Gladstone Fish Health Investigation 2011 - 2012 - Western Basin ...

Department of Agriculture, Fisheries and ForestryGladstone HarbourFish Health Investigation2011–2012Great state. Great opportunity.

Front cover photo supplied by Johnny Mitchell, as featured in his report Time line BarraThis publication has been compiled by Dr Stephen Wesche, Dr Tim Lucas, Dr David Mayer, DavidWaltisbuhl and Dr Ross Quinn of the Department of Agriculture, Fisheries and Forestry.© State of Queensland, 2013.The Queensland Government supports and encourages the dissemination and exchange of itsinformation. The copyright in this publication is licensed under a Creative Commons Attribution 3.0Australia (CC BY) licence.Under this licence you are free, without having to seek our permission, to use this publication inaccordance with the licence terms.You must keep intact the copyright notice and attribute the State of Queensland as the source of thepublication.For more information on this licence, visit http://creativecommons.org/licenses/by/3.0/au/deed.enThe information contained herein is subject to change without notice. The Queensland Governmentshall not be liable for technical or other errors or omissions contained herein. The reader/user acceptsall risks and responsibility for losses, damages, costs and other consequences resulting directly orindirectly from using this information.

ContentsTable of tables 1Table of figures 1Executive summary 5Phase 1 (August 2011 – February 2012) 6Phase 2 (April 2012 – September 2012) 6Barramundi 6Other species 6Conclusions 7Background 9Stocking in Lake Awoonga 10Calliope and Boyne Rivers flooding event 10Recreational catch 11Phase 1: Fish Health Survey – Gladstone Harbour (August 2011–February 2012) 13Overview 13Study area 13Candidate species 14Sampling regime 14Data collection and processing 15Phase 2: Expanded Gladstone Fish Health Survey (March 2012–September 2012) 16Overview 16Conceptual model 16Study area 17Candidate species 17Sampling regime 18Data collection and processing 18Summary of samples collected 21Significant findings and discussion 24Analysis of all species combined 24Barramundi 25Commercial catch 25Field observational Phase 1 27Laboratory findings Phase 1 28Observational findings Phase 2 29Targeted barramundi surveys to examine Neobenedenia sp. infection 32Laboratory findings Phase 2 34Mud crabs 37Commercial catch 37Crab size structure 38Shell disease and abnormalities 41i

Observational findings Phase 1 42Laboratory testing Phase 1 43Laboratory findings Phase 2 44Mullet 46Commercial catch 47Observational findings Phase 1 48Histopathology and residue testing Phase 1 50Observational findings Phase 2 50Laboratory findings Phase 2 53Sharks and rays 54Commercial catch 54Observational findings Phase 1 56Laboratory testing Phase 1 57Observational findings Phase 2 58Laboratory findings Phase 2 60Banana prawns 60Commercial catch 61Observational findings Phase 1 63Laboratory testing Phase 1 63Observational findings Phase 2 64Laboratory testing Phase 2 65Trawl caught finfish species 65Observational findings Phase 1 65Laboratory testing Phase 1 66Observational findings Phase 2 66Laboratory findings Phase 2 66Pelagics 67Observational findings Phase 1 67Histopathology and residue testing Phase 1 68Observational findings Phase 2 69Histopathology and residue testing Phase 2 70Other species 70Scallops 70Finfish 72Conclusions 74Barramundi 74Mullet and other finfish 75Sharks 75Mud crabs 76Summary 76References 77Appendix A – Methods for Expanded Fish Health Sampling Program 79Animal ethics 79Study area 79Candidate species 80Barramundi (Lates calcarifer) 80- ii -

Mullet (Mugilidae) 80Banana prawn (Fenneropenaeus merguiensis) 80Mud crab (Scylla serrata) 80Bull shark (Carcharhinus leucas) 80Trawl species: Grinner (Saurida sp), Australian threadfin (Polydactylus multiradiatus) andCastelnau’s herring (Herklotsichthys castelnaui) 80Pelagic species: Queenfish (Scomberoides sp.) 80Sampling regime 81Timing 81Gear type 81Sampling 83Field observations 83Samples selected for further processing 83Sample processing 84Step 1: Observations and data recorded by field staff at the time of capture. 84Mullet, barramundi, queenfish and sharks 84Mud crabs 84Banana prawns 85Herring/grinner/Australian threadfin 86Specimen identification number 87Step 2: Laboratory processing 88Mullet, barramundi, queenfish and sharks 88Trawl Species: Banana prawns, herring/grinner/Australian threadfin 89Mud crabs 89Appendix A-1: Eye and Skin lesion grading system 91Appendix A – 2 Necropsy data sheets 99Appendix A – 3 Fish health assessment variables 100Appendix B – Statistical report: Gladstone Harbour sampling trips 2012 102Appendix C – Statistical report: Histology results 129Appendix D – Statistical analysis chemical residues 137Appendix E – Time line Barra 138December, 2010- The Overflow 139January, February, March- Mystery and Intrigue. 140April- Booming 143May/June- Fins 144July/ August- Getting Ugly 147September- Deaths and Sickness 150October 152November 153- iii -

Table of tablesTable 1. Number of fish from each site and sampling trip used for histopathology andchemical residue testing. Note: 1 Denotes Gladstone Harbour and adjacent waterways, 2denotes reference sites. Note due to low sample numbers, sharks were used for histologyonly. 21Table 2. The number of fish, crustaceans and molluscs observed during Phase 1 and 2 ofthe Gladstone Fish Health Investigation. Note * indicates multiple species included incategory 22Table 3. Numbers of samples tested by Biosecurity Queensland during Phase 1. 23Table 4. Numbers of samples used for necropsy examination during Phase 2. 23Table 5. Highest lesion grade for each crab as assessed by Fisheries Queensland in the fieldand Biosecurity Queensland in the laboratory. The grey numbers indicate agreementbetween the two assessments. 44Table 6. Mean values and standard errors (s.e.) for the crab categories. 45Table 7. Mean values and average standard errors for hepato-somatic index and per centabnormal for each location and sampling event. 45Table 8. Lesion categories as observed in the field, according to categories described byAndersen, 2001. Where individual crabs had more than one lesion, only the most severelesion was included. 45Table 9. The sharks and rays observed during Phase 1. 56Table of figuresFigure 1. A school of large fish in the Boyne River detected on a fish finder duringrecreational fishing in 2011. 11Figure 2. Sampling sites within the principal study area. 14Figure 3. The annual catch and catch per unit effort of barramundi from the Gladstone region(CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 – 2012). 26Figure 4. The number of days fishing where commercial fishers have reported barramundifrom the Gladstone region (CFish grid S30), as reported in commercial fisher logbooks(calendar years 2000 – 2012). 26Figure 5. Skin condition observed in barramundi at each site sampled during Phase 1 of fishhealth investigations, September 2011 – February 2012. Note, (n) refers to the total numberof fish examined at each site. 27Figure 6. Eye condition observed in barramundi at each site sampled during Phase 1 of fishhealth investigations, September 2011 – February 2012. Note, (n) refers to the total numberof fish examined at each site. 281

Figure 7. Barramundi caught in the Boyne River showing physical damage consistent withbeing washed over the Awoonga spillway 29Figure 8. Conditions observed (skin colour and lesions) in barramundi sampled at each siteduring April 2012, Phase 2. Note, (n) refers to the total number of fish examined at each site.Some fish may have been observed with multiple conditions. 30Figure 9. Barramundi caught while trawling in the Gladstone Harbour. 30Figure 10. A barramundi caught in the Boyne River during the June/July 2012 samplingevent with recovering injuries to the jaw and body. 31Figure 11. A barramundi caught in the Burnett River, Bundaberg, during the June/July 2012sampling event with an old injury in an advanced stage of healing. Note the new skincompletely covering the old injury with new scales forming over the affected area. 31Figure 12. A barramundi caught in the Calliope River during the June/July 2012 samplingevent with an old injury in an advanced stage of healing. Note the new skin completelycovering the old injury. 32Figure 13. Conditions observed (skin colour and lesions) in barramundi sampled at each siteduring June/July, Phase 2. Note, (n) refers to the total number of fish examined at each site.Some fish may have been observed with multiple conditions. 32Figure 14. Levels of mesentery fat in barramundi from Gladstone sites, according to theirlesion scores as recorded during necropsy. 35Figure 15. The annual catch of mud crabs from the Gladstone region (CFish grid S30), asreported in commercial fisher logbooks (calendar years 2000 – 2012). 37Figure 16. The number of days fishing where commercial fishers have reported mud crabfrom the Gladstone region (CFish grid S30), as reported in commercial fisher logbooks(calendar years 2000 – 2012). 38Figure 17. The size structure of mud crabs observed at each site during the Gladstone FishHealth Monitoring Program Phase 1, December 2011 – January 2012. Note, (n) refers to thetotal number of crabs examined at each site. 39Figure 18. The size structure of mud crabs observed in the Gladstone region (sitescombined) and the reference sites during the Gladstone Fish Health Monitoring ProgramPhase 2, April 2012 – July 2012. Note, (n) refers to the total number of crabs examined ateach site. 39Figure 19. The size structure of mud crabs observed at each site during the Gladstone FishHealth Monitoring Program, Phase 2, April 2012 – July 2012. Note, (n) refers to the totalnumber of crabs examined at each site. 40Figure 20. The size structure of mud crabs observed at sites monitored during a fisheryindependent sampling program conducted along the Queensland coastline 2000 - 2009.Note, (n) refers to the total number of crabs examined at each site. 41Figure 21. Examples of shell abnormalities (circled) observed in mud crabs during theGladstone Fish Health Monitoring Program. 422

Figure 22. The percentage of mud crabs with shell abnormalities observed at all sites duringthe Gladstone Fish Health Monitoring Program, Phase 1. Note, (n) refers to the total numberof crabs examined at each site. 42Figure 23. The combined percentage of mud crabs with shell abnormalities observed at allsites during the Gladstone Fish Health Monitoring Program, Phase 2. Note, (n) refers to thetotal number of crabs examined at each site. 43Figure 24. Mullet species encountered during sampling (a) sea mullet; (b) diamond scalemullet; (c) goldspot mullet; (d) bluespot mullet. 47Figure 25. The annual catch of mullet from the Gladstone region (CFish grid S30), asreported in commercial fisher logbooks (calendar years 2000 – 2012). 47Figure 26. The number of days fishing where commercial fishers have reported mullet catchfrom the Gladstone region (CFish grid S30), as reported in commercial fisher logbooks(calendar years 2000 – 2012). 48Figure 27. Conditions observed at each site sampled during Phase 1, September 2011 –January 2012. Note, (n) refers to the total number of fish examined at each site. Some fishmay have been observed with multiple conditions. 49Figure 28. (a) Mullet caught in Rodds Bay displaying minor redness/pinpoint marks on theventral surface and caudal peduncle. The redness around the operculum was a result of thecapture by net; (b) Mullet caught in the Calliope River showing general redness; (c) Mulletcaught in Rodds Bay with ulcerative lesion on the caudal peduncle. 50Figure 29. Minor redness and pinpoint marks observed on fins and the body surface of mulletsampled from the Fitzroy River (a) and (b), and Bundaberg (c) and (d). 51Figure 30. A mullet captured in the lower reaches of the Boyne River with a fresh bite mark 51Figure 31. A mullet sampled from the upper reaches of the Boyne River displaying generalredness and lesions comprised of small areas with detached scales. 52Figure 32. A mullet sampled from the Fitzroy River with ulcerative lesions on (a) the left and(b) right side of the gill covering (operculum), exposing the gills. Note also the cloudy righteye 52Figure 33. The combined results for conditions observed in mullet at all sites sampled duringPhase 2, 2012. Note, (n) refers to the total number of fish examined at each site. Some fishmay have been observed with multiple conditions. 53Figure 34. Bull shark 54Figure 35. The annual catch of shark from the Gladstone region (CFish grid S30), asreported in commercial fisher logbooks (calendar years 2000 – 2011). Data includes both lineand net catch for the region combined. 55Figure 36. The number of days fishing where commercial fishers have reported shark catchfrom the Gladstone region (CFish grid S30), as reported in commercial fisher logbooks(calendar years 2000 – 2011). Data includes both line and net effort for the region combined.55Figure 37 Conditions observed in bull sharks at the time of capture from all sites sampledduring April 2012. Note, (n) refers to the total number of sharks examined at each site. Somesharks may have been observed with multiple conditions. 583

Figure 38. Sharks caught in the Calliope River illustrating the increase in skin redness postmortem. (a) and (c) Shark #0005 and #0007 photographed in the Calliope River at the timeof capture; (b) and (d) The same sharks photographed four hours later in the lab. 59Figure 39. Bull sharks caught at Bundaberg and the Calliope River showing evidence of ectoparasites(lighter areas) between the dorsal fins. 59Figure 40. Banana prawns, a targeted species during the sampling program. 61Figure 41. The annual catch of banana prawns from the Gladstone region (CFish grid S30)as reported in commercial fisher logbooks from calendar years 2000 – 2011. Data includesboth otter trawl and beam trawl catch for the region. 62Figure 42. The number of days fishing where commercial fishers have reported bananaprawn catch from the Gladstone region (CFish grid S30), as reported in commercial fisherlogbooks (calendar years 2000 – 2011). Data includes both otter trawl and beam trawl catcheffort for the region. 62Figure 43. Banana prawns collected during the sampling program from (a) the GladstoneHarbour and (b) offshore Fitzroy River sites showing signs of shell erosion (circled). 64Figure 44. Banana prawns collected during the sampling program with the isopod presentunder the carapace (a) Prawn caught from the offshore Fitzroy River site with an isopodpresent under the carapace (circled); (b) Prawn caught at Bundaberg with part of itscarapace removed to reveal the isopod. 65Figure 45. Queenfish caught at Bundaberg during the sampling program. 67Figure 46 . Conditions observed in queenfish and other pelagic species at each site duringPhase 1. Note, (n) refers to the total number of fish examined at each site. 68Figure 47. Conditions observed in queenfish and other pelagic species at each site duringPhase 2. Note, (n) refers to the total number of fish examined at each site. Some fish mayhave been observed with multiple conditions. 704

Executive summaryThe Gladstone Harbour Fish Health Investigation occurred in response to public concernsregarding fish health in Gladstone Harbour, which were raised with the former Department ofEmployment, Economic Development and Innovation (now Department of Agriculture,Fisheries and Forestry (DAFF)) in August 2011. The majority of concerns were regardingskin redness, lesions and eye damage observed on barramundi caught from the Boyne Riveror within its vicinity. Further reports included unusual skin redness and occasional lesionsobserved on a range of fish species from Gladstone Harbour (including sharks), as well asan increase in the incidence of shell erosion on mud crabs. The subsequent investigationaimed to determine the cause of the reported issues and specifically to address publicconcern regarding the potential impact of industrial activity including dredging in GladstoneHarbour. The investigation, conducted by DAFF, was part of a whole-of-governmentresponse, which included sediment and water quality testing; and investigation into humanhealth concerns.The first reports of abnormalities in fish were received months after a major flood eventhappened in Gladstone from December 2010 until early 2011. The flood event caused adramatic influx of large barramundi (estimated 30 000), as well as other species from LakeAwoonga into the Boyne River when the Awoonga Dam spilled over. The dam had not spilledover since the 1990s, well before the dam wall was raised in 2002. Many of these fish wereinjured (some fatally) from the force of the impact when passing over the spillway. It isbelieved that a smaller number of fish were washed over during a second flood event in early2012.The initial response investigated fish with visible abnormalities (e.g. missing scales andredness), and signs of disease to indicate the cause of the observed issues. This earlysampling (Phase 1) between August 2011 and February 2012 was based on observations ofcommercial fishing activity, contracted fish surveys and submissions by members of thepublic. Over 5000 fish, crustaceans and molluscs were visually assessed during thissampling.In January 2012, the Independent Gladstone Fish Health Scientific Advisory Panel releaseda report recommending a structured sampling program to gauge fish health in GladstoneHarbour sites against comparable unaffected sites (reference sites), through time. It alsorecommended the development of a conceptual model of possible cause-effect relationshipsto help guide studies and eliminate potential causal factors.Structured sampling in Gladstone Harbour (Phase 2) was conducted in two major surveysduring April–May (Trip 1) and June–July (Trip 2) 2012. These surveys covered 11 sitesincluding two reference sites and focussed on a subset of 10 species. The selected speciesrepresented a range of different life cycles (e.g. catadromous and estuarine) and trophiclevels (e.g. predatory, omnivorous detritivores and scavengers), and had been reported witha variety of abnormalities. A third targeted survey was conducted in September 2012 tomonitor potential seasonal reoccurrence of parasitic infestations in the barramundi from theBoyne River. A total of 3699 fish and crustaceans were visually assessed in the field, with452 specimens examined by necropsy, and tissues from 120 specimens subjected tohistopathology and chemical residue testing.5

