7, undertaken from June 2012 to October 2013), it is possible to forecast expected levels of mortality in Season Two dredging. Using observed proportions of coral mortality (total dead/total coral count), a generalized linear mixed model (GLMM) was fitted for interpolating and forecasting global coral mortality for each site. The GLMM does not attempt to adjust model predictions for the possible presence of mortality where there are missing observations. Above average mortality of colonies that were not sampled from D8 onwards may cause the measured mortality to be underestimated relative to the modelled mortality and the mortality trigger values. A model that provides a Bayesian model which uses a probabilistic approach for adjusting predicted site-specific mortality for potential mortality (i.e. adjusts for missing observations) will be used if empirical evidence suggests that the mortality of missing colonies (as measured in these colonies prior to them becoming missing, that is, from measures of partial mortality in the preceding survey) is biased toward above average mortality. To ensure the GLMM is appropriate, we compared the mortality of missing colonies with the mortality of sampled colonies. If the mortality of non- sampled colonies (missing colonies) is significantly higher than the sampled colonies then it is proposed to use a Bayesian model that accounts for the non- sampled colonies. Methods The number of missing colonies at the five monitoring sites was a small fraction of the total number of coral colonies that was measured. As a consequence of the severe unbalanced number of replicates, it was not possible to use a parametric test (such as a t-test) to compare the average mortality of the missing colonies with that of the measured colonies. Instead, a resampling technique was used to calculate the average difference between the two means and obtain a balanced design for a parametric test. Mortality data were divided into two groups: > Sampled colonies: mortality of all colonies at survey S that were sampled in survey S + 1; > Not sampled colonies: mortality of all colonies at survey S that were not sampled in survey S + 1. If a colony was not sampled for two or more consecutive surveys only the mortality of the last survey in which the colony was sampled was included. For each site, the average mortality of the non-sampled colonies was calculated. The average mortality of the sampled colonies was calculated for 20,000 random samples (with replacement) of the same number of replicates as the number of colonies that were missing. In this way the same number of replicates for sampled and not sampled colonies groups was obtained. The distribution of the difference between the two groups (average mortality of non-sampled colonies average mortality of sampled colonies) was plotted and the upper and lower 95% confidence limits were obtained based on the 2.5 and 97.5 percentiles. For each iteration, the difference in mortality between the averages of the groups was tested for significance with a t- test. The average mortality of the non-sampled colonies was considered above the average mortality of sampled colonies when more than 5% of t-tests (1000 tests) indicated a significantly greater mortality for the non-sampled colonies. Results The number of missing colonies between D8 and D12 surveys was ranged between 6 and 11 (Table B-1). Sampled colonies ranged between 328 and 362. Weed Reef 2 was the only site in which the average mortality of non- sampled colonies was higher than sampled colonies and the mean difference was 1.5%. Resampling indicated that mortality of colonies that were not sampled was less than or equal to the mortality of sampled colonies at all sites (Table B-1). This is supported by a high number of iterations (between Prepared for INPEX Cardno 63

73 99.7% and 100.0%) in which the mortality of the two groups was not statistically different or the non-sampled group mortality was significantly less than the group of sampled colonies. Further, the 95% confidence intervals overlap zero indicating the difference between the average mortality of the sampled and nonsampled colonies is not significantly different from zero (Figure B-1). Density (%) Density Channel Island CHI Density Weed Reef 1 WR Missing - Measured Mortality (%) Density (%) Density WR2 Missing - Measured Weed Reef Mortality Missing - Measured (%) Figure B-1 Histogram of the difference between average mortality of non-sampled corals to the average mortality of 20,000 random selections from the sampled coral. Dotted line indicates the median, red lines indicate the 2.5 and 97.5 percentiles Prepared for INPEX Cardno 64

74 Table B-1 Comparison of average mortality of non-sampled corals to the average mortality of 20,000 random selections from the sampled corals Site CHI WR1 WR2 Summary statistics Number of not sampled colonies Number of sampled colonies Average mortality of not sampled colonies Average mortality of sampled colonies Resampling summary statistics Number of resampling iteration 20,000 20,000 20,000 Mean Average Difference (not sampled minus sampled colonies) Number of non-significant t-tests 19,989 19,585 19,997 Number of iterations where mortality of non-sampled was significantly greater than mortality of sampled colonies (t-test) Number of iterations where mortality of non-sampled was significantly less than mortality of sampled colonies (t-test) Prepared for INPEX Cardno 65

75 Appendix B-2 Multivariate analysis using PERMANOVA testing for differences in the benthic composition among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p 0.05) p-values in bold. RED = redundant terms due to significant interaction. MC = Monte Carlo simulation was used to calculate p-values where unique permutations < 100. A. PERMANOVA analysis Source of Variation df SS MS Pseudo-F p(perm) Survey RED Site RED Transect (Site) Survey x Site Residual Total B. Pair-wise for the Survey x Site Interaction term; for comparisons between Surveys Comparisons between Surveys t p-value CHI B1, B B1, B B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D Comparisons between Surveys t p-value B3, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D Prepared for INPEX Cardno 66

76 Comparisons between Surveys t p-value D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D D3, D D3, D D3, D D3, D D4, D D4, D D4, D D4, D D4, D D5, D D5, D D5, D D5, D D6, D D6, D D6, D D7, D D7, D D8, D WR1 B1, B B1, B B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D B3, D Comparisons between Surveys t p-value B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D D3, D Prepared for INPEX Cardno 67

77 Comparisons between Surveys t p-value D3, D D3, D D3, D D4, D D4, D D4, D D4, D D4, D D5, D D5, D D5, D D5, D D6, D D6, D D6, D D7, D D7, D WR2 B1, B B1, B B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D1 Negative B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D D1, D Comparisons between Surveys t p-value D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D10, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D11, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D D12, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D D3, D D3, D D3, D D3, D D4, D D4, D D4, D D4, D D4, D D5, D D5, D D5, D Prepared for INPEX Cardno 68

78 Comparisons between Surveys t p-value D5, D D6, D D6, D D6, D D7, D Comparisons between Surveys t p-value D7, D D8, D C. Pair-wise for the Survey x Site Interaction term; for comparisons between Sites. Comparisons between Surveys t p-value B1 CHI, WR MC CHI, WR MC WR1, WR MC B2 CHI, WR MC CHI, WR MC WR1, WR MC B3 CHI, WR MC CHI, WR MC WR1, WR MC D1 CHI, WR MC CHI, WR MC WR1, WR MC D2 CHI, WR MC CHI, WR MC WR1, WR MC D3 CHI, WR MC CHI, WR MC WR1, WR MC D4 CHI, WR MC CHI, WR MC WR1, WR MC D5 CHI, WR MC CHI, WR MC Comparisons between Surveys t p-value WR1, WR MC D6 CHI, WR MC CHI, WR MC WR1, WR MC D7 CHI, WR MC CHI, WR MC WR1, WR MC D8 CHI, WR MC CHI, WR MC WR1, WR MC D9 CHI, WR MC CHI, WR MC WR1, WR MC D10 CHI, WR MC CHI, WR MC WR1, WR MC D11 CHI, WR MC CHI, WR MC WR1, WR MC D12 CHI, WR MC CHI, WR MC WR1, WR MC Prepared for INPEX Cardno 69

79 Appendix B-3 A. PERMANOVA analysis Multivariate analysis using PERMANOVA testing for differences in the percent cover of corals grouped by family among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p(perm) 0.05) terms in bold. RED = redundant terms due to significant interaction Source of Variation df SS MS Pseudo-F p(perm) Survey RED Site RED Transect (Site) Survey x Site Residual Total B. Pair-wise for the Survey x Site Interaction term; for comparisons between Surveys Comparisons between Surveys t p-value CHI B1, B B1, B B1, D B1, D B1, D3 Negative B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D Comparisons between Surveys t p-value B3, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D D3, D D3, D7 Negative D3, D D3, D D3, D D3, D D3, D D4, D D4, D D4, D D4, D D4, D D4, D D4, D D4, D Prepared for INPEX Cardno 70