Phase 1 (August 2011–February 2012)Both commercial catch records as well as field observations confirmed there was anabnormally large population of barramundi in the Boyne River, with many of these fish beingexceptionally large. Recreational fishers indicated that the increased population ofbarramundi had been present since the flooding event in early 2011, with fish showing signsof damage but also recovery at first. During winter, when the water temperature was reportedto be as low as 18.1 o C, anecdotal evidence suggested deteriorating health in the barramundifrom the Boyne River including swollen, opaque eyes and ulcerated areas on the skin. Theinvestigation found that the fish were suffering from skin lesions and parasitic infestation(monogeneans on the body surface), conditions not previously documented from fish in awild environment, but are often recorded from fish held in heavily stocked aquaculturesystems.Other finfish species were observed with a range of generally mild skin conditions, includnglocalised skin inflammation, scale loss, skin erosion and redness. A fungal infection thatcauses skin lesions (Epizootic Ulcerative Syndrome (EUS) or ‘red-spot’ disease) wasidentified in two fish, both of which were outside of Gladstone Harbour. EUS was excludedas a potential cause of the lesions seen in Gladstone Harbour. No bacterial, parasitic orfungal pathogens were found that could explain the skin conditions.Phase 2 (April 2012–September 2012)BarramundiThe barramundi from the Boyne River caught in April 2012 were observed to have highprevalence of skin lesions thought to be caused by physical trauma, while fish captured inSeptember 2012 were observed with lesions that were healing. Infestations of themongenean parasite identified in Phase 1 were not observed during the Phase 2 structuredsampling.Based on weight-length relationships, mesentery fat assessment and livermelanomacrophage centres, barramundi sampled from Gladstone Harbour sites were in poornutritional condition compared with those from the reference sites. There were somebarramundi caught during sampling with relatively high levels of mesentery fat, indicatingexcessive intake of food. These fish tended to be found in the Boyne River and had lesionstypical of physical trauma caused by fish passing over a spillway, suggesting that theyprobably originated from Lake Awoonga.Chemical residue analysis of barramundi livers showed that fish collected from GladstoneHarbour had significantly higher concentrations of iron and cadmium during April 2012, andhigher arsenic and zinc levels than the reference sites in June/July 2012 (P ≤ 0.05).Comparison of concentrations of metals in barramundi livers with toxicological effects dataindicated that the concentrations in the barramundi liver were of no concern. In addition,there were no significant differences in these metal concentrations between fish with andwithout visible abnormalities.Other speciesSpecies other than barramundi were generally observed to be in good condition. Someabnormalities were observed in other species (e.g. shell lesions in mud crabs), howeveranalysis of the data indicated there was no evidence that these conditions were moreprevalent in fish or crustaceans from Gladstone Harbour when compared to those from thereference sites, or when compared with historical data.6

Areas of detached scales associated with parasitic infection were observed on a number ofshark species caught during the sampling program from Gladstone Harbour and thereference sites. It is unknown whether the numbers of parasites observed on sharks duringthis program are abnormal, however it is evident that these parasites and associatedpathology are not unique to Gladstone. Severe skin redness was observed during necropsy,however this was not apparent in freshly caught sharks; redness developed in the hours afterdeath. This post-mortem redness was not unique to Gladstone.Large numbers of prawns and fish were caught on trawlers both in Gladstone Harbour andthe reference sites, with no significant redness or other abnormalities observed.No significant abnormalities were identified from the sampling and necropsy examination ofpelagic species.Histopathology investigation indicated that gill, liver and skin tissue of grinner, and the gill,hepatopancreas and muscle tissue of mud crabs was similar in Gladstone samples toreference sites. Some bleeding and swelling was observed in shark skin from both Gladstoneand reference sites.Chemical residue testing of crabs from both Gladstone and reference sites found asignificant association between aluminium and selenium concentrations, and shell lesions.Given that concentrations of metals in mud crab hepatopancreas were similar to historicaldata from sites across the east coast of Queensland, any effect of these metals on shelllesions is unlikely to be unique to Gladstone. The available data also indicated that the suiteof organic chemicals tested did not contribute to crab disease.Chemical residue testing of muscle tissue from grinner showed no significant differencebetween Gladstone and reference sites.ConclusionsIt is obvious from all the reports that something happened to aquatic life in GladstoneHarbour in 2011. The findings provided in this study indicate that the fish health in GladstoneHarbour had returned to a more normal situation in 2012. Other than the physical damage orrecovery from physical damage, this study found the sampled fish in a health status similar tothat observed in the Bundaberg and Fitzroy reference sites.All industrial activities in Gladstone that were operating in Gladstone in 2011 have continuedto operate, including dredging. The only factor different in 2011 compared to previous yearswas the significant rainfall, the subsequent flooding and the introduction of a significantbiomass of fish from Lake Awoonga. Flooding has also occurred in 2012 and 2013.However, other than continuing reports of barramundi showing physical damage as the resultof being washed over the Awoonga spillway in these flooding events, reports of anyreoccurence of sick fish have not been received by DAFF.The most likely cause of the abnormalities observed during this investigation is the suddenintroduction of barramundi and other fish from Lake Awoonga during December 2010 to early2011. Although barramundi are naturally able to move between fresh and salt waterenvironments, it appears that these fish were overcrowded and unable to feed normally afterthey were suddenly re-located from Lake Awoonga into the Boyne River. The stresspreceded parasitic infestation during the colder winter water temperatures in 2011. TheNeobenedenia infestation subsided as the water temperature increased and fish densityreduced, as shown by improving fish health in late 2011 and 2012.7

The results in this study support that it was flooding combined with the introduction of asignificant biomass of fish (including barramundi, mullet, catfish and bony bream) thatstressed the ecosystem in Gladstone Harbour and adjacent waterways. This study cannotrule out the possibility that the activity of dredging and associated turbidity providedadditional stress to the ecosystem, but it was not the primary stressor.8

BackgroundIn August 2011, the Queensland Government received reports of barramundi andsubsequently other species being caught with obvious signs of ill health, includingbulging/red eyes, blindness, severe skin lesions and skin discolouration.On 16 September 2011, Fisheries Queensland, in the former Department of Employment,Economic Development and Innovation (now Department of Agriculture, Fisheries andForestry (DAFF)), closed Gladstone Harbour and the surrounding area to fishing for a periodof 21 days, under section 96 of the Fisheries Act 1994 in response to concerns about humanhealth and to allow further testing to be carried out on the conditions affecting locally-caughtfish.From the initial testing of nine ill barramundi, two conditions were identified that wereaffecting barramundi in the Gladstone area: Red-spot disease (Epizootic Ulcerative Syndrome (EUS) or ‘red-spot’ disease), which iscaused by a fungus endemic to finfish species of mainland Australia. This condition wasonly confirmed in one barramundi from Port Alma, near the Fitzroy River; An external parasite Neobenedenia sp., which was affecting the eye and skinparticularly on some of the barramundi in Gladstone Harbour.Barramundi and other finfish species displayed a range of generally mild skin conditions fromlocalised skin inflammation, skin erosion and redness. No bacterial, parasitic or fungalpathogens were found that could explain the skin conditions.Reports were also received from the Gladstone Area Water Board that an estimated 30 000 1large barramundi entered the Boyne River in early 2011 when the Awoonga Dam spilled overfor the first time since 1996. Commercial catches of barramundi in 2011 were in excess of20 times the average annual Gladstone Harbour (Grid S30) catch from 2005 to 2010,providing support for this observation.In response to the fish health issues, the Queensland Government set up an investigationprogram that included fish sampling and testing, water quality sampling and testing, andinvestigation into human health concerns.Fisheries Queensland commenced monitoring fish health in Gladstone Harbour after theclosure of the Harbour was declared, and continued to monitor fish health in Gladstone andadjacent areas in conjunction with commercial fishers until September 2012. BiosecurityQueensland conducted examination of animals of particular concern, which were submittedby either Fisheries Queensland or members of the public.In October 2011, an Independent Scientific Advisory Panel was established to provideindependent scientific advice to the Queensland Government on the fish health investigationin Gladstone Harbour. The membership of the panel comprised eminent scientists with1Fisheries Queensland was advised verbally that an estimated 30 000 barramundi were washedover the spillway in 2010/11 and has used this figure in all its publications. However, in theGAWB Annual report 2011, 20 000 barramundi are reported as being washed over the spillway.Given that the commercial fishers caught almost 250 t of barramundi in Gladstone in 2011and theaverage weight of barramundi in October 2011 was 9 kg; it is estimated that the commercialfishers caught more than 27 700 barramundi. In the last quarter of 2011, catch rates were stillmore than two and a half times the pre-flood rate for this quarter, therefore Fisheries Queenslandbelieves that the figure of 30 000 barramundi being washed out of Awoonga Dam is anunderestimate.9

ecognised expertise and research publications concerning aquatic environmental scienceincluding water quality; fish health and toxicology; and human health especially in relation tothe potential for transmission of diseases from marine species to humans.The Panel reviewed the Queensland Government’s monitoring regimes, results collected in2011 and analysis primarily focusing on fish health in Gladstone Harbour and surrounds, butalso considered water quality monitoring and human health issues where relevant andappropriate.The Panel acknowledged that this was a complex issue and supported the Government’songoing investigation of the issue, noting that good progress had been made. The Panelmade specific comments and recommendations in relation to the issues of fish health, waterquality and human health with a view to identifying a possible cause(s) of the fish healthissues being observed in Gladstone Harbour.In particular, the Panel emphasised the need for comparative information from similarunaffected systems to determine the scale of the problem being observed in barramundi andother fish species in the Gladstone area. It was recommended that a more sophisticatedstudy design be conducted, including baselines and trends during ’normal’ periods, andappropriate areas outside the Harbour to act as a form of control for comparative analysis.The Panel also recommended constructing a conceptual model to illustrate possible causeeffectrelationship(s) to help guide studies and eliminate potential causal factors.Stocking in Lake AwoongaBarramundi have been stocked in Lake Awoonga since 1996. According to the GladstoneArea Water Board website, over 4.54 million fish had been released into Lake Awoongaincluding 3.69 million barramundi, 465 718 sea mullet and 70 942 mangrove jack. Since2006, an additional 1.3 million barramundi have been stocked in Lake Awoonga. Until waterflowed over the spillway on 12 December 2010, the stocked fish had not been able tomigrate downstream because the dam had not overflowed for the previous 14 years. In 2002,the height of Awoonga Dam was raised.Calliope and Boyne Rivers flooding eventAbove average rainfall occurred in November and December 2010. Totals of 93 mm and587.4 mm respectively were recorded at the Calliope weather station, compared with themedian level of 69.4 mm and 95.9 mm respectively. The total annual rainfall in 2010 was1702 mm, the second highest level since records started in 1906 up until 2010 (the highestlevel was 1801 mm, recorded in 1956).For comparison, 1973 was a wet year across Queensland, which resulted in the Brisbanefloods in January 1974. The total annual rainfall at the Calliope weather station in 1973 was1326 mm, which was the eighth highest on record. The mean total annual rainfall is 854 mmand the median total annual rainfall is 811 mm.The rainfall in 2010/11 resulted in an annual stream flow in the Boyne River of 1 194 848 ML(measured between October 2010 to September 2011), the highest stream flow sincerecords started in 1984. This is greater than the full supply capacity of Wivenhoe Dam, whichis 1,165,238 ML. The monthly stream flow for December 2010 was 634 999 ML, also thehighest since records started.On 12 December 2010, Awoonga Dam overflowed and did not cease spilling until June 2011.On Tuesday 28 December 2010, the water level in the dam had reached its highest point10

ever at 44 m; 4 m above the spillway level. In 2013, a higher flood peak in the Boyne Riverwas reached at 48.3 m, more than twice the height above the spillway in 2010/11.It was estimated that 30 000 large barramundi were washed out of Awoonga Dam into theBoyne River between December 2010 and March 2011, as the result of water flow over thespillway. If the average weight of these fish was ten kilograms, this represents an additional300 tonnes of barramundi introduced into the Boyne River and Gladstone Harbour.Recreational catchA charter operator in the Gladstone region, Johnny Mitchell, provided a report to FisheriesQueensland of his observations of the recreational catch of barramundi in the Boyne Riversystem in 2011 (Appendix E).His report documented the movement of large schools of large barramundi (Figure 2), witheach school exceeding an estimated one thousand barramundi moving from the freshwatersection of the Upper Boyne through the tidal sections to the mouth of the Boyne River andinto the Harbour. Within a week of the initial overflow at Awoonga Dam, he “witnessed manynew fresh and lively arrivals (escapee lake fish), detected with the use of high definitiondepth sounder in areas within the harbour up to 43 km from the dam wall” and reported highcatches of barramundi in the Harbour up to 50 km north and south of the Boyne River in lateJanuary 2011. Anglers reported catching 12 to 15 barramundi averaging around 100 cm injust one to two hours of fishing. Many of the fish showed signs of physical damage to the jawor head; some wounds were severe, although advanced healing was obvious.Figure 1. A school of large fish in the Boyne River detected on a fish finder during recreational fishing in2011.11

Mr Mitchell reported high catches until late April 2011 in the freshwater, tidal reaches of theBoyne River, Gladstone Harbour and extending into adjoining areas. It is evident that manyof these fish migrated out of the Harbour as tagged Awoonga fish were reported in theBurnett River (Bundaberg) and Ross River (Townsville) (Tagged 9/2/2011 at AwoongaSpillway, recaptured 31/8/2011 at Ross River).By April/May 2011, Mr Mitchell reported that the anal and tail fins were eroded and the fishwere in thinner condition compared with the barramundi caught shortly after being washedover the spillway. In late July 2011, Mr Mitchell observed the first signs of eye problems andthe “skinny fish, tattered fins and the loss of that glossy gleaming iconic barra glint” becamemore common and obvious. “The fish were duller and the body slime seemed more whitishrather than opaque.”Sawynok, et al (2013) also document the change in recreational fishing in Gladstone afterthe 2010/11 flood. Information about fishing trips was obtained from boat ramp surveys anddirectly from recreational fishers and summarised seasonally for the period Autumn 2006 toSpring 2012. The influx of fish from Lake Awoonga changed the catch composition of fishreported by recreational fishers. Barramundi for the period 2006–10 represented only 3.1%of the catch. This changed to 28.4% for the period 2010–12. In the year 2010/11, barramundirepresented more than 50% of the fish caught.12

Phase 1: Fish Health Survey – Gladstone Harbour(August 2011–February 2012)OverviewThe first fish samples were provided to Biosecurity Queensland from August 2011. In lateSeptember 2011, Fisheries Queensland commenced monitoring fish health in GladstoneHarbour. The monitoring was based on a visual examination of fish and crustaceans. Theinitial monitoring focused on understanding the distribution of fish within the Harbour withexternal symptoms and the severity of those symptoms.A number of specific species were monitored in more detail after reports of high levels ofabnormalities were received from recreational and commercial fishers. These included trawlspecies (banana prawns and trawled fish), mud crabs, sharks, scallops and barramundi.Field staff were given training on the collection of biological samples (skin swabs, tissuesamples from various organs, collection of observed parasites), and photographedabnormalities as reference material. After the initial monitoring trips in September andOctober 2011, the definitions of skill discolouration, eye condition and lesion categories wererefined and used in the rest of Phase 1 and in Phase 2 (Appendix A). After the initialmonitoring, Fisheries Queensland established a routine monthly fish health monitoringprogram.Study areaThe principal study area was Gladstone Harbour and adjacent waterways, encompassing theareas closed to fishing during September 2011. Sampling took place at a range of siteswithin the principal study area to allow an understanding of fish presenting with signs of illhealth and their potential movement within the principal study area. The sites sampled werethe Narrows, Hamilton Point, Calliope River, Gladstone Harbour (trawl), spoil ground, UpperBoyne River, Lower Boyne River and Rodds Bay (Figure 1).The main reference sites (i.e. for comparison with the principal study area) for the structuredmonitoring program included the nearby Fitzroy River to the north and waters adjacent toBundaberg with its nearby coastal waters to the south. Both of these sites have various landuse activities conducted adjacent to the waterways. These sites are useful as reference sitesgiven their environments, fish communities and geographic location. It must be noted that notall species of finfish and crustaceans were sampled at each site because of differences intheir distribution and the suitability of sampling methods in different areas. In Phase 1 only,occasional samples were also collected from within the principal study area but outside of thelocations described in Figure 2.13

Figure 2. Sampling sites within the principal study area.Candidate speciesA range of species were sampled, and observations recorded in accordance with proceduresin Appendix A. Most observations were done as part of normal fishing activities. Subject tolimiting the impact on commercial fishing activities, all species caught were examined.Sampling regimeTiming: Sampling commenced in September 2011 and continued until February 2012. Allspecies caught during sampling events were assessed for signs of ill health.Although Fisheries Queensland stopped sampling for Phase 1 in February 2012, the generalpublic continued to provide samples for testing by Biosecurity Queensland up until April2012.Gear type: Multiple gear types were used during the sampling program depending on thesite and fishing activities. The gear types were consistent with commercial fishing apparatusand included: gill net / haul net – finfish and sharks prawn trawl gear – to target banana prawn, demersal finfish crab pots – mud crabs electrofisher – fish species in freshwater.14

Data collection and processingFisheries Queensland observers were placed on commercial fishing vessels during normalcommercial fishing trips and asked to record information on gross signs of ill health observedin the catch. On occasion, commercial fishers were contracted to fish in designated locationswhen no normal commercial fishing activity was taking place at that time.Fish were observed upon capture and assessed visually for signs of ill health. Data collectedincluded an assessment of skin discolouration, eye condition, lesions and presence ofectoparasites.Prawns were assessed visually for signs of ill health including shell erosion and the presenceof parasites. Crabs were assessed for shell abnormalities. Shell lesions were gradedaccording to the methods described by Andersen (2003).When fish or crustaceans showed signs of ill health, samples were collected (either as wholeanimals or tissue samples) and forwarded to Biosecurity Queensland for furtherhistopathology investigations and residue testing.Some samples were submitted to Biosecurity Queensland by members of the public.15

Phase 2: Expanded Gladstone Fish Health Survey(March 2012–September 2012)OverviewBased on the Independent Scientific Advisory Panel’s recommendations, the QueenslandGovernment conducted an expanded integrated program to understand the causes of fishhealth issues in Gladstone Harbour through further monitoring and research.The objectives of the expanded Gladstone fish health sampling program were: To continue monitoring the fish health in Gladstone Harbour and the surroundingwaterways so that Fisheries Queensland had 12 months of data to account for seasonalinfluences. To determine the health status of fish and crustacean species in the Gladstone Harbourand adjacent waterways. For the purposes of this study, health status was defined asthe observed prevalence and severity of significant infectious diseases and pathologicallesions. To determine whether the health status of fish and crustacean species in GladstoneHarbour and adjacent waterways was significantly different to other areas along thecentral Queensland coast. To provide information for the conceptual model being developed by the QueenslandGovernment to help narrow the range of possible causes for the observed health issues,and provide focus for further investigations.The Extended Gladstone Fish Health Survey (Phase 2) continued and improved on the fishhealth investigations conducted by Fisheries Queensland during Phase 1 in the GladstoneHarbour, and adjacent waterways and reference sites. Phase 1 sampling events hadprovided important information for a range of health issues being displayed by fish andcrustacean species, in particular the relationship between parasite burden (Neobenedeniasp.) and skin discolouration in barramundi. This information was used to guide the structureof the extended sampling program (see Fish health survey report 1 March 2012, availablefrom www.qld.gov.au/gladstoneharbour).A more intensive sampling program was required to better understand the variation intemporal (seasonal) and spatial prevalence of symptoms displayed in fish and crustaceanspecies. Phase 1 sampling provided important information on the health status of a numberof fish and crustacean species during late spring and early summer. The extended samplingprogram was completed by 30 September 2012 and, when combined with results fromprevious investigations, provided Fisheries Queensland with observations for a 12-monthperiod from both within and outside Gladstone Harbour.The intensive and structured nature of the sampling program has facilitated more robuststatistical analysis of results than was possible in Phase 1. It has also provided a referencepoint for any future monitoring that considers longer-term temporal variation (e.g. annual) inhealth status of fish and crustacean species in Gladstone Harbour and adjacent waterways.Conceptual modelA conceptual model was developed to document and graphically display the potential factorsthat may influence the health of fish and crabs in the waterways around Gladstone. Thesefactors and pathways were documented and assessed by reviewing scientific literature and16

from advice from a panel of experts. Next, the likelihood that they were affecting (positively ornegatively) the health of fish and crabs was considered. Development of the modelconsidered the impact on fish health of changes to the ecosystem, such as changes to thefood web and competition among predators, as well as the introduction of toxicants thatcould directly affect fish health.The model allows for different hypotheses (e.g. competition for food versus effects ofturbidity) to be visualised and their likelihood to be compared. It can also be used to identifywhich steps in the pathway can be monitored to determine whether or not the particularpathway is actually affecting fish health. For example, to test the ’competition for food’pathway, the abundance of prey fish should be assessed and compared to other systems.The model was never intended to be a quantitative tool with the sophistication to predict thehealth of fish given various levels of driving factors. It can, however, be utilised to explain thethinking and justification for the likely pathways affecting fish health and help improve theunderstanding of the ecological processes that potentially result in a reduction in the healthof fish in Gladstone waterways.The conceptual model indicated that water quality, including turbidity, was unlikely to be theprimary contributing cause for the fish health issues observed in Gladstone Harbour andadjacent waterways.Study areaThe study area during Phase 2 was the same as the study area described for Phase 1, withsampling sites illustrated in Figure 2.Candidate speciesThe monitoring program focused on seven species of finfish, one species of shark, onespecies of prawn, and one species of crab (species listed below). These species represent arange of different life cycles (e.g. catadromous and estuarine) and trophic levels (e.g.predatory, omnivorous detritivores and scavengers). They are species that were encounteredduring Phase 1 sampling in the principal study area and reference areas, and had displayeda variety of abnormalities.Barramundi (Lates calcarifer): Barramundi is a predatory species and a principle target inthe region for recreational line and commercial net fishers. It is catadromous (live in freshand salt water), but must migrate to salt waters to spawn. Barramundi fingerlings are stockedinto freshwater impoundments throughout the Port Curtis and Fitzroy River catchments,including Awoonga Dam.Mullet (Mugil sp.): Mullet is an omnivorous detritivore with a catadromous lifecycle. Thisspecies is caught mainly by commercial net fishers, although smaller numbers are alsocaught by recreational fishers (mainly for bait) using cast nets. Sea mullet have been stockedinto Lake Awoonga.Banana prawn (Fenneropenaeus merguiensis): Banana prawns are omnivorousdetritivores and a principle target for the commercial trawl fishery in the region, as well as forrecreational fishers using cast nets. Banana prawns use the numerous intertidal mangrovelinedcreeks as nursery habitats, then move into waters including Gladstone Harbour as theygrow.17

Mud crab (Scylla serrata): Mud crab is the principle target species for recreational andcommercial crabbers in the region. The species is an active omnivorous scavenger thatoccurs in estuarine and coastal habitats with mud substratum.Bull shark (Carcharhinus leucas): Bull shark is a predatory estuarine and coastal speciescaught frequently in the region by commercial net fishers and recreational line fishers. Theyare known to inhabit freshwaters, particularly as juveniles.Trawl fish species – Grinner (Saurida sp), Australian threadfin (Polydactylusmultiradiatus) and Castelnau’s herring (Herklotsichthys castelnaui): These three taxaare small in size and common in the local demersal fish assemblages, which makes themcommon in bycatch of trawlers operating inside the study area. They are caught occasionallyby recreational anglers.Pelagic species - Queenfish (Scomberoides sp.): Queenfish is a pelagic fish occurringthroughout the region and commonly caught in commercial nets.Sampling regimeTiming: The expanded sampling regime commenced in April 2012 and was repeated inJune/July 2012.A third sampling event was scheduled for September 2012, however observations madebetween April and July 2012 indicated that fish health was improving compared toobservations made in 2011. Therefore, sampling in September 2012 was refocused oninvestigating whether Neobenedenia sp. identified as affecting barramundi in the BoyneRiver in October 2011 reoccurred in the Boyne River at a similar time in 2012.Gear type: Multiple gear types were used during the sampling program depending on siteand candidate species. The gear types are consistent with commercial fishing apparatus andinclude: gill net / haul net – barramundi, sea mullet, bull sharks and queenfish prawn trawl gear – banana prawn, grinner, Australian threadfin, Castelnau’s herring crab pots – mud crab electrofisher – fish species in freshwater.Data collection and processingFisheries Queensland contracted commercial fishers to sample at designated samplinglocations using appropriate fishing apparatus to catch candidate species (as describedabove). Exceptions to this included barramundi collected from Lake Awoonga using anelectrofisher, and some mud crab/finfish sampling that involved Fisheries Queensland staffaccompanying commercial fishers during normal commercial fishing activities.Observations in accordance with procedures in Appendix A were made of all fish andcrustaceans at the time of capture by Fisheries Queensland staff. A sub-sample of the catchwas forwarded to the lab where observations were made by Biosecurity Queensland staff.The samples underwent necropsy examination, and tissue samples were collected for furtherhistopathology investigations and chemical residue testing. The sub-sample selected forfurther testing included individuals with and without obvious abnormalities. Individuals without18

obvious abnormalities were included to ensure that health conditions not apparent byexternal observation alone could be detected during further laboratory investigations.Statistical analysis was performed to determine significant relationships and trends in thefield and necropsy data, histopathology data and chemical residue data (Appendices B, Cand D respectively).Field observations by Fisheries Queensland: Fish were observed upon capture andassessed visually for external signs of ill health. Data collected included an assessment ofskin discolouration, eye condition, lesions and presence of ecto-parasites. For the purposeof this data collection, skin lesions were distinguished from skin discoloration by observablesigns other than colour change, such as missing scales, ulceration or bleeding (seeAppendix A).Prawns were assessed visually for signs of ill health including shell erosion/lesions and thepresence of parasites. Crabs were assessed for shell abnormalities. Shell lesions weregraded according to the methods described by Andersen (2001).Fish collected using trawl methods (prawns, herring, grinner, Australian threadfin) werenecropsied while onboard the vessel by a veterinarian, and tissue samples were collected forfurther histopathology investigations and residue testing.Laboratory observations by Biosecurity Queensland: Fish were assessed visually forsigns of ill health once received in the lab by a veterinarian. Data collected included, but wasnot limited to, an assessment of skin discolouration, eye condition, lesions and presence ofecto-parasites. Observations of internal organs were made during necropsy, and tissuesamples were collected for further histopathology investigations and residue testing.Crabs were assessed for external shell abnormalities once received in the lab by an aquaticveterinary officer. Shell lesions were graded and location on the carapace recordedaccording to the methods described by Andersen (2003). Further observations were made ofinternal organs during necropsy and, in order to allow a statistical comparison of the fish fromdifferent sites, discrete categories were assigned to capture these observations based onAdams et al. (1993). Chemical residue test results were interpreted by the Department ofScience, Innovation Technology, Innovation and the Arts. For a detailed description of thesampling methods, refer to Appendix A.Histopathology and chemical residue testingA subset of individuals were selected for histopathology examination and chemical residuetesting (Table 1).For barramundi, gills, liver and skin (with muscle) were chosen for microscopic investigation,due to their known sensitivity as indicators of environmental stress. Specifically, thefollowing abnormalities were rated from zero to three to enable quantitative comparisonbetween sites and time points: gill hyperplasia; gill parasites; liver melanomacrophagecentres; skin lesions; and muscle lesions.Barramundi gills and liver were used for chemical residue testing, as they were deemed themost likely tissues to show evidence of short to medium term exposure to contamination.Since there were no significant findings in organic testing of barramundi tissues during Phase1, only metals were tested in Phase 2.19

For mud crabs, hepatopancreas, gills and muscle were examined histopathologically.Hepatopancreas tissue was also used for chemical residue testing, which included testing formetals and organic contaminants.Grinner were chosen due to their benthic and relatively immobile nature, size (ease /feasibility of processing) and previous evidence provided by members of the public of red /sick grinner from Gladstone Harbour. Since no grinner were caught from Bundaberg duringApril, a total of 25 fish were tested. Gill, liver and muscle were used for histopathologyexamination. For chemical residue testing, only muscle tissue could be sampled in sufficientvolume due to the small size of these fish.Since insufficient numbers of sharks were caught for a balanced statistical analysis, theywere used for histopathology only.Chemical residue test results were interpreted by the Department of Science, InformationTechnology, Innovation and the Arts.20

Table 1. Number of fish from each site and sampling trip used for histopathology and chemical residuetesting. Note: 1 Denotes Gladstone Harbour and adjacent waterways, 2 denotes reference sites. Note dueto low sample numbers, sharks were used for histology only.Site Species collected per trip 2012LakeAwoongaShark Mud crab Barramundi GrinnerApril/MayApril/MayJune/JulyApril/May5June/JulyApril/May5 51 5 55 5Upper Boyne5 5River 1Port Dev’tArea 1HamiltonPoint 1The Narrows 1 5 5GladstoneTrawl 1Fitzroy River 2 5 5 5 5 5 5 5Bundaberg 2 1 5 5 5 5 5CalliopeRiver 1 3June/JulySummary of samples collectedDuring the course of fish health investigations conducted by Fisheries Queensland inGladstone Harbour and surrounding waterways, approximately 9000 fish, crustaceans andmolluscs were sampled. Observations were made across a wide range of species and aresummarised in Table 2. Samples provided to Biosecurity Queensland for further testing havebeen summarised in Tables 3 and 4. Biosecurity numbers include some samples provided bysourcesother than Fisheries Queensland during Phase 1 (Table 3).21

Table 2. The number of fish, crustaceans and molluscs observed during Phase 1 and 2 of the GladstoneFish Health Investigation. Note * indicates multiple species included in categoryFish species(taxonomic group eg. Family or Genus or species)Phase 1 Phase 2 TotalAnchovies * (Engraulidae) 355 355Australian threadfin * (Polydactylus spp.) 361 545 906Barramundi (Lates calcarifer) 281 216 497Batfish * (Ephippidae, Drepaneidae) 3 1 4Beach salmon (Leptobrama muelleri) 2 2Black jew (Protonibea diacanthus) 1 1Black pomfret (Parastromateus niger) 3 3Blue threadfin (Eleutheronema tetradactylum) 51 1 52Bony bream (Nematalosa erebi) 42 42Bream * (Acanthopagrus spp.) 46 46Butter bream (Monodactylus argenteus) 2 2Catfish (Ariidae) 129 129Cod/Groupers * (Epinephelus spp.) 17 1 18Flathead * (Platycephalus spp.) 9 1 10Mackerels and Bonitos* (Scombridae) 7 7Grinner * (Bathysauridae, Synodontidae) 80 179 259Herring * (Clupeidae, Pristigasteridae, Elopidae) 72 346 418Javelin fish * (Pomadasys spp.) 35 35King threadfin (Polydactylus macrochir) 21 21Milk fish (Chanos chanos) 1 1Mullet * (Mugilidae) 149 125 274Ponyfish * (Leiognathidae) 366 366Queenfish * (Scomberoides spp.) 89 71 160River jew *(Johnius spp.) 419 419Scad * (Carangidae) 18 18Scats * (Scatophagidae) 10 10Sharks and rays * (multiple Families) 227 26 253Silverbiddies (Gerreidae) 24 24Snappers (Lutjanidae) * 2 2Snubnose dart (Trachinotus blochii) 42 42Sole * (Soleidae, Cynoglossidae) 1 1Sweetlips and emperors* (Haemulidae, Lethrinidae) 3 2 5Trevally * (Carangidae) 49 5 54Tripletail (Lobotidae) 3 3Whiting * (Sillago spp) 303 303Total 3211 1526 4737Crustaceans and molluscs Phase 1 Phase 2 TotalMoreton Bay bug * (Thenus spp.) 1 1Scallops (Pectinidae) 23 23Banana prawns (Fenneropenaeus merguiensis)266 (+ an additional 574 84085kg not quantified)Other prawns * (Penaeidae) 294 294Crabs - Not mud crabs * (Portunidae) 5 5Mud crabs (Sylla serrata) 1435 1599 3034Total2024 + (+ = 85kgbanana prawns)2173 4197+22