80 Comparisons between Surveys t p-value D5, D D5, D D5, D D5, D D5, D D5, D D5, D D6, D7 Negative D6, D D6, D D6, D D6, D D6, D D7, D D7, D D7, D D7, D11 Negative D7, D D8, D D8, D D8, D D8, D D9, D D9, D D9, D D10, D D10, D D11, D WR1 B1, B B1, B B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D1 Negative B3, D2 Negative Comparisons between Surveys t p-value B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D5 Negative D3, D D3, D D3, D D3, D9 Negative D3, D D3, D D3, D D4, D D4, D D4, D D4, D8 Negative D4, D9 Negative D4, D D4, D11 Negative D4, D12 Negative D5, D D5, D D5, D D5, D D5, D D5, D D5, D D6, D D6, D D6, D Prepared for INPEX Cardno 71

81 Comparisons between Surveys t p-value D6, D D6, D D6, D D7, D D7, D D7, D D7, D D7, D D8, D D8, D D8, D D8, D D9, D D9, D D9, D D10, D D10, D12 Negative WR2 B1, B B1, B B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B1, D B2, B B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B2, D B3, D1 Negative B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D B3, D D1, D Comparisons between Surveys t p-value D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D1, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D2, D D3, D D3, D D3, D D3, D D3, D D3, D D3, D D3, D D3, D D4, D D4, D D4, D D4, D D4, D D4, D D4, D D4, D D5, D D5, D D5, D D5, D D5, D10 Negative D5, D D5, D D6, D D6, D D6, D D6, D D6, D D6, D D7, D D7, D D7, D D7, D D7, D D8, D D8, D D8, D Prepared for INPEX Cardno 72

82 Comparisons between Surveys t p-value D8, D D9, D D9, D D9, D D10, D Comparisons between Surveys t p-value D10, D D11, D C. Pair-wise for the Survey x Site Interaction term; for comparisons between Sites during Dredging survey 12. Comparisons between Surveys t p-value B1 CHI, WR MC CHI, WR MC WR1, WR MC B2 CHI, WR MC CHI, WR MC WR1, WR MC B3 CHI, WR MC CHI, WR MC WR1, WR MC D1 CHI, WR MC CHI, WR MC WR1, WR MC D2 CHI, WR MC CHI, WR MC WR1, WR MC D3 CHI, WR MC CHI, WR MC WR1, WR MC D4 CHI, WR MC CHI, WR MC WR1, WR MC D5 CHI, WR MC CHI, WR MC Comparisons between Surveys t p-value WR1, WR MC D6 CHI, WR MC CHI, WR MC WR1, WR MC D7 CHI, WR MC CHI, WR MC WR1, WR MC D8 CHI, WR MC CHI, WR MC WR1, WR MC D9 CHI, WR MC CHI, WR MC WR1, WR MC D10 CHI, WR MC CHI, WR MC WR1, WR MC D11 CHI, WR MC CHI, WR MC WR1, WR MC D12 CHI, WR MC CHI, WR MC WR1, WR MC Prepared for INPEX Cardno 73

83 Appendix B-4 A. PERMANOVA analysis Multivariate analysis using PERMANOVA testing for differences in the percent cover of corals grouped by growth form among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p(perm) 0.05) terms in bold. RED = redundant term due to significant interaction. MC = Monte Carlo simulation was used to calculate P values where unique permutations < 100 Source of Variation df SS MS Pseudo-F p(perm) Survey Site Transect (Site) Survey x Site Residual Total B. Pair-wise for the term Site Comparison between Sites t p-value CHI, WR CHI, WR WR1, WR Prepared for INPEX Cardno 74