Table 3. Numbers of samples tested by Biosecurity Queensland during Phase 1.Fish Species Necropsy Histopathology Microbiology ResidueHeldfrozenBarramundi (Lates calcarifer) 33 69 30 6 27Black jew (Protonibea diacanthus) 0 1 1 0 0Bony bream (Nematalosa erebi) 1 1 0 0 1Catfish (Ariidae) 9 6 6 6 0Flathead (Platycephalus sp.) 0 0 0 0 2Grinner (Bathysauridae,Synodontidae)1 1 0 0 1Mullet (Mugilidae) 2 2 3 4 2Queenfish (Scomberoides spp.) 2 2 2 0 2River jew (Johnius spp.) 7 7 3 7 0Scats (Scatophagidae) 2 2 0 0 1Spangled emperor (Lethrinusnebulosus)1 1 1 0 1Cod/Groupers (Epinephelus sp.) 1 1 1 0 0Threadfin (Polynemidae) 2 2 0 0 2Trevally (Carangidae) 2 2 0 0 1Whiting (Sillago spp) 1 6 1 4 1Sharks (multiple families) 5 21 21 0 2Total 69 124 69 27 43Crustaceans and molluscs Necropsy Histopathology Microbiology ResidueHeldfrozenCrabs (Portunidae) 9 8 8 0 2Moreton Bay bug (Thenus sp.) 1 0 0 0 3Prawns (Penaeidae) 2 11 2 7 1Scallops (Pectinidae) 0 7 0 63 0Total 12 26 10 70 6Table 4. Numbers of samples used for necropsy examination during Phase 2.BullOtherLocation BarramundiMullet QueenfishMud crab TotalsharkpelagicBundaberg 17 1 20 20 21 79Fitzroy River 20 7 21 11 2 21 82Calliope River 20 3 17 18 58Hamilton Point 20 1 20 41Lower Boyne 20 20 1 41Rodds Bay 9 20 20 49Spoil Ground 5 15 20Upper Boyne 16 20 36Narrows 20 20Port Dev’t Area 21 21Lake Awoonga 5 5Total 127 12 118 37 17 141 45223

Significant findings and discussionAnalysis of all species combinedThe structured sampling in Phase 2 allowed a statistical analysis to be done on fieldobservations and necropsy data from all species combined (Appendix B). Overall, locationhad a significant effect on most variables, and for many variables there was a significantinteraction (P < 0.05) between location and trip, indicating a changing pattern through spaceand time.Health assessment index (HAI) was calculated based on the methods of Adams et al. (1993),as a coarse measure of overall health, based on combined scores from a range of internaland external observations made during necropsy. Higher scores indicated a higherprevalence and/or severity of abnormalities. In Trip 1, average HAI was highest in Bundaberg(32.8), followed by the Lower (32.7) and Upper Boyne (30.9) respectively. In Trip 2, theUpper Boyne had by far the highest HAI score (38.8) followed by Lower Boyne (28.2) andBundaberg (24.6). There was no significant difference between pooled Gladstone sites 2 andpooled reference sites for HAI in either Trip 1 or 2.In Trip 1, external observations and measurements showed that Gladstone fish hadsignificantly lower condition factor (weight×10 5 /Length 3 ) and higher proportion of tuckedabdomen (poor condition) than fish from the reference sites (P ≤ 0.05). Internal observationssupport this finding with a significantly lower proportion of mesentery fat in fish fromGladstone. In particular, fish from Hamilton Point had the lowest condition factor and lowestlevels of mesentery fat in both trips, while Bundaberg fish had the highest condition factors inboth trips.The estimated annual growth rate (2011) of barramundi in Gladstone (Boyne River44.0±50.5mm and Calliope River 49.1±40.0m) compared with the Fitzroy River(169.8±67.6mm) (Sawynok, et al 2013) for barramundi greater than 650 mm could indicatelimited food supply that results in poor condition as found in Trip 1.In Trip 2, the proportion of fish with parasites identified during necropsy was significantlygreater in fish from Gladstone (P < 0.05), with the highest level observed in barramundi fromHamilton Point.Prevalence of lesions and fin abnormalities was significantly higher in Gladstone than in thereference sites for both trips (P < 0.05). The lesions in particular can be largely explained bythe high numbers of barramundi with lesions observed from the Boyne River during both tripsduring Phase 2.Externally visible health ‘status’ was assessed in the field for every individual fish. Necropsydata were tested for differences between apparently normal and abnormal fish. Liverabnormalities were significantly higher in apparently abnormal fish (P < 0.05). However,despite significant differences in lesions and fin abnormalities between Gladstone andreference sites, the only significant difference in livers between Gladstone and referencesites was in Trip 2 when liver colour abnormalities in fish from Gladstone were significantlyless prevalent than at the reference sites.2 Gladstone sites refers to Gladstone Harbour and adjacent waterways including the Boyne andCalliope Rivers and the Narrows.24

There were significant differences between the two reference sites for most conditions in Trip1, Trip 2 or both (P < 0.05). Some differences would be expected due to the difference inlatitude and habitat structure. Reference sites were chosen to the north and south ofGladstone to balance these effects.Skin condition: Skin redness has been reported and investigated in a wide range of fishspecies including sharks. Pathology investigation of skin redness in Phase 1 showedinflammation, erosion, scarring, bleeding, cell death and swelling, but no causative agentwas identified.Skin redness was further investigated in Phase 2. The broad range of time taken to processsamples allowed changes in observed skin redness over time to be documented. Fish werephotographed in the field as they were caught, and again in the laboratory as part of thenecropsy examination. Fish were stored in plastic bags on ice, and took between zero and17 hours to process. Data showed that skin condition and fin condition ratings were affectedby time since death (P < 0.05), and that there was a significant interaction with fish species.This means that changes in skin condition through time are different according to species. Inboth barramundi and mullet, the proportion of fish with a skin condition other than normalincreased over time (P = 0.010 and 0.056, respectively). There was no significant differencebetween the relationships observed in Gladstone and reference sites. This indicates that thereddening of fish over time is not a phenomenon specific to Gladstone Harbour.Not enough bull sharks were caught for the above statistical analysis, although the limitedevidence did suggest a similar pattern. Photographic evidence allowed for visiblecomparison of the fish over time, and showed that skin redness developed in some sharksand other fish post mortem (Figure 38).BarramundiCommercial catchBetween 2000 and 2010, the reported commercial catch of barramundi in the entireGladstone region (S30) ranged between 3.97 t (2009) and 16.77 t (2005) with an annualaverage of 10.8 t. In 2011, the influx of barramundi from Lake Awoonga into the system sawthe commercial catch rise to an unprecedented high of 248.31 t (Figure 3). Of this total, 130 twas captured between July and September 2011. Logbook data supplied by commercialfishers for 2012 shows that the recorded catch returned to historical levels.25

Commercial barramundi catch and catch per day for the Gladstone region2000 ‐ 2011300Total Catch (t)Catch (kg) per day600Weight (tonnes)25020015010050500400300200100Catch per day (kg/day)02000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012Year0Figure 3. The annual catch and catch per unit effort of barramundi from the Gladstone region (CFish gridS30), as reported in commercial fisher logbooks (calendar years 2000 – 2012).Commercial fishing effort for barramundi in the Gladstone region (days fished withbarramundi catch recorded) increased between 2000 and 2004, peaking at 370 days. After2004, effort dropped steadily to a low of 108 days in 2009 with a rise in 2010 to 223 days.Following the introduction of fish from Lake Awoonga into the Boyne River in 2011, effortincreased to a record high of 457 days (Figure 4). Although catches of barramundi for 2012returned to historical levels, the fishing effort expended (141 days) was the second lowestnumber of days resulting in the second highest recorded catch rate (135kg/day).Days500450400350300250200150100500Commercial effort (days fished) for barramundi in the Gladstone region2000 ‐ 20112000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 4. The number of days fishing where commercial fishers have reported barramundi from theGladstone region (CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 –2012).26

Field observational Phase 1During Phase 1 of the fish health investigations in Gladstone, 267 barramundi were observedand their conditions recorded. This figure included fish observed at the reference sites, butexcluded those sampled during a targeted Neobenedenia survey in the Boyne River, whichwill be discussed separately.While the deep ulcerative lesions seen on barramundi prior to the commencement ofsampling in September 2011 were not detected, a number of barramundi were observed tohave lesions comprising areas of detached scales, varying skin discoloration as well as eyeabnormalities. The results of observations are summarised in Figures 5 and 6.Skin conditions observed in barramundi at each site during Phase 1No Abnormality Minor Redness General Redness Extensive Redness Areas of detached scales706050Number of Fish403020100Fitzroy River(n = 26)Narrows(n = 2)Hamilton Point(n = 48)Calliope River(n = 22)Boyne River(n = 92)Rodds Bay(n = 76)Bundaberg(n = 1)LocationFigure 5. Skin condition observed in barramundi at each site sampled during Phase 1 of fish healthinvestigations, September 2011 – February 2012. Note, (n) refers to the total number of fish examined ateach site.Eye conditions observed in barramundi at each site during Phase 1No Abnormality Cloudy Red‐Haemorrhaging Ruptured807060Number of Fish50403020100Fitzroy River(n = 26)Narrows(n = 2)Hamilton Point(n = 48)Calliope River(n = 22)Boyne River(n = 92)Rodds Bay(n = 76)Bundaberg(n = 1)Location27

Figure 6. Eye condition observed in barramundi at each site sampled during Phase 1 of fish healthinvestigations, September 2011 – February 2012. Note, (n) refers to the total number of fish examined ateach site.Laboratory findings Phase 1The majority of the barramundi submitted for necropsy, including fish submitted prior to 16September 2011, showed eye lesions ranging in severity from inflammation to ruptured eyes.Severe skin lesions and abnormalities including generalised redness were also observed.Copepods (small crustaceans) were found on the gills of some barramundi in lowabundance. Copepods are not considered to be a serious pathogen at the observed level ofabundance.A number of barramundi showed a high number and intensity of melanomacrophage centres(MMCs) in organs (e.g. liver, spleen and kidney). An increase in MMCs is a commonhistopathological indicator of a range of factors or stressors including poor water quality,starvation, disease or ageing. The general function of MMCs is the focalisation, destruction,detoxification and recycling of internal and external materials. An increase in MMCs isgenerally likely to be the result of these catabolic physiological processes in the fish,commonly a negative nutritional balance (Agius & Roberts, 1981)There was no detectable sediment in the gills of the barramundi samples except for twosubmissions—see P11-76883 (Boyne River) and P11-76885 (Hamilton Point) in reports onwww.qld.gov.au/gladstoneharbourParasitic flatworm infection – Pathological examination of barramundi by BiosecurityQueensland identified a parasitic flatworm (Neobenedenia sp.) as the most likely causativeagent of the eye abnormality in the barramundi. Bacteria were excluded as the cause of thelesions. Flatworms can cause eye injuries by way of feeding and attachment over the surfaceof the cornea. In eyes that are cloudy, eroded, inflamed, swollen or where there is evidenceof haemorrhage, vision is likely to be reduced. In severe cases where the eye has beenruptured, this will cause blindness.The parasite moves over the surface area of the eye and the skin of the entire fish. A numberof barramundi showed hyperaemic (reddened) areas on the skin surface, which is consistentwith skin damage caused by the parasite feeding on mucous and skin cells on the skinsurface. This has been the first report of a significant outbreak of Neobenedenia sp. in wildbarramundi.Skin lesions and abnormalities – The causes of the deep, ulcerative lesions in barramundicollected in the early stages of the investigation from Gladstone Harbour were not able to bedetermined. Specific histopathology testing showed that Epizootic Ulcerative Syndrome(EUS), which is caused by a fungus, was not present in the samples collected fromGladstone Harbour, therefore EUS has been ruled out as the cause of these lesions. Onlytwo cases of EUS were found in testing: a barramundi taken in September 2011, which wascollected from Port Alma (at the mouth of the Fitzroy River), and in a whiting from theBundaberg region.Chemical residue testing – Barramundi samples were taken from five locations ─ PortAlma, Lake Awoonga and Fitzroy River (reference sites), and China Bay and Boyne River(Gladstone Harbour sites). Testing was conducted on the gill, liver and muscles of each ofthe fish or pooled fish samples.28

Samples were tested for more than 160 organic chemicals. Only two organic chemicals weredetected in the fish samples analysed. These were p’p’-DDE and DEET. p’p’-DDE is adegradation product of the organochlorine pesticide DDT, which was a commonly usedpesticide in the agricultural sector, but has not been permitted in Australia for more than 20years. The samples were from barramundi from China Bay and Turkey Beach in Gladstone.DEET is a very commonly used ingredient in personal insect repellents. Given that it wasfound in only one sample, it is possible that this fish sample may have been contaminated.Concentration of metals, aluminium, barium, copper and zinc were slightly higher in the fishtissue from the reference sites than Gladstone Harbour sites. Cadmium, nickel, selenium andsilver were very similar among sites. Iron and arsenic were present at higher levels in theGladstone barramundi population. Further details can be found atwww.qld.gov.au/gladstoneharbourObservational findings Phase 2April 2012: A total of 111 barramundi were collected from all planned netting sites. NoNeobenedenia sp. or eye abnormalities were observed on fish collected from any of thesampling locations. In the Boyne River, a high prevalence (50-75%) of physical damage wasobserved on the barramundi caught. Conditions described as physical damage appeared asgraze–type injuries and included large areas of scale loss often on one side of the body,operculum damage and some mandible fractures. Physical damage (Figure 7) was attributedto fish passing over the Awoonga spillway earlier this year (Awoonga overtopped in lateJanuary to early February and late March 2012). The observed damage is consistent withinjuries seen in fish after being washed over spillways of other impoundments, andconsistent with observations made by recreational anglers following the 2011 overtopping ofAwoonga Dam.Seven barramundi caught at the reference sites and six barramundi caught in the Gladstonearea (Hamilton Point and Rodds Bay) showed varying levels of redness on the skin. Thisranged from pinpoint marks to more general redness over areas of the body and fins. Noother signs of ill health were observed, with most fish considered to be in good condition.Observations are summarised in Figure 8.Seven barramundi were caught incidentally during trawl sampling in Gladstone Harbour(Figure 9). The commercial fisher involved in the sampling identified this as an unusualcapture, which may indicate a larger number of barramundi than normal being present inGladstone Harbour. No significant signs of ill health were observed on these barramundi.Figure 7. Barramundi caught in the Boyne River showing physical damage consistent with beingwashed over the Awoonga spillway.29

Conditions observed in barramundi sampled at each site during April, Phase 2Skin Colour Normal Skin Minor Redness Skin General rednessSkin Extensive Redness Minor Lesion Physical Damage3530Number of Fish2520151050Fitzroy(n = 33)Hamilton Point(n = 11)Calliope(n = 10)lower Boyne River(n = 22)upper Boyne River(n = 15)Rodds Bay(n = 7)Bundaberg(n = 13)LocationFigure 8. Conditions observed (skin colour and lesions) in barramundi sampled at each site during April2012, Phase 2. Note, (n) refers to the total number of fish examined at each site. Some fish may have beenobserved with multiple conditions.Figure 9. Barramundi caught while trawling in the Gladstone Harbour.June/July 2012: A total of 93 barramundi were collected from all planned netting sites. NoNeobenedenia or eye conditions were observed on barramundi at any of the samplinglocations.Most notable in the barramundi samples collected was the presence of two fish from theupper Boyne River showing recovering lesions. The lesions were well infiltrated withconnective tissue, indicating recovery from injuries that have occurred at least one monthago, as described by the attending veterinarian. The recovering injuries observed in thesefish were consistent with the fresh injuries observed on fish during the April/May samplingevent in the Boyne River, which were attributed to fish passing over the Awoonga Damspillway in late January to early April 2012 (see Figure 10).Barramundi were also caught in the Calliope and Burnett Rivers with injuries in advancedstages of healing. New skin had completely covered the old injuries and new scales wereforming over the affected areas (see Figures 11 and 12). Given the advanced stage of thehealing, it is believed these injuries did not happen during 2012.30

Most barramundi sampled in Gladstone in June/July were considered to be in goodcondition. The majority of skin discolouration or lesion conditions observed during June/Julywere from barramundi caught in the Fitzroy River and were represented by red pin pointmarks, individual scales detached or those attributed to physical injuries (Figure 13).Figure 10. A barramundi caught in the Boyne River during the June/July 2012 sampling event withrecovering injuries to the jaw and body.Figure 11. A barramundi caught in the Burnett River, Bundaberg, during the June/July 2012 samplingevent with an old injury in an advanced stage of healing. Note the new skin completely covering the oldinjury with new scales forming over the affected area.31

Figure 12. A barramundi caught in the Calliope River during the June/July 2012 sampling event with anold injury in an advanced stage of healing. Note the new skin completely covering the old injury.Conditions observed in barramundi sampled at each site during June/July, Phase 2Skin Colour Normal Skin Pale Pink Areas Ventrally Skin Minor Redness Skin General rednessSkin Extensive Redness Minor Lesion Areas of detached scales Physical Damage3025Number of Fish20151050Fitzroy River(n = 35)Hamilton Point(n = 20)Calliope River(n = 13)lower Boyne River(n = 10)upper Boyne River(n = 6)Rodds Bay(n = 2)Bundaberg(n = 7)LocationFigure 13. Conditions observed (skin colour and lesions) in barramundi sampled at each site duringJune/July, Phase 2. Note, (n) refers to the total number of fish examined at each site. Some fish may havebeen observed with multiple conditions.Targeted barramundi surveys to examine Neobenedenia sp. infectionTargeted barramundi surveys in the Boyne River were undertaken by Fisheries Queenslandin October 2011 and September 2012 to investigate parasitism by the monogeneanNeobenedenia sp. Neobenedenia had been implicated as a causative agent for some of theobserved conditions affecting barramundi during Phase 1 of the fish health investigation.Neobenedenia is known to exist naturally in the marine environment in the Gladstone areagiven previous occurrences of Neobenedenia on barramundi broodstock in the GladstoneArea Water Board fish hatchery (pers. comm. Kirt Hutchby, Hatchery Manager).Neobenedenia does not survive in freshwater.Neobenedenia are known to cause disease and mortalities in aquaculture. Deveney et al.(2001) documented an outbreak of Neobenedenia that occurred in the Hinchinbrook Channelon barramundi cultured in sea cages where parasite burdens exceeded 400 specimens perfish, however these parasites have not previously been associated with documented fishhealth issues in wild populations.32

In October 2011, approximately 28 barramundi were caught during sampling. Of these, 14fish were retained for closer study. The remaining fish were tagged and released alive.These fish were tagged to determine upon recapture if they moved out of the Boyne Riverand if their health status changed.Retained fish were selected based on their external appearance. The objective was to collectfish that exhibited the widest range of health conditions, including fish that appeared healthy,to determine if a relationship existed between parasite burden and signs of ill health.Of the 14 retained fish, one had distinct skin lesions, 11 fish had cloudy, red, swollen orruptured eyes, eight fish had general skin discolouration, while a further five showed minorskin discolouration. Samples of the gills and skin were taken from each of the 14 fish. Watersamples were also collected at the sampling site.Neobenedenia counts conducted on each of the retained fish ranged from none to 187parasites per fish. The eye damage observed in many of these fish was consistent withdocumented Neobenedenia infestation. Although bacteria have been excluded as the causeof eye conditions observed, the severity of some eye conditions may have been furtheraffected by opportunistic secondary bacterial infections.There was no clear correlation between observed severity of skin conditions and parasitecount, other than in the extreme cases where the healthiest fish caught had noNeobenedenia while the fish with 187 Neobenedenia had the most severe skin conditions.During sampling in September 2012, only three barramundi were caught. All sampling wasundertaken at the same locations and with the same sampling effort (two nights fishing with acommercial fisher) as in October 2011. All fish caught were observed to be in good healthwith no eye or skin conditions observed. One fish had an old wound on the caudal peduncle(tail) that was in an advanced stage of healing with new skin and scales present. All fish wereretained for closer examination, however no Neobenedenia were found on these fish.The Neobenedenia outbreak described by Deveney et al. (2001) in the HinchinbrookChannel was thought to be caused by environmental stress (uncharacteristically low watertemperatures of 19 o C for three weeks) compromising the ability of barramundi to combatinfestation. This stress, combined with the high stocking densities used in aquacultureenvironments, allowed the number of Neobenedenia to reach epidemic proportions quickly.The density of barramundi in Gladstone was extremely high, as shown by catch rates. Theaverage commercial catch of barramundi in Gladstone in 2011 was 543 kg / day and wasrecorded as 900 kg / day in August 2011, whereas the average catch rate for the period2000-2010 was 50 kg / day for Gladstone (Figure 4). Based on these catch rates, the densityof barramundi was 10 times higher than normal levels in 2011 and 18 times in August 2011.There were a number of factors that would have been stressful to the barramundi inGladstone, rendering them more susceptible to infestation. Although some of the stressorswould have been different to the husbandry practices and seasonal conditions identified byDeveney et al. (2001), many are the same for barramundi in the Boyne River in the winter of2011. The densities of Gladstone barramundi were more than 10 times and possibly up to 18times higher than that of natural barramundi populations. Many of the fish had physicaldamage caused by being washed over the Awoonga Dam spillway. Food supply would havebeen limited due to the large number of large predators present for a lengthy period. TheBoyne River recorded temperatures below 19 o C for July and August 2011 from the mouth to8.6 km upstream. The lowest temperature measured was 18.1 o C at 4 m depth at the site 8.6km upstream from the mouth of the Boyne River. Finally, these cold water temperatures33

provided another environmental stress. Under these conditions, Neobenedenia populationswere able to multiply, resulting in the symptoms observed.The outbreak of Neobenedenia infection in Hinchinbrook cage culture resulted in the loss ofan estimated 200 000 fish. All sick, dying and dead fish were removed from the cages anddisposed of on land. Wild sick barramundi in Gladstone were not removed as would occur inan aquaculture operation, but remained in Gladstone waters. Although low numbers ofbarramundi were reported to have died in the second half of 2011, it is believed that most ofthe infected fish survived.Given the severity and high frequency of health conditions observed in barramundi inOctober 2011, the fish in Gladstone Harbour and adjacent waterways would have showedsigns of poor health for an extended period of time. This is what was observed for theremaining samples taken in 2011 by Fisheries Queensland. In 2012, Fisheries Queenslandobserved fish recovering from the loss of significant areas of scales (Figures 11 and 12).As the water temperature increased, the Neobenedenia infection resolved in the populationof barramundi in the Gladstone area. Barramundi from Gladstone Harbour and adjacentwaterways sampled by Fisheries Queensland after October 2011 and reported by Landos(2012) were not identified as being infected with Neobenedenia.Although more barramundi entered the Boyne after the Awoonga Dam spillway overtoppedagain in late January to early April 2012, it is thought that the number of barramundi that mayhave entered the system was less than during the flooding event of 2010-2011.Neobenedenia are not expected to cause further issues in wild barramundi in the area undernormal environmental conditions.Laboratory findings Phase 2NecropsyNecropsy examination of barramundi in both trips one and two found a significant associationbetween field allocated disease status and liver abnormalities (P < 0.05) (Appendix B).Those fish that were allocated as diseased in the field (based on externally visible signs)were almost four times more likely to have liver form abnormalities than externally normalfish. It is not clear from this data whether the liver abnormalities are a consequence of theskin trauma or whether they are linked to another common factor, such as water quality ordiet.Field-based disease assessment was higher in fish from Gladstone than the reference sites.This was driven by the greater proportion of lesions in barramundi from the Boyne River inparticular. Interestingly, observed liver abnormalities in fish from the Boyne River were onlysignificantly higher than reference sites during Trip 2, despite significantly higher prevalenceof lesions in both trips. In Trip 2 only, skin condition, fin abnormalities and parasites weresignificantly higher in Gladstone barramundi than reference sites (P =

Barramundi found in freshwater impoundments generally have high fat levels. WithinGladstone Harbour, high levels of mesentery fat appeared to be an indicator of fish that hadrecently passed over the Awoonga Dam spillway. All five fish from Lake Awoonga hadgreater than 50% mesentery fat. Of the fish examined from the Boyne River, those with highmesentery fat tended to also have lesions. In total, 19 of the 20 fish with lesions in the BoyneRiver had greater than 50% mesentery fat. Of the remaining 16 fish from the Boyne Riverwith no observed lesions, only one had greater than 50% mesentery fat. Elsewhere inGladstone Harbour, 43 of the 49 barramundi examined had no observable mesentery fat(see Figure 14). The proportion of fish with notable mesentery fat decreased across all sitesbetween April and June/July 2012. In both reference sites, barramundi had relatively highlevels of mesentery fat, which did not appear to be related to lesion scores.Mesentary fat levelsNone 50% caecum covered 100% caecum coveredNumber of Fish20181614121086420L0 L0 L1 L0 L2 L0 L2 L0 L1 L2 L3 L4 L0 L2 L3 L4LakeAwoongaCalliopeRiverHamilton Point Rodds Bay lower Boyne River upper Boyne RiverSite and Lesion ScoreFigure 14. Levels of mesentery fat in barramundi from Gladstone sites, according to their lesion scores asrecorded during necropsy.HistopathologyDue to technical difficulties, there were some samples of barramundi that were not able to beproperly evaluated for the analysis of skin, so the skin histology data could not be included inthe statistical analysis. All other values were combined to allow comparison of cumulativescore as a very crude measure of overall health for each fish. Liver lipid scores were alsocategorised in this way, but were not tallied with other scores because liver storagevacuolation is not an abnormality.Due to the variation in time taken to transport barramundi specimens to the lab and toperform necropsies, there were varying levels of post-mortem degeneration in the tissues.Pathologists were provided with time between capture and necropsy for each specimen toenable estimation of this effect, and quantified only the abnormalities deemed not attributableto post-mortem degeneration.The barramundi from Lake Awoonga had the lowest score for gill hyperplasia, gill parasites,muscle lesions and cumulative score. All five fish from Lake Awoonga were rated moderatefor liver lipid, which is consistent with necropsy findings of relatively high condition factor,mesentery fat and hepato-somatic index in these fish. The differences between fresh andsalt water fish are expected, given the substantial difference between habitats.Gill hyperplasia was significantly higher in barramundi from the reference sites compared toGladstone in April 2012 sampling. There were no significant differences between sites in theJune/July sampling trip. Gill hyperplasia is an abnormal increase in the cell numbers and canbe a sensitive indicator of water-borne irritants. The sensitivity of this analysis in barramundi35

was somewhat reduced because post-mortem degeneration can mask subtle changes in gilltissue. This problem was much reduced in grinner and mud crab, because tissues were fixedimmediately following euthanasia of the specimens.Melanomacrophage centres (MMCs) are distinct clusters of cells containing pigment that areknown to increase in size or frequency during conditions of environmental stress or inresponse to starvation or disease (Agius and Roberts, 2003). Age is also known to have asimilar effect on MMCs, however this has been accounted for in our analysis using length asa proxy for age and adjusting the data as though all fish are the same length.Liver MMCs were rated significantly higher (P < 0.05) in barramundi from Gladstone than thereference sites during June/July 2012. They were similarly higher in Gladstone fish thanthose from reference sites caught during April, but the difference was not statisticallysignificant. The fish from Hamilton Point were rated highest for MMCs in both trip one andtwo. There was a significant relationship between MMC scores and mesentery fat (r = -0.57;P < 0.01). Barramundi with lower levels of mesentery fat tended to have higher MMC scores.Barramundi caught from Bundaberg had the lowest ratings for liver MMCs for both samplingtrips.The barramundi from Hamilton Point also had low levels of mesentery fat, low conditionfactor and low hepato-somatic index. This evidence combined suggests that normal feedingmay not have been occurring at the time of capture or in the recent past in the fish caughtfrom Hamilton Point.For further details and statistical analysis of histopathology findings, see Appendix C.Chemical residue testingThe detailed report outlining chemical residue analysis of barramundi and interpretation isfound in Appendix D. This section is a summary of the main findings. The measuredconcentrations of metals in barramundi gills from Gladstone Harbour were similar or lessthan those from the reference sites. The one exception was arsenic, which was twice aslarge in the Gladstone sites than the reference sites; however, this was not a statisticallysignificant difference (P > 0.05). There are no historical data for metals in barramundi liversand gills for similar Queensland estuarine and near-coastal sites available to compare withthe results from the current study.Analysis of barramundi livers showed that the mean measured concentrations of iron andcadmium in fish collected during the first sampling round were significantly (P ≤ 0.05) higherat some Gladstone sites compared to reference sites, and arsenic and zinc were significantlyhigher in Trip 2. The significantly higher levels of these metals were driven by results fromspecimens sampled at Hamilton Point. However, the small number of samples taken at eachsite and trip (n = 5) means that these results should be interpreted with caution. There wasno relationship between the concentrations of these metals and the external signs of health.Arsenic was regularly detected, and iron and zinc were occasionally detected in watersamples collected during the water quality monitoring conducted by the QueenslandGovernment. In contrast, cadmium was rarely detected during the Water Quality MonitoringProgram. The measurement of elevated concentrations of these metals does not necessarilyindicate a toxicological problem, but does indicate exposure to them.Comparison of concentrations of these metals in barramundi livers with toxicological effectsdata did not, overall, indicate that the concentrations in the barramundi liver were of concern.In addition, as there were no significant associations between these metals and observed36

fish health, it is unlikely that they were associated with the fish health issues in GladstoneHarbour. Overall, there is no strong evidence of a link between observed fish health at thetime of sampling and tissue residue concentrations.Mud crabsThe monitoring program collected information on catch composition (size and sex), presenceand severity of observed shell damage and abnormalities, as well as commercial catchinformation supplied by crabbers via commercial logbooks. This information was comparedwith historical data sets when possible to provide perspective.Sampling sites for mud crabs during Phase 1 included the Narrows, Port Development Area,Calliope River, South Trees Inlet (joins the lower reaches of the Boyne River), Rodds Bayand the Fitzroy River (reference site).Sampling sites for mud crabs during Phase 2 included the Narrows, Port Development Area,Hamilton Point, Calliope River, Rodds Bay, as well as the Fitzroy River and Bundaberg(reference sites).Commercial catchThe commercial catch of mud crabs in the Gladstone region between 2000 and 2012hasvaried, with the highest catch recorded in 2003 (163.11 t) and the lowest catch in 2012(67.65 t). In 2011, 112.66 t was landed, the fourth highest catch in the period (Figure 15).Weight (tonnes)180160140120100806040200Commercial mud crab catch for the Gladstone region2000 - 20122000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 15. The annual catch of mud crabs from the Gladstone region (CFish grid S30), as reported incommercial fisher logbooks (calendar years 2000 – 2012).Commercial fishing effort for mud crab in the Gladstone region (days fished with mud crabcatch recorded) closely follows the trend in catch, with peaks in catch corresponding withpeaks in fishing effort (Figure 16). Logbook data supplied by commercial fishers for 2012showseffort and catch has reduced from 2010 to a level similar to the period 2006 to 2008.37

4,000Commercial effort (days fished) for mud crab in the Gladstone region2000 - 20123,5003,0002,500Days2,0001,5001,00050002000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 16. The number of days fishing where commercial fishers have reported mud crab from theGladstone region (CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 –2012).Crab size structureConcerns were raised that fishers were not catching large numbers of juvenile (undersize)crabs during normal commercial fishing activities. These concerns were investigated, as theyrelate to recruitment and mud crab catches into the future for the Gladstone region.To illustrate the size structure of the catch, crabs caught during the program werecategorised as legal male, large female (≥ legal size limit for males) or undersize (males andfemales combined).Phase 1: During Phase 1, mud crab sampling was conducted between December 2011 andJanuary 2012. A total of 1435 crabs were caught with their size and sex recorded. Thepercentage of undersized crabs varied between sites within Gladstone, with the percentageof undersized crabs ranging between 34.67% (Rodds Bay) and 50.91% (South Trees Inlet).The percentage of undersized crabs caught in the Fitzroy River was 42.14%, while nosampling of mud crabs was conducted in Bundaberg during this phase of the program (seeFigure 17).38