84 Appendix B-5 A. PERMANOVA analysis Univariate analysis using PERMANOVA testing for differences in hard coral cover among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p(perm) 0.05) terms in bold. RED = redundant terms due to significant interaction Source of Variation df SS MS Pseudo-F p(perm) Survey Site Transect (Site) Survey x Site Residual Total Prepared for INPEX Cardno 75

85 Appendix B-6 A. PERMANOVA analysis Univariate analysis using PERMANOVA testing for differences in silt and sand cover among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p 0.05) terms in bold. RED = Redundant term due to significant interaction. MC = Monte Carlo simulation was used to calculate P values where unique permutations < 100. Source of Variation df SS MS Pseudo-F p(perm) Survey Site Transect (Site) Survey x Site Residual Total B. Pair-wise for the term Survey Comparisons between Survey t p-value B1, B MC B1, B MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B2, B MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC D1, D MC Comparisons between Survey t p-value D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D2, D MC D2, D MC D2, D MC Prepared for INPEX Cardno 76

86 Comparisons between Survey t p-value D2, D MC D2, D MC D2, D MC D2, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D4, D MC D4, D MC D4, D MC Comparisons between Survey t p-value D4, D MC D4, D MC D5, D MC D5, D MC D5, D MC D5, D MC D6, D MC D6, D MC D6, D MC D7, D MC D7, D MC D8, D MC C. Plot showing mean (±SE) silt and sand cover at Channel Island, Weed Reef 1 and Weed Reef 2 combined during surveys Cover (%) Survey Prepared for INPEX Cardno 77

87 Appendix B-7 A. PERMANOVA analysis Univariate analysis using PERMANOVA testing for differences in turf algae cover among Sites (CHI, WR1 and WR2) and across surveys (B1 to D12). Significant (p 0.05) terms in bold. RED = Redundant term due to significant interaction. MC = Monte Carlo simulation was used to calculate P values where unique permutations < 100. Source of Variation df SS MS Pseudo-F p(perm) Survey Site Transect (Site) Survey x Site Residual Total B. Pair-wise for the term Survey Comparisons between Surveys t p-value B1, B MC B1, B MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B2, B MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC D1, D MC D1, D MC Comparisons between Surveys t p-value D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D10, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D11, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D12, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC Prepared for INPEX Cardno 78

88 Comparisons between Surveys t p-value D2, D MC D2, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D4, D MC D4, D MC D4, D MC D4, D MC C. Pair-wise for the term Site Comparison between Sites t p-value CHI, WR CHI, WR WR1, WR Comparisons between Surveys t p-value D4, D MC D5, D MC D5, D MC D5, D MC D5, D MC D6, D MC D6, D MC D6, D MC D7, D MC D7, D MC D8, D MC D. Plot showing mean (±SE) turf algae cover at Channel Island, Weed Reef 1 and Weed Reef 2 combined during surveys: Cover (%) Survey Prepared for INPEX Cardno 79

89 E. Plot showing mean (±SE) turfing algae cover at each site across all surveys combined Cover (%) Channel Island Weed Reef 1 Weed Reef 2 Site F. Scatter plot of turf algae and silt and sand cover at Channel Island, weed Reef 1 and Weed Reef 2 across all surveys combined 100 y = x R² = Silt and Sand Cover (%) Turfing Algae Cover (%) Prepared for INPEX Cardno 80