Size structure of mud crabs observed at each site during Phase 1.Legal Male Crabs Large Female Crabs Undersize Crabs90%80%70%60%Percentage50%40%30%20%10%0%Fitzroy River(n = 140)Narrows(n = 607)Port DevelopmentArea(n = 150)LocationCalliope River(n = 108)South Trees Inlet(n = 55)Rodds Bay(n = 375)Figure 17. The size structure of mud crabs observed at each site during the Gladstone Fish HealthMonitoring Program Phase 1, December 2011 – January 2012. Note, (n) refers to the total number ofcrabs examined at each site.Phase 2: During Phase 2, mud crab sampling was conducted between April 2012 and July2012. A total of 1599 crabs were caught, with their size and sex recorded. Results show thata large proportion of the crabs caught within the Gladstone region were undersized crabs(43.59%) (Figure 18). The percentage of undersized crabs varied between sites withinGladstone, with numbers of undersized crabs ranging between 35% (Rodds Bay) and 55%(Port Development Area). The percentage of undersized crabs caught in the Fitzroy Riverand Bundaberg were 49% and 78% respectively (Figure 19).90%80%70%Size structure of mud crabs observed in the Gladstone region and reference sitesduring Phase 2.Legal Male Crabs Large Female Crabs Undersize CrabsPercentage60%50%40%30%20%10%0%Fitzroy River(n = 387)Gladstone(n = 959)Bundaberg(n = 253)LocationFigure 18. The size structure of mud crabs observed in the Gladstone region (sites combined) and thereference sites during the Gladstone Fish Health Monitoring Program Phase 2, April 2012 – July 2012.Note, (n) refers to the total number of crabs examined at each site.39

Size structure of mud crabs observed at each site Phase 2.Legal Male Crabs Large Female Crabs Undersize Crabs90%80%70%60%Percentage50%40%30%20%10%0%Fitzroy River(n = 387)Narrows(n = 147)PortDevelopmentArea(n = 161)Hamilton Point(n = 192)LocationCalliope River(n = 94)Rodds Bay(n = 365)Bundaberg(n = 253)Figure 19. The size structure of mud crabs observed at each site during the Gladstone Fish HealthMonitoring Program, Phase 2, April 2012 – July 2012. Note, (n) refers to the total number of crabsexamined at each site.A notable difference in the size structure between Phase 1 and 2 is the higher proportion ofall crabs caught that were legal male crabs during Phase 1 (summer), highlighting theseasonal nature of the mud crab fishery.The percentage of undersize crabs observed during Phases 1 and 2 of the program is verysimilar to information collected from a long-term, fishery-independent sampling programconducted between 2000 and 2009 along the Queensland coastline where 12 280 crabswere caught and their size and sex recorded. The number of crabs caught within theGladstone region during the fishery independent survey was 1492 with 40.48% undersize( see Figure 20).40

Size structure of mud crabs observed during fishery independent sampling2000 - 2009.Legal Male Crabs Legal Female Crabs Undersize Crabs90%80%70%60%Percentage50%40%30%20%10%0%Gulf(n = 1635)North QLD(n = 4371)Gladstone(n = 1492)Hervey Bay(n = 1898)Moreton Bay(n = 2884)LocationFigure 20. The size structure of mud crabs observed at sites monitored during a fishery independentsampling program conducted along the Queensland coastline 2000 - 2009. Note, (n) refers to the totalnumber of crabs examined at each site.As the size structure from this investigation is similar to historical data for Gladstone there isno evidence to suggest issues with local recruitment.Shell disease and abnormalitiesIn late 2011, commercial fishers were reporting that the occurrence of shell (carapace)abnormalities within the Gladstone mud crab fishery was increasing. These concerns wereinvestigated, as they relate to the health of crab stocks and the viability of commercialcrabbing in the region.Symptoms reported in the abnormal mud crabs were similar to those documented byAndersen (2000) for the description of rust spot shell disease. Lesions on the shell, called‘rust spots’, due to their orange colour, were described by Andersen (2000) along with theassociated histopathology.Rust spot shell disease was first reported by commercial crab fishers in 1994 in mud crabsfrom the Port Curtis area. Investigations into the prevalence of rust spot disease in the PortCurtis area across three sampling occasions between 1998 and 2001 determined prevalencehad declined from 18.3% during 1998/99 to 14.5% in 1999/2000 and dropped further to10.2% by the 2000/2001 sampling event (Andersen and Norton 2001). Although the authorsacknowledge that prevalences reported for 1998/1999 may have been inflated slightly due tothe recording of lesions caused by other factors. The average prevalence across all samplingevents was 14.3%. Andersen and Norton (2001) restricted their analysis to observations oflesions on the dorsal carapace. This information was used to provide a baseline for thepresent survey program.It should be noted that Fisheries Queensland figures include all shell abnormalities, as it wasnot possible to definitively identify rust spot shell disease in the field. Fisheries Queenslandalso documented shell abnormalities regardless of where they occurred on the crab shell(e.g. legs, claws, carapace). Examples of shell abnormalities observed in the field are shownin Figure 21.41

Figure 21. Examples of shell abnormalities (circled) observed in mud crabs during the Gladstone FishHealth Monitoring Program.Observational findings Phase 1Of the 1435 mud crabs observed during Phase 1, 5.33% were identified as having shellabnormalities from sites within the Gladstone region and 2.14% from the Fitzroy River. Nosamples were collected from Bundaberg during Phase 1.The percentage of crabs showing abnormalities varied between sites within Gladstone, withnumbers of abnormal crabs ranging from 1.87% (Rodds Bay) to 9.26% (South Trees Inlet)(Figure 22).Percentage of mud crabs showing signs of shell abnormalities at each site, Phase 1.NormalAffectedPercentage100%90%80%70%60%50%40%30%20%10%0%Fitzroy River(n = 140)Narrows(n = 607)Port DevelopmentArea(n = 150)Calliope River(n = 108)South Trees Inlet(n = 55)Rodds Bay(n = 375)LocationFigure 22. The percentage of mud crabs with shell abnormalities observed at all sites during theGladstone Fish Health Monitoring Program, Phase 1. Note, (n) refers to the total number of crabsexamined at each site.42

Laboratory testing Phase 1The mud crabs examined showed evidence of erosive shell disease. Signs of this diseaseare consistent with bacterial infection by Vibrio spp., which are naturally occurring in marinewaters. Photobacterium (Vibrio) damaselae was isolated in one mud crab sample. Thesebacteria are opportunistic and proliferate on damaged shell to cause erosion due to theirchitinolytic enzymes.Observational findings Phase 2Mud crab sampling during Phase 2 was conducted in April/May 2012 and was repeated inJune/July 2012. The results from these sampling events are recorded below, along with asummary for Phase 2.April/May 2012: A total of 853 crabs were observed during sampling from seven sites withinthe study area (385 in the reference sites and 468 within Gladstone). The prevalence of shellabnormalities ranged from 2.1% in the Rodds Bay area to 8.1% within the Port DevelopmentArea and averaged 4% in the reference sites and 4.4% in the combined Gladstone sites.June/July 2012: A total of 746 crabs were observed during sampling in June/July 2012 fromseven sites within the study area (491 in Gladstone and 255 in the reference sites). Theprevalence of all shell abnormalities ranged from 1.4% in the Fitzroy River area to 4.7%within the Narrows, and averaged 1.96% in the reference sites and 3.26% in Gladstone.Phase 2 summary: In total, 1599 crabs were observed during Phase 2 sampling from sevensites within the study area (640 in the reference sites and 959 within Gladstone). Of these,58 crabs (3.6%) were observed to have shell abnormalities.The prevalence of shell abnormalities as shown in Figure 23 ranged from 2.4% in the RoddsBay and Bundaberg Reference sites to 6.2% within the Port Development Area, andaveraged 3.2% in the reference sites and 3.9% in Gladstone. Grade 5 lesions (most severeshell damage – perforated partially or fully > 20 mm in diameter) were found throughout thestudy area, including both reference sites (Fitzroy River and Bundaberg).Percentage of mud crabs showing signs of shell abnormalities at each site Phase 2.NormalAffectedPercentage100%90%80%70%60%50%40%30%20%10%0%Fitzroy River(n = 387)Narrows(n = 147)PortDevelopmentArea(n = 161)Hamilton Point(n = 192)Calliope River(n = 94)Rodds Bay(n = 365)Bundaberg(n = 253)LocationFigure 23. The combined percentage of mud crabs with shell abnormalities observed at all sites duringthe Gladstone Fish Health Monitoring Program, Phase 2. Note, (n) refers to the total number of crabsexamined at each site.43

Of the 58 crabs observed to have shell abnormalities in the field, 9 (15.5%) had lesionslocated on areas of the shell other than the dorsal carapace (legs, claws, abdomen andthorax).Lesion grade ratings recorded by Fisheries Queensland field staff and those recorded byBiosecurity Queensland from June/July sampling showed a significant (P < 0.001) degree ofassociation, with the counts listed in Table 5. There was agreement between field andlaboratory assessments in 79.71% of these cases. Interestingly, the field assessment washigher than the laboratory assessment in 14.49% of these cases, whereas the reverse onlyoccurred in 5.8% of the crabs. These results provide confidence for the field assessmentsmade by Fisheries Queensland staff when recording the presence and severity of lesions inthe field for both Phase 1 and Phase 2 of the program.Table 5. Highest lesion grade for each crab as assessed by Fisheries Queensland in the field andBiosecurity Queensland in the laboratory. The grey numbers indicate agreement between the twoassessments.LaboratoryAssessmentLesion Grade(0 – 5)Field AssessmentLesion Grade (0 – 5)0 1 2 3 4 50 47 0 2 4 1 01 2 1 2 0 0 02 0 1 2 0 0 13 0 0 0 1 0 04 0 0 1 0 1 15 0 0 0 0 0 3Considering the prevalence of rust spot shell disease documented by Andersen and Norton(2001), the potential bias by Fisheries Queensland staff to overestimate the prevalence ofshell lesions during field observations, and that observations of shell lesions were notrestricted to the dorsal carapace during this program, these results of 3.9% shell lesionssuggest that the prevalence of rust spot shell disease has decreased from levels of 10.2%previously recorded by Andersen and Norton (2001).It was assumed that the number of crabs with shell lesions that were reported in Phase 1were not under reported, as the above statistical analysis of Phase 2 for mud crabs did notidentify under reporting, and the staff and field methods were the same between Phases 1and 2. The combined results of Phase 1 and 2 do not indicate changes in the frequencies ofoccurrence over the sampling period or sampling location in the Harbour. The level of shelllesions observed during Phase 1 and 2 was significantly lower than the 37.8% reported formud crabs from Gladstone by Landos (2012), and lower than Andersen and Norton (2001).Andersen and Norton (2001) found the highest average percentage in Gladstone of 18.3% in1998/99, which reduced to 10.2% in 2001.Laboratory findings Phase 2NecropsyThere was no significant difference (P = 0.68) between the hepato-somatic index ofapparently normal and abnormal crabs (0.057 ± 0.002 vs. 0.054 ± 0.002). Hepato-somaticindex (HSI) is defined as the ratio of liver weight to body weight. HSI is generally used as anindication of energy stores, and these results would suggest that the presence of shelllesions is not having a significant effect on feeding.44

The key variables analysed were the percentage abnormal carapace assessed in the fieldusing all 1599 observations (a Binomial proportions model), the distribution of countsaccording to highest lesion class (a Poisson linear model), and HSI (Normal model). In theseanalyses, ‘crab category’ was always significant (P ≤ 0.05), justifying its inclusion as astratifying term. The relevant means are listed in Table 6.Table 6. Mean values and standard errors (s.e.) for the crab categories.Hepato-somatic index s.e. Percent abnormal s.e.Female large 0.059 0.002 3.8 0.8Female small 0.064 0.004 1.8 0.7Male legal 0.043 0.003 7.1 1.7Male undersize 0.056 0.002 2.5 0.7Regarding locations and sampling events, the only significant (P ≤ 0.05) effect was for theHSI, where combined results for June/July (0.061 ± 0.002) were higher than those forApril/May (0.050 ± 0.002). Results for HSI are to be expected given HSI is generally highestduring the non-reproductive period for crabs (winter), and then decreases as crabs move intotheir reproductive season and energy stores are utilised in gonad production (Table 7).Table 7. Mean values and average standard errors for hepato-somatic index and per cent abnormal foreach location and sampling event.Location Hepato-somatic index Per cent abnormalApril/May June/July April/May June/JulyFitzroy River 0.046 0.060 4.5 1.4Bundaberg 0.050 0.061 3.7 3.1The Narrows 0.055 0.059 3.2 5.2Port Development Area 0.057 0.068 8.8 4.4Hamilton Point 0.050 0.061 3.1 3.7Calliope River 0.049 0.063 5.8 2.7Rodds Bay 0.043 0.056 2.3 2.4(average s.e.) 0.004 (average s.e.) 2.0June/July also sees similar or lower percentages of abnormal crabs than April/May for allsites except for the Narrows, while no significant difference was detected between Gladstoneand the reference sites.Possibly due to the low number of abnormal crabs (58), there were no significant effects forlocation or trip on the distributions of these counts (P = 0.14 and 0.13 respectively). Thedistribution of observed lesions across sites is shown in Table 8.Table 8. Lesion categories as observed in the field, according to categories described by Andersen, 2001.Where individual crabs had more than one lesion, only the most severe lesion was included.LocationLesion CategoryL1 L2 L3 L4 L5Fitzroy River 2 6 1 1 4Bundaberg 0 1 1 2 2The Narrows 1 0 2 2 1Port Development Area 2 5 1 1 1Hamilton Point 0 2 1 2 2Calliope River 1 1 2 1 0Rodds Bay 1 3 0 4 145

HistopathologyLipid storage vacuolation in the digestive organ (hepatopancreas) was evaluated accordingto the categories none, mild, moderate or severe. The location had no effect on thesevalues, however there was a significant reduction (P = 0.006) in lipid between April 2012 andJune/July 2012 sampling.When comparing the combined Gladstone sites with the combined reference sites, therewere no significant differences for any findings for the June/July 2012 sampling. However, inthe April 2012 sampling, crabs from the reference sites had significantly higher ratings of gillparasites (P = 0.015) and muscle lesions (P = 0.000) than the score for the crabs from thereference sites (P = 0.015). The findings of muscle lesions should be interpreted with cautionbecause there were only five crabs that were rated any score other than zero.The crabs assessed were essentially normal based on histopathology findings, other thansome shell lesions, which are not specific to Gladstone Harbour and occur naturally asdiscussed previously.Residue testingThe concentration of aluminium and selenium was found to be significantly higher in crabswith shell lesions than crabs without. As this occurred at all sites, these elevated metalconcentrations may be associated with mud crab with shell lesions in general, not just inGladstone Harbour.Comparison of measured concentrations of metals and metalloids in the hepatopancreascollected from the Gladstone Port Development Area and the Narrows were similar to othercatchments along the east coast of Australia when compared to historical data. The fact thatthe tissue concentrations from Gladstone were typical for the area indicates that the metalsinvestigated in tissues are not responsible for health issues particular to Gladstone Harbour.None of the 114 organic chemicals analysed were detected in the crabs from either of theGladstone harbour sites.For a detailed analysis of chemical residue results, please see Appendix D.MulletThe monitoring program collected information on mullet size and sex, presence and severityof observed signs of abnormalities, as well as commercial catch information supplied byfishers via commercial logbooks.Several species were encountered during sampling including sea mullet (Mugil cephalus),diamond scale mullet (Liza vaigiensis), goldspot mullet (Liza argentea) and bluespot mullet(Valamugil sp.). These species are commonly caught by commercial fishers in the region andare often known under different common names, so examples of each are shown in Figure24.46