90 Appendix B-8 Univariate analysis using PERMANOVA testing for differences in the number of coral recruits ( 2 cm) among Sites (CHI, WR1, WR2) and across surveys (B1 to D12). n = 4. Significant (p(perm) 0.05) terms in bold. RED = redundant terms due to significant interaction. MC = Monte Carlo simulation was used to calculate P values where unique permutations < 100. A. PERMANOVA analysis for number of coral recruits Source of Variation df SS MS Pseudo-F p(perm) Survey Site Transect (Site) Survey x Site Residual Total B. Pair-wise for the term Survey Comparisons Between Surveys t p-value B1, B MC B1, B MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B1, D MC B2, B MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B2, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC B3, D MC D1, D MC Comparisons Between Surveys t p-value D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D1, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D2, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D3, D MC D4, D MC D4, D MC D4, D MC D4, D MC D4, D MC D4, D MC D4, D MC D4, D MC D5, D MC D5, D MC D5, D MC Prepared for INPEX Cardno 81

91 Comparisons Between Surveys t p-value D5, D MC D5, D MC D5, D MC D5, D MC D6, D MC D6, D MC D6, D MC D6, D MC D6, D MC D6, D MC D7, D MC D7, D MC D7, D MC Comparisons Between Surveys t p-value D7, D MC D7, D MC D8, D MC D8, D MC D8, D MC D8, D MC D9, D MC D9, D MC D9, D MC D10, D MC D10, D MC D11, D MC C. Plot showing mean (±SE) number of coral recruits at all sites combined during each survey Number of Recruits Survey Prepared for INPEX Cardno 82

92 APPENDIX C RAW DATA Prepared for INPEX Cardno 83

93 Appendix C-1 A. Likely perceived mortality Mean (±SE) percentage cover of sub-categories contributing to total mortality for B0 (average of surveys B1 to B3) and D12 Sediment Mean SE Mean SE Channel Island Weed Reef Weed Reef Entire Colony Sediment Burial Channel Island Weed Reef Weed Reef Macroalgae Channel Island Weed Reef Weed Reef Mobile Fauna Channel Island Weed Reef Weed Reef Total Perceived Dead Coral Channel Island Weed Reef Weed Reef B0 D12 Prepared for INPEX Cardno 84

94 B. Actual mortality B0 D12 Total Dead Coral Mean SE Mean SE Channel Island Weed Reef Weed Reef Entire Colony Mortality Channel Island Weed Reef Weed Reef Missing section Channel Island Weed Reef Weed Reef Turf algae Channel Island Weed Reef Weed Reef Encrusting algae Channel Island Weed Reef Weed Reef Immobile Fauna Channel Island Weed Reef Weed Reef Total Actual Dead Coral Channel Island Weed Reef Weed Reef Prepared for INPEX Cardno 85

95 Appendix C-2 Paver sediment depth and coral sediment cover on adjacent pavers and corresponding tagged colony in Dredging surveys D10 and D12 Site Colony Family D11 D12 Paver Sediment Depth (mm) Coral Sediment Cover (%) Paver Sediment Depth (mm) Coral Sediment Cover (%) CHI 1 Faviidae Faviidae Poritidae 4.0 No CPCe Pectiniidae Pectiniidae Dendrophylliidae Poritidae Poritidae WR1 3 Dendrophylliidae Pectiniidae Poritidae Dendrophylliidae Faviidae Pectiniidae Faviidae Poritidae WR2 1 Faviidae Faviidae Poritidae Pectiniidae Pectiniidae Poritidae 3.0 No CPCe 2.0 No CPCe 21 Dendrophylliidae Dendrophylliidae Prepared for INPEX Cardno 86

96 Appendix C-3 Coral surface area in pixels and percentage change in surface areas of tagged colonies between Baseline survey B3 and Dredging surveys D3, D4, D5, D6, D7 and D12 Colony Surface Area (pixels) Percent Change B3 D3 D4 D5 D6 D7 D12 B3 to D3 B3 to D4 B3 to D5 B3 to D6 B3 to D7 B3 to D12 CHI CHI CHI CHI CHI WR WR WR WR WR WR WR WR WR WR WR WR WR WR Prepared for INPEX Cardno 87