Figure 24. Mullet species encountered during sampling (a) sea mullet; (b) diamond scale mullet; (c)goldspot mullet; (d) bluespot mullet.Sampling sites for mullet during Phase 1 included, Calliope River, Boyne River, Rodds Bay inthe Gladstone area and Bundaberg and the Fitzroy River (reference sites).Sampling sites for mullet during Phase 2 included the Calliope River, Boyne River (upper andlower reaches), Rodds Bay, as well as the Fitzroy River and Bundaberg (reference sites).Commercial catchThe commercial catch of mullet in the Gladstone region (CFish grid S30, as reported incommercial logbooks) between 2000 and 2012 peaked in 2004 at 30.36 t. Since 2004, thecatch has shown a downward trend, with 2012 the lowest year in the period (7.33 t)(Figure 25).Commercial mullet catch for the Gladstone region2000 - 20123530Weight (tonnes)25201510502000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 25. The annual catch of mullet from the Gladstone region (CFish grid S30), as reported incommercial fisher logbooks (calendar years 2000 – 2012).47

Commercial fishing effort for mullet in the Gladstone region (days fished with mullet catchrecorded) has varied since 2000, with a general downward trend. Effort in 2012 was thelowest recorded since 2000 (96 days) and is approximately a quarter of the effort recorded in2000 (451 days) (Figure 26). The number of fishers catching mullet during this period hasalso dropped from a peak in 2001 (30 fishers) to lows in 2012 (7 fishers). Although the catchand number of fishers has decreased, the catch per day fished (76.3kg/day) is aboveaverage for the period (62.3 kg/day) and is the fourth highest catch rate for the period.Days500450400350300250200150100500Commercial effort (days fished) for mullet in the Gladstone region2000 - 20112000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 26. The number of days fishing where commercial fishers have reported mullet catch from theGladstone region (CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 –2012).Observational findings Phase 1Between September 2011 and January 2012, approximately 149 mullet caught bycommercial fishers were examined for external signs of ill health both within Gladstone andat the reference sites. The majority of these fish were caught in Rodds Bay and theBundaberg area.All mullet caught at reference sites (37 fish) were observed to be in good condition, while 15of 112 fish caught within the Gladstone area were observed to have some abnormal skinconditions. These observations were made from a single catch each at Rodds Bay (13 of 40mullet) and Calliope River (2 of 2 mullet) (Figure 27).48

Conditions observed in mullet sampled at each site Phase 1Skin Colour Normal Skin Minor Redness Skin General Redness Ulceration100908070Number of fish6050403020100Fitzroy(n = 2)Calliope(n = 2)Rodds Bay(n = 110)Bundaberg(n = 35)LocationFigure 27. Conditions observed at each site sampled during Phase 1, September 2011 – January 2012.Note, (n) refers to the total number of fish examined at each site. Some fish may have been observed withmultiple conditions.Conditions observed consisted mostly of minor redness on the ventral surface/pin point redmarks, with a single fish from the Calliope River having significant skin discolouration. Onefish captured in Rodds Bay displayed an ulcerative lesion on the caudal peduncle (Figure28).49

Figure 28. (a) Mullet caught in Rodds Bay displaying minor redness/pinpoint marks on the ventral surfaceand caudal peduncle. The redness around the operculum was a result of the capture by net; (b) Mulletcaught in the Calliope River showing general redness; (c) Mullet caught in Rodds Bay with ulcerativelesion on the caudal peduncle.Histopathology and residue testing Phase 1Only two mullet were examined by necropsy and histology during Phase 1 of theinvestigation. They were submitted by a member of the public from the Gladstone area, butno location was provided. Both fish exhibited skin redness with inflammation of the dermisand loss of epidermis. No bacteria or fungi were present in the skin lesions. Some inorganicsediment was found in association with sloughed epidermal cells and mixed with sloughedcells between gill filaments. A range of minor lesions related to parasites were presentinternally, however these are normal.Mullet from the Kolan River (Bundaberg) and Turkey Beach (Gladstone Harbour) were testedfor chemical residues using gill, liver and muscle tissue from two fish from each site. Themetal concentrations for the reference site (Kolan River) exceed those from Gladstone formost metals (i.e. aluminium, iron, copper, zinc, arsenic, selenium, silver and barium). Othermetals, including lead and mercury, showed no obvious difference between the referenceand Gladstone Harbour sites. No organic chemicals were detected.Observational findings Phase 2Mullet sampling during Phase 2 was conducted in April 2012 and was repeated in June/July2012. The results from these sampling events are recorded below, along with a summary forPhase 2.50

April: Mullet samples were collected from all proposed sites with a total of 63 caught andexamined for signs of ill health. The species of mullet sampled at each site varied, with seamullet dominating the catch in Bundaberg and the Calliope River. The catch in the FitzroyRiver and the upper and lower reaches of the Boyne River were almost exclusively goldspotmullet, while bluespot mullet dominated the catch at Rodds Bay. A single diamond scalemullet was also caught at Rodds Bay. No significant signs of abnormalities were observed inany mullet caught during the April sampling events. Minor redness was observed on samplescollected from both the Fitzroy River (100% of samples) and Bundaberg (25%) referencesites, but was not observed in samples collected in Gladstone.The conditions observed in the Fitzroy River and Bundaberg consisted of minor redness andsmall pin point marks on the body and fins, as illustrated in Figure 29.Figure 29. Minor redness and pinpoint marks observed on fins and the body surface of mullet sampledfrom the Fitzroy River (a) and (b), and Bundaberg (c) and (d).A single mullet was captured in the lower reaches of the Boyne River with a fresh bite mark.It was most likely that the bite mark occurred while the fish was in the net. Given the type andrecent nature of the wound, this fish was recorded as being in good health for analysis(Figure 30).Figure 30. A mullet captured in the lower reaches of the Boyne River with a fresh bite mark51

June/July: Mullet samples were collected from all proposed sites with a total of 62 caughtand examined for signs of ill health. The species of mullet sampled at each site varied, withthe catch in Bundaberg, Fitzroy and the Boyne River being almost exclusively sea mullet. Asingle diamond scale mullet was sampled from the Fitzroy River. The mullet catch in theCalliope River and Rodds Bay was exclusively goldspot mullet.Mullet were observed with minor redness/red pin point marks on the body surface from allsites except Rodds Bay and the upper reaches of the Boyne River. This condition was mostprevalent in the Fitzroy River (six of 11 mullet sampled). The ventral pale pink areasobserved in three out of 11 mullet caught in the Calliope River were due to the captureprocess (removal from the net).A single mullet from the upper reaches of the Boyne River displayed both general rednessand lesions comprising of small areas of detached scales (see Figure 31).Figure 31. A mullet sampled from the upper reaches of the Boyne River displaying general redness andlesions comprised of small areas with detached scales.A single mullet caught in the Fitzroy River was observed to have ulcerative lesions on bothsides of its gill covering (operculum) and a cloudy eye (Figure 32).Figure 32. A mullet sampled from the Fitzroy River with ulcerative lesions on (a) the left and (b) right sideof the gill covering (operculum), exposing the gills. Note also the cloudy right eyePhase 2 summary: In total, 125 mullet were observed during Phase 2 sampling from sixsites within the study area (47 in the reference sites and 78 within Gladstone). Seven mullet(8.97%) observed from the Gladstone area showed external signs of abnormalities and theywere all caught during Trip 2. This value does not include three fish from the Calliope Riverobserved to have pale pink areas ventrally resulting from the capture process. In the52

eference sites, 22 mullet (46.81%) displayed external signs of abnormalities and these wereseen during both trips. The conditions observed are summarised in Figure 33.Conditions observed in mullet sampled at each site Phase 225Skin Colour Normal Skin pale pink areas ventrally Skin Minor RednessSkin General redness Cloudy Eye Areas of detached scalesUlceration20Number of fish151050Fitzroy River(n = 21)Calliope River(n = 18)lower Boyne River(n = 20)upper Boyne River(n = 20)Rodds Bay(n = 20)Bundaberg(n = 26)LocationFigure 33. The combined results for conditions observed in mullet at all sites sampled during Phase 2,2012. Note, (n) refers to the total number of fish examined at each site. Some fish may have beenobserved with multiple conditions.In mullet, skin lesions were far less prevalent than barramundi, with only two lesionsrecorded.Laboratory findings Phase 2Collectively, Gladstone sites appeared to contain healthier mullet than the reference sites inApril 2012, but in June/July 2012 the trend was reversed.In April, the only significant differences (P < 0.05) between Gladstone and reference siteswere that Gladstone mullet had significantly lower incidence of skin redness than thereference sites. The cause of the difference is not known. In June/July, fin and kidneyabnormalities were significantly more prevalent in Gladstone sites than reference sites (P

Sharks and raysThe monitoring program collected information on shark species, size, sex, presence andseverity of observed signs of ill health, as well as commercial catch information supplied byfishers via commercial logbooks.During Phase 1, no candidate species of shark had been identified, therefore all sharks andrays encountered during routine, at sea observing or targeted sampling events wereobserved for signs of ill health. During Phase 2, the sampling identified bull sharks as thecandidate species given they were the dominant species in the catch during Phase 1 (Figure34).Figure 34. Bull sharkSampling sites for shark during Phase 1 included the Calliope River, Rodds Bay, SpoilGround and nearby offshore waters, the Narrows, Hamilton Point, the upper and lowerreaches of the Boyne River, as well as the Fitzroy River and Bundaberg (reference sites).Sampling sites for shark during Phase 2 included the Calliope River, Boyne River (upper andlower reaches), Rodds Bay, as well as the Fitzroy River and Bundaberg (reference sites).Commercial catchThe commercial catch of shark species in the Gladstone region (CFish grid S30, as reportedin commercial logbooks for both line and net combined) between 2000 and 2011 peaked in2008 at 157.40 t. The recorded catch then declined sharply to 44.55 t in 2009 and hascontinued to fall to 26.32 t in 2011 (Figure 35). In 2012, minimal catch was reported (0.44 t).54

Weight (tonnes)180160140120100806040200Commercial shark catch for the Gladstone region2000-20122000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 35. The annual catch of shark from the Gladstone region (CFish grid S30), as reported incommercial fisher logbooks (calendar years 2000 – 2011). Data includes both line and net catch for theregion combined.Commercial fishing effort for sharks in the Gladstone region (days fished with shark catchrecorded) has varied since 2000 and follows a similar trend to catch with a peak of 706 daysin 2008. In recent years, there has been a decline in the number of fishers recording sharkcatch from a peak in 2004 (34 fishers) to 2011 (14 fishers) (Figure 36). Effort recorded vialogbooks for 2012 is very low, with only five fishers recording shark catch for a total of 11days fishing.800Commercial effort (days fished) for shark in Gladstone region2000-2012700600500Days40030020010002000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 36. The number of days fishing where commercial fishers have reported shark catch from theGladstone region (CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 –2011). Data includes both line and net effort for the region combined.55

Observational findings Phase 1Phase 1 sampling of sharks included observations made between September 2011 andFebruary 2012 across a range of species (227 sharks and rays). Table 9 lists the number ofeach species encountered during targeted sampling events or during routine, at-seaobserving trips.Table 9. The sharks and rays observed during Phase 1.SpeciesNumberBlacktip whaler complex (Carcharhinus limbatus & Carcharhinus tilstoni) 45Bull shark (Carcharhinus leucas) 65Creek whaler (Carcharhinus fitzroyensis) 5Eagle ray (Myliobatidae sp.) 3Giant shovelnose (Rhinobatos typus) 1Graceful shark (Carcharhinus amblyrhynchoides) 3Grey carpet shark (Chiloscyllium punctatum) 2Lemon shark (Negaprion acutidens) 18Milk shark (Rhizoprionodon acutus) 22Narrow sawfish (Anoxypristis cuspidatus) 2Pigeye shark (Carcharhinus amboinenis) 5Scalloped hammerhead (Sphyrna lewini) 7Sharks unidentifed in field (Carcharhinus spp.) 31Spinner shark (Carcharhinus brevipinna) 2Spot-tail shark (Carcharhinus sorrah) 1Stingrays (Dasyatis spp.) 2Weasle shark (Hemigaleus australiensis) 2Whitecheek shark (Carcharhinus coatesi) 1Whitespotted guitarfish (Rhynchobatus australiae) 9Zebra shark (Stegostoma fasciatum) 1Total 227During September 2011, nine sharks caught in the Fitzroy River by a commercial fisher wereobserved to be in good health. The sharks comprised a number of species, including bullsharks (2), creek whalers (3), pigeye sharks (3) and a blacktip whaler.A total of 87 sharks and rays were sampled in October and November 2011 from a range ofsites including the Fitzroy River (5), Narrows (1), Calliope River (16), Gladstone Harbour(21), Rodds Bay (43) and Bundaberg (1). Sharks showed a range of abnormal conditionsthroughout the study area, including the reference sites. A number of sharks showed varyingdegrees of skin redness generally on the flanks and underside of the body. Although thecause of the redness is unknown, sharks can display varying levels of red, discoloured skindepending on the length of time in the net and post-harvest handling. One stingray wascaught during a trawl survey that had an injured, shortened tail with the barb missing.A number of sharks also had areas of lighter skin discolouration, particularly between the firstand second dorsal fins. Closer examination revealed the placoid scales (tough modifiedscales found on sharks and rays) were absent, exposing the underlying layers of skin(epidermis and dermis). Some of these areas were bleeding and appeared to haveassociated scratch marks. Many of these sharks were observed to have numerous flatworm56

parasites (monogeneans) on the body surrounding the affected areas. While this conditionwas most commonly observed on bull sharks, it was not restricted to this species.A report from an independent expert stated ”it is well established that fish infected bymonogeneans on the skin may ‘flash’ (i.e. rub affected parts of the body against substrate),presumably in an attempt by the host to remove the source of irritation” (Whittington andChisholm, 2008). It is possible that the clinical signs (e.g. skin redness and scratch marksobserved in sharks) could be caused indirectly by the parasites. These parasites occurnaturally, although normal parasite loads are not well documented, so it is not possible todetermine whether observed infections are abnormal.In December 2011, 29 sharks or rays were captured from several sites including the CalliopeRiver (1), the lower reaches of the Boyne River (3) and in Rodds Bay (25). Of these, 14 wereassessed as being healthy with no skin discolouration, 13 sharks showed discolouration on asmall area of the body, one was described as showing a large area of skin discolouration andone displayed skin discolouration over the majority of the body. No sharks or rays showedulcerative lesions.In January 2012, of the 50 sharks and rays captured, 12 were caught in the Fitzroy River (allbull sharks) while the remaining came from sites within Gladstone including the CalliopeRiver (7), Hamilton Point (1), the Spoil Ground (11) and Rodds Bay (19). All the sharks fromthe Fitzroy River had ecto-parasites and skin discolouration. Within the Gladstone sites,some of the black tip whaler, bull, lemon, milk and weasel sharks caught had skindiscolouration. Most of these sharks also had ecto-parasites.In February 2012, at sea observations were conducted on two net fishing operations for twoconsecutive nights. The fishing was conducted at Rodds Bay and at Colosseum (halfwaybetween Rodds Bay and the mouth of the Boyne River). Over the four nights, the fisherscaught 52 sharks and rays. Of the 21 bull sharks caught, 19 had ecto-parasites and skindiscolouration. Four of the other sharks and rays exhibited redness on the ventral surface.Laboratory testing Phase 1Tissue from a total of 21 sharks was examined microscopically during Phase 1, with 21bacteriology tests. Five necropsies were conducted on whole sharks.Shark species examined during Phase 1 included lemon shark, bull shark, pig eye shark,whitecheek shark, blacktip whaler shark, graceful whaler shark and narrow sawfish. Thesharks examined had scale pocket hyperaemia (pockets of reddening), dermal haemorrhage(bleeding of the skin) and epidermal necrosis (cell death). When sections were taken and theskin lesions were observed under a microscope, observations included congestion in theconnective tissue immediately below the epidermal basement membrane, occasionalhaemorrhage into the epidermis and/or skin surface, and a mild inflammatory cell infiltrationinto the outer layer of the dermis. No bacterial or fungal pathogens were found that couldexplain these skin conditions on the sharks.Monogenean parasites found on the skin of several bull sharks were identified asDermophthirius maccallumi from the family Microbothriidae. Similar parasites were found ongraceful whaler sharks and pig eye sharks. This genus is known to cause disease in captivesharks, including excess mucous production and ulcerated skin lesions. This disease canlead to secondary infection by bacteria or viruses, but has also been implicated in skinlesions, denticle loss and other clinical signs in wild sharks without co-infection by bacteria orviruses.57