97 Appendix C-4 HARD CORALS Benthic composition (%) of the different sites in field survey D11, including coral families and growth form, algae, seagrass and other invertebrates CHI WR1 WR2 Mean SE Mean SE Mean SE Acroporidae Encrusting Foliose Agariciidae Foliose Dendrophylliidae Encrusting Foliose Euphyllidae Other Faviidae Encrusting Foliose Massive Submassive Fungiidae Encrusting Foliose Solitary Merulinidae Encrusting Foliose Submassive Oculinidae Encrusting Submassive Pectiniidae Encrusting Foliose Submassive Poritidae Encrusting Massive Submassive Siderastreidae Encrusting Submassive Growth Forms Encrusting Foliose Massive Prepared for INPEX Cardno 88

98 CHI WR1 WR2 Mean SE Mean SE Mean SE Solitary Submassive SOFT CORALS Leather Tree Coral Sea Whip Other ALGAE Encrusting Macroalgae Turf Algae OTHER BIOTA Anemones Ascidian Bryozoa Hydroids Sponge Tube Worm Mucus Tubes SUBSTRATA Rock Rubble Sand/Silt Prepared for INPEX Cardno 89

99 Appendix C-5 Number of coral recruits (<20 mm) at different sites during Baseline and Dredging Phase surveys. Blanks indicate sites not sampled Survey CHI WR1 WR2 SSI NEW CHP MAN Mean SE Mean Mean SE Mean Mean SE Mean SE Mean SE Mean SE B B B D D D D D D D D D D D D Prepared for INPEX Cardno 90

100 APPENDIX D TAGGED CORALS CONDITION: SUPPLEMENTARY MATERIAL Prepared for INPEX Cardno

101 Appendix D-1 a) CHI TAG 03 Time series from survey B1 to D11 of tagged corals that have recorded between 95% and 100% mortality B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 92

102 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Prepared for INPEX Cardno 93

103 D10 Mar 14 D11 Apr 14 D12 May 14 b) CHI TAG 08 B1 Jun 12 B2 Jul 12 B3 Aug 12 Prepared for INPEX Cardno 94

104 D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 D6 Aug 13 Prepared for INPEX Cardno 95

105 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Prepared for INPEX Cardno 96

106 c) CHI TAG 10 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 NO PHOTO Prepared for INPEX Cardno 97

107 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Removed from Survey Prepared for INPEX Cardno 98

108 d) CHI TAG 31 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 99

109 D4 Apr 13 D5 Jun 13 D6 Aug 13 NO PHOTO D7 Oct 13 D8 Dec 13 D9 Feb 14 Removed from Survey Prepared for INPEX Cardno 100

110 e) CHI TAG 61 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 101

111 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Prepared for INPEX Cardno 102

112 D10 Mar 14 D11 Apr 14 Missing Colony f) CHI TAG 64 B1 Jun 12 B2 Jul 12 B3 Aug 12 Prepared for INPEX Cardno 103

113 D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 D6 Aug 13 Prepared for INPEX Cardno 104

114 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Prepared for INPEX Cardno 105

115 g) WR1 TAG 62 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 106

116 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Removed from Survey Prepared for INPEX Cardno 107

117 h) WR2 TAG 02 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 108

118 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Prepared for INPEX Cardno 109

119 D10 Mar 14 D11 Apr 14 D12 May 14 i) WR2 TAG 10 B1 Jun 12 B2 Jul 12 B3 Aug 12 Prepared for INPEX Cardno 110

120 D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 D6 Aug 13 Prepared for INPEX Cardno 111

121 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Prepared for INPEX Cardno 112

122 j) WR2 TAG 20 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 113

123 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Prepared for INPEX Cardno 114

124 D10 Mar 14 D11 Apr 14 D12 May 14 k) WR2 TAG 26 B1 Jun 12 B2 Jul 12 B3 Aug 12 Prepared for INPEX Cardno 115

125 D1 Oct 12 D2 Dec 12 D3 Feb 13 D4 Apr 13 D5 Jun 13 D6 Aug 13 Prepared for INPEX Cardno 116

126 D7 Oct 13 D8 Dec 13 D9 Feb 14 D10 Mar 14 D11 Apr 14 D12 May 14 Prepared for INPEX Cardno 117

127 l) WR2 TAG 45 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 118

128 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Removed from Survey Prepared for INPEX Cardno 119

129 m) WR2 TAG 46 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 120

130 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Removed from Survey Prepared for INPEX Cardno 121

131 n) WR2 TAG 47 B1 Jun 12 B2 Jul 12 B3 Aug 12 D1 Oct 12 D2 Dec 12 D3 Feb 13 Prepared for INPEX Cardno 122

132 D4 Apr 13 D5 Jun 13 D6 Aug 13 D7 Oct 13 D8 Dec 13 D9 Feb 14 Prepared for INPEX Cardno 123

133 D10 Mar 14 D11 Apr 14 D12 May 14 Prepared for INPEX Cardno 124

134 Appendix D-2 Coral Health Assessment based on Surveys B1 to D8 Tumours and other growth anomalies A tumour is defined as a raised lump of undifferentiated tissue, typically without polyps and often paler in colour compared with healthy tissue (Figure D-1a). Growth anomalies are also raised lumps of tissue, but they typically have polyps and no colour differentiation (Figure D-1b). They may be caused by different agents, such as borers or turfing algae. a) b) Figure D-1 Tumours and growth anomalies on Dendrophyllidae; a) Channel Island Tag 8 showing evidence of tumours; and b) Weed Reef 2 Tag 67 showing growth anomalies Tumours were identified only on two corals, both from the family Dendrophyllidae (Table D-1). The tumours in each case were present from the first survey (B1). At Channel Island, Tag 8, almost the entire colony died between surveys D2 and D3. While Northeast Wickham Point Tag 69 is still alive, mortality has increased by almost 60%. Table D-1 Occurrence of coral tumours in Darwin Harbour Family Site Tag Number First Record Dendrophyllidae Maximum Number Change in Mortality (%) Channel Island 08 B % Northeast Wickham Point 69 B % Coral growth anomalies (Table D-2) are more prevalent than tumours. Once again the Dendrophyllidae were primarily affected, with 18 of the 20 records belonging to this family. Prepared for INPEX Cardno 125

135 Table D-2 Occurrence of coral growth anomalies in Darwin Harbour, the maximum number of growth anomalies recorded on each tagged coral and the percentage change in mortality between the survey when the first growth anomaly was recorded and D10 Family Site Tag Number First Record Maximum Number Change in Mortality (%) Dendrophyllidae Channel Island 30 D B Charles Point 20 B B Mandorah 43 B B Northeast Wickham Point 2 B B B B Weed Reef 1 11 B B B B Weed Reef 2 21 B B B B Faviidae Weed Reef 2 4 B Poritidae Weed Reef 2 2 D D4 mortality data used. 2 D7 mortality data used. 3 Complete Mortality in D8. Macroborers Macroborers are organisms that take shelter within the primary framework of a coral colony and are one class of a community of bioeroders that erode the calcium carbonate coral skeleton. Other bioeroders include microborers such as algae, bacteria and fungi, and grazers such as fish and urchins. The numbers of borers were assessed for each of the tagged corals at each of the sites throughout monitoring (B1 to D8). Borers were predominantly associated with corals of the family Faviidae (Table D-3) and to a lesser extent with the families Agariciidae, Dendrophyllidae, Fungiidae, Pectinidae, Poritidae and Siderastreidae. Borers appear to be more prevalent in corals of the family Faviidae (Figure D-2). Given both the high prevalence of borers and their role as bioeroders on reefs, we tested whether there was an associated increase in coral mortality with increased numbers of borers. The change in mortality from Baseline to D10 was plotted as a function of number of borers (Figure D-3). No relationship between the number of borers occurring on a coral and percentage change in mortality was found (R 2 =0.001, p=0.705), indicating that the presence of borers is not adversely affecting the health of the corals as measured by mortality. Prepared for INPEX Cardno 126