Encysted larval parasites were found internally, but were not expected to be causingsignificant effects on the health of the sharks.There was minimal or no detectable sediment in the gills of the shark samples.Observational findings Phase 2April: Bull sharks were collected from all proposed sites except for Rodds Bay. Sampling ofbull sharks proved difficult during the April period with a total of 23 caught. Anecdotalevidence from commercial fishers suggests that bull sharks start leaving the estuaries at thistime of year. Three sharks of other species were also caught (two blacktip whaler, one Creekwhaler) and their condition recorded.Conditions observed with sharks caught during the April sampling period consisted of skinredness and areas of detached scales predominantly between the first and seconc dorsalfins exposing the skin, as well as some scratch marks around the affected areas. No othersigns of ill health were observed at the time of capture. A summary of results is shown inFigure 37.Conditions observed in Bull Sharks sampled at each site during April, Phase 212Skin Colour Normal Skin Minor Redness Areas of detached scales (indicative of ectoparasites)10Number of sharks86420Fitzroy River(n = 7)Hamilton Point(n = 1)Calliope River(n = 3)lower Boyne River(n = 10)upper Boyne River(n = 1)Bundaberg(n = 1)LocationFigure 37 Conditions observed in bull sharks at the time of capture from all sites sampled during April2012. Note, (n) refers to the total number of sharks examined at each site. Some sharks may have beenobserved with multiple conditions.Skin redness was observed on sharks caught from most areas, including those caught at thereference sites. The redness observed at the time of capture was classified as minor rednessand was most obvious on the lighter ventral surface of the sharks. The extent of redness wasdifficult to quantify consistently, given redness increased significantly post mortem asillustrated in Figure 38. The two sharks pictured were captured in the Calliope River and bothwere classified as having no redness in the field immediately after capture. After beingindividually bagged and kept on ice for approximately four hours, both sharks displayedsignificant redness.58

Figure 38. Sharks caught in the Calliope River illustrating the increase in skin redness post mortem. (a)and (c) Shark #0005 and #0007 photographed in the Calliope River at the time of capture; (b) and (d) Thesame sharks photographed four hours later in the lab.Areas of skin between the first and second dorsal fin with placoid scales absent were alsoobserved on sharks from most areas, including reference sites (Figure 39). This symptomhas been previously associated with the presence of the monogenean parasiteDermophthirius maccallumi during Phase 1 sampling. Therefore, it was considered to beevidence that monogenean ecto-parasites were present even when the parasites themselveswere not seen. This assumption was also based on personal communications with Dr IanWhittington, who suggested that only adult monogeneans from the family Microbothriidaewould be visible on the skin in the field, as juvenile worms are generally located underneaththe scales and would be too small and hidden from sight for visual identification.Figure 39. Bull sharks caught at Bundaberg and the Calliope River showing evidence of ecto-parasites(lighter areas) between the dorsal fins.June/July: No bull sharks were collected during the June/July sampling event. Commercialfishers suggested that bull sharks leave the study area during the winter months.When considering the findings from the sampling program (Phase 1 and Phase 2) the mostcommon abnormalities observed in sharks and rays was skin redness, and lesions locatedbetween the dorsal fins. Pathology investigating skin redness during Phase 1 found noassociated pathogenic organisms, so the cause of this condition is unknown. Given sharks,59

in particular bull sharks, have been observed outside Gladstone both during this survey andduring routine observer trips, it is known that this skin condition is not unique to Gladstone.Furthermore, the degree of redness increases post mortem. It is thought that capturemethods and handling practises of sharks influence the degree of redness exhibited, butfurther work is needed to quantify this effect.The parasitic monogenean, Dermophthirius maccallumi, identified during the program isknown to cause symptoms similar to those observed on many sharks sampled during theprogram, and may also indirectly be the cause of many of the scratch marks around theaffected areas if sharks are ’flashing’ or rubbing themselves on hard substrates in an attemptto rid themselves of the irritation. Given there have been no previous studies onDermophthirius maccallumi on wild sharks that quantify a normal parasite load, it is difficult todetermine whether the number of parasites observed on sharks during this program areabnormal. What is known is that the presence of these parasites and the associated skinlesions are not unique to the Gladstone area.Laboratory findings Phase 2NecropsyA total of 12 sharks were received for laboratory testing, all during April. Only four of thesewere from Gladstone Harbour (three from Calliope River and one from Hamilton Point),seven were caught in the Fitzroy River and one from Bundaberg. The sharks from GladstoneHarbour, particularly the three from the Calliope River, had higher skin condition ratings thanother sharks, although as discussed above this redness was not apparent on capture. It ispossible that subtle differences in handling, such as net mesh size, soak time and time spentin transit, could have influenced this post-mortem skin redness, but particularly with such lownumbers it is not possible to draw any conclusions from these observations.No significant abnormalities were observed internally in any of the sharks during necropsy.HistopathologyLiver, skin and gills from 10 sharks were assessed histologically, including the only four thatwere caught from Gladstone Harbour, five from the Fitzroy River and one from Bundaberg.Common pathology indicators for each tissue were rated from zero to three to allowquantitative comparison to be made. There were no significant abnormalities observed inshark tissues, and there were no significant differences between fish from Gladstone and thereference sites for any tissues. Microscopic examination of skin tissue identified mildinflammation and bleeding in one shark from Bundaberg, mild inflammation in a shark fromHamilton Point, and mild bleeding and degeneration of skin tissue in a shark from theCalliope River. No causal agent was identified for the observed skin redness.Samples collected for chemical residue testing were archived for future testing if required.Banana prawnsThe monitoring program collected information on prawn size, presence and severity ofobserved signs of ill health, as well as commercial catch information supplied by fishers viacommercial logbooks. Samples were collected and provided to Biosecurity Queensland forfurther testing.Sampling sites for banana prawns during Phase 1 and Phase 2 included Gladstone Harbourand waters offshore of Bundaberg and the Fitzroy River (reference sites).60

Figure 40. Banana prawns, a targeted species during the sampling program.Commercial catchThe commercial catch of banana prawns in the Gladstone region (CFish grid S30, asreported in commercial logbooks for both otter and beam trawl combined) between 2000 and2011 peaked in 2003 at 125.53 t. The recorded catch declined to 6.36 t in 2007 and thensteadily increased to 109.18 t in 2011, the second highest recorded catch since 2000 (Figure41). In 2012, the reported banana prawn catch in Gladstone had dropped to 12.88 t..61

Commercial banana prawn catch for the Gladstone region 2000 - 2012140120Weight (tonnes)1008060402002000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 41. The annual catch of banana prawns from the Gladstone region (CFish grid S30) as reported incommercial fisher logbooks from calendar years 2000 – 2011. Data includes both otter trawl and beamtrawl catch for the region.Commercial fishing effort for banana prawns in the Gladstone region (days fished withbanana prawn catch recorded) has varied since 2000 and follows a similar trend to catchhowever between 2008 and 2011 similar effort was recorded while the commercial catchrose steadily (Figure 42). The number of fishers catching banana prawns during this periodpeaked in 2003 (30 fishers). Effort recorded via logbooks for 2012 was very low with onlyseven fishers recording banana prawn catch.Commercial effort (days fished) for banana prawns in Gladstone region2000 - 2012Days80070060050040030020010002000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012YearFigure 42. The number of days fishing where commercial fishers have reported banana prawn catch fromthe Gladstone region (CFish grid S30), as reported in commercial fisher logbooks (calendar years 2000 –2011). Data includes both otter trawl and beam trawl catch effort for the region.62

Observational findings Phase 1Between October 2011 and February 2012, catches of banana prawns were observed andinformation on the external signs of health was recorded from a range of sites by FisheriesQueensland observers, as well as during targeted banana prawn surveys.During October 2011, observers noted the small number of banana prawns caught (sixprawns) while aboard a boat fishing for scallops in the Harbour were all in good health.Trawl sampling in Gladstone Harbour in November 2011 caught 85 banana prawns with asingle prawn infected by a parasitic isopod located under the carapace and one prawnshowing signs of minor shell erosion. The remaining 83 prawns were observed to beexternally normal.A further four prawns were collected from the Gladstone Fish Markets. These prawns werecaught on 11 November 2011 in an area opposite Auckland Point Wharf and provided toFisheries Queensland by a local seafood wholesaler. A preliminary examination of one ofthem revealed a large parasitic isopod under the carapace. These prawns were provided toBiosecurity Queensland for further testing.In December 2011, two banana prawns were caught in the trawl sampling in GladstoneHarbour. These prawns were in good condition. In the Fitzroy River, 173 banana prawnswere caught and all observed to be in good condition.During February 2012, Fisheries Queensland observers working aboard commercial fishingboats noted that of the 25 kg of banana prawns caught in the Gladstone Harbour, twoprawns were infected with the gill isopod, while catches in the Fitzroy River (56 kg) andBundaberg (4 kg) were all observed to be in good health.While the above results refer specifically to banana prawns, a number of other prawn specieswere caught during sampling with samples provided to Biosecurity Queensland for furthertesting. Other species included eastern king, endeavour, rainbow and tiger prawns. Based onthe same methods applied in the assessment of banana prawns, no signs of ill health wereobserved in these other species.Laboratory testing Phase 1Prawns showed evidence of shell erosion due to Vibrio spp. A number of prawns includingbanana, eastern king and tiger showed evidence of endoparasite infection by immaturestages of tapeworm and gregarines. These parasites are normal in wild prawn populationsand would not have had a significant impact on the health of prawns (Owen, 1983; Owen &Rothlisberg, 1985). As noted in the field observations, the isopod observed under thecarapace was identified as a bopyrid isopod and is also common in wild prawn populations.Tails of four healthy looking prawns and three prawns with visible abnormalities caused bybopyrid isopods from Gladstone Harbour were tested for chemical residues. Organicstesting found no detectable levels of any chemical. Metal testing showed some differencesbetween the samples, however without replication or any samples from a reference site,meaningful interpretation of this data is not possible.Please refer to www.qld.gov.au/gladstoneharbour for a complete description of the pathologyresults, and the chemical residue test results.63

Observational findings Phase 2April 2012: During trawl sampling in April, the catch of banana pawns at each location was150 kg or greater. Due to the size of the catch, observations were made from a sub-sampleof at least 100 prawns at each location.A single prawn was collected in the Gladstone Harbour and one from the offshore FitzroyRiver site with shell erosion. A single prawn at Bundaberg and a single prawn from theoffshore Fitzroy River sites were observed to be affected by an isopod parasite under the gillcarapace. No other signs of ill health were observed in banana prawns caught during thetrawl sampling.June/July 2012: Observations were made from the complete banana prawn catch at boththe Gladstone Harbour (approximately 1.5 kg) and offshore Fitzroy River sites (approximately2 kg). Due to the size of the catch at Bundaberg (approximately 200 kg) observations weremade from a sub-sample of 100 prawns.Two prawns collected from the Gladstone Harbour and one prawn collected from theoffshore Fitzroy River sites had minor shell erosion. Two prawns collected from the offshoreFitzroy River site and a single prawn collected from Bundaberg had isopod parasites locatedunder their gill carapace (see Figures 43 and 44). No other signs of ill health were observedin banana prawns caught during the trawl sampling.Figure 43. Banana prawns collected during the sampling program from (a) the Gladstone Harbour and (b)offshore Fitzroy River sites showing signs of shell erosion (circled).64

Figure 44. Banana prawns collected during the sampling program with the isopod present under thecarapace (a) Prawn caught from the offshore Fitzroy River site with an isopod present under the carapace(circled); (b) Prawn caught at Bundaberg with part of its carapace removed to reveal the isopod.Bopyrid isopods and larval stages of tapeworms are common parasites of prawns and are anormal occurrence in wild populations along the Queensland coast. The prevalence ofbopyrid isopods observed during this sampling program was low across all areas (Gladstoneand reference sites), suggesting a normal background prevalence in the populationthroughout. This information, along with the low prevalence of shell erosion observedthroughout the study, provide no evidence of health issues in the wild banana prawnpopulation in the Gladstone region.Laboratory testing Phase 2Samples collected for histopathology and residue testing during Phase 2 have been archivedfor future testing if required.Trawl caught finfish speciesObservational findings Phase 1In November 2011, trawl sampling was conducted in Gladstone Harbour targeting prawns.Approximately 350 of each of the following finfish species were caught as incidental catch:river jew, anchovy, Australian threadfin and ponyfish. Three river jew displayed some minorredness on the ventral surface, and all other fish caught were in good condition. Three out of31 herring caught showed injuries on the dorsal side between the dorsal fin and head.In December 2011, trawled finfish (bycatch) were observed in catches from the Fitzroy River(149) and Gladstone Harbour (62). All fish caught by trawling were assessed as beinghealthy with no skin discolouration.In February 2012, Fisheries Queensland observers monitored the health condition of fishcaught in Gladstone by trawl net while aboard commercial fishing boats. In total,approximately 490 bycatch fish were observed. None of these fish showed any signs of skindiscolouration.65

Laboratory testing Phase 1Seven river jewfish and one ghost grinner were submitted to Biosecurity Queensland inDecember 2011 from Port Curtis where they were collected floating on the surface inresponse to a report of a fish kill. Chemical residue testing and pathology investigations wereconducted on these fish, and found no likely cause of death. Based on these findings, thefish were most likely caught by fishers and subsequently discarded as bycatch. Seewww.qld.gov.au/gladstoneharbour for more details.Observational findings Phase 2April 2012 sampling: The following numbers of the candidate fish were observed at thethree sites during the April sampling event. Bundaberg: no grinner; 105 herring; 105 Australian threadfin Gladstone Harbour: 20 grinner; 20 herring; 100 Australian threadfin. Fitzroy River: 16 grinner; 15 herring; 100 Australian threadfin.No significant signs of ill health were observed in trawl fish species caught during the trawlsampling from offshore Bundaberg, the Gladstone Harbour or from offshore of the FitzroyRiver.June/July sampling: The following numbers of the candidate fish were observed at thethree sites during the June/July sampling event: Bundaberg: 18 grinner; 10 herring; 25 Australian threadfin. Gladstone Harbour: 26 grinner; 100 herring; 100 Australian threadfin Fitzroy River: 100 grinner; 100 herring; 100 Australian threadfin.Two Australian threadfin caught at Bundaberg and a single Australian threadfin caught inGladstone Harbour had isopod parasites present either on the gills or body surface. Otherthan the presence of the isopod parasites, no signs of ill health were observed in the fishspecies collected during the trawl survey at the three sites.Laboratory findings Phase 2The grinner, Australian threadfin and herring specimens were preserved fresh on-board atrawl vessel by a veterinarian with aquatic animal expertise. Other than gross observationsmentioned above and length records, no necropsy data was collected for these species. Asub-sample of grinner was chosen for histopathology and chemical residue testing, and allother samples were archived for further testing if required.HistopathologySkin, liver and gill samples from each of the 25 fish were assessed histologically and rated inthe same way as has been described previously for barramundi. There were no significant ornotable differences between Gladstone Harbour and the reference sites for any tissues – thefish appeared normal. These fish tissues were preserved immediately following euthanasia toeliminate any deterioration of tissues and allow detection of very subtle abnormalities. Therewas no evidence of any unusual gill or skin irritation occurring in the fish from GladstoneHarbour.Chemical residue testingAluminium, iron, copper and zinc were significantly higher (P < 0.05) in grinner musclecollected at Bundaberg (reference site) compared to other sites, but there were no otherdifferences in metal concentrations between sites. Differences in metal concentrations66

etween sites is normal and expected. This data indicates that the grinner caught inGladstone Harbour had not been exposed to any chronic elevated levels of metals.PelagicsSampling sites for pelagic fish species during Phase 1 included the Calliope River, GladstoneHarbour, Boyne River (upper and lower reaches), Rodds Bay, the Spoil Grounds as well asthe Fitzroy River and Bundaberg (reference sites).Sampling sites for pelagic fish species during Phase 2 included the Calliope River, BoyneRiver (lower reaches), Spoil Grounds, as well as the Fitzroy River and Bundaberg (referencesites). The candidate pelagic species for Phase 2 were queenfish due to conditions observedin this species during Phase 1 (Figure 45).Figure 45. Queenfish caught at Bundaberg during the sampling