136 Table D-3 Numbers of tagged corals where borers were present across sites and families. Total number corals of each family shown in brackets Family Channel Island Charles Point Mandorah Northeast Wickham Point South Shell Island Weed Reef 1 Weed Reef 2 Agariciidae 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1 (1) 0 (0) Dendrophyllidae 2 (7) 1 (23) 3 (39) 3 (12) 0 (1) 0 (14) 0 (14) Faviidae 19 (33) 19 (39) 14 (34) 19 (43) 0 (5) 22 (29) 6 (25) Fungiidae 0 (0) 0 (0) 0 (0) 1 (1) 0 (1) 2 (2) 2 (3) Merulinidae 0 (0) 0 (0) 0 (0) 0 (0) 0 (1) 0 (1) 0 (0) Mussidae 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (1) 0 (0) Oculinidae 0 (0) 0 (1) 0 (0) 0 (0) 0 (1) 0 (0) 0 (0) Pectinidae 4 (26) 0 (0) 0 (0) 3 (11) 1 (54) 2 (21) 0 (16) Poritidae 2 (9) 3 (12) 1 (2) 2 (7) 0 (12) 0 (5) 1 (16) Siderastreidae 0 (0) 0 (0) 0 (0) 1 (1) 0 (0) 0 (1) 1 (1) Borers No Borers Number of Corals Coral Family Figure D-2 Numbers of tagged corals with and without borers across families Prepared for INPEX Cardno 127

137 % Change in Mortality Agariciidae Dendrophyllidae Faviidae Fungiidae Pectinidae Poritidae Siderastreidae Numbers of Borers Figure D-3 Percentage change in coral mortality from the Baseline Phase to D10, plotted as a function of the number of borers present on tagged corals categorised by coral family. A negative value represents a decrease in mortality for the period Prepared for INPEX Cardno 128

138 Turf Algae Turf algae is often a rapid coloniser following stress or other causes of tissue mortality; however it is unclear it is unclear whether turf algae itself can cause mortality of coral tissue as there is limited documented evidence of this occurring. Once turf algae is established on a coral, it is likely to cause stress and tissue mortality and may expand across the colony. One mechanism for this spread is enhanced microbial activity at the coral algal interface due to chemical compounds released by the algae. Turf algae has been widely prevalent on corals in Darwin Harbour, with between 40% of tagged corals having turf algae present at Weed Reef 2 and 71% at Northeast Wickham Point during the Baseline Phase (Figure D-4). Turf algae has been recorded in surveys after the Baseline Phase on corals not previously colonised, but generally only for a few colonies each Survey at any site. The fact that so many colonies had turf algae present during the baseline, and yet only a few have gone on to show substantial or complete mortality, indicates that the presence of turf algae does not necessarily cause coral mortality. 80% 70% CHI CHP MAN NEW SSI WR1 WR2 60% 50% 40% 30% 20% 10% 0% Baseline D1 D2 D3 D4 D5 D6 D7 D8 Figure D-4 Percentage of tagged corals at each site when turf algae was first observed on the colony Prepared for INPEX Cardno 129

139 APPENDIX E WATER QUALITY DATA Prepared for INPEX Cardno 130

140 Appendix E-1 Daily mean, maximum and minimum water temperature data from Channel Island, Weed Reef 1 and Weed Reef 2, between 11/04/2014 and 11/05/2014. Extract from Cardno Water Quality database 22/05/2014 Prepared for INPEX Cardno 131

141 Appendix E-2 Daily mean, maximum and minimum water turbidity data from Channel Island, Weed Reef 1 and Weed Reef 2 between 11/04/2014 and 11/05/2014. Extract from Cardno Water Quality database 22/05/2014 Prepared for INPEX Cardno 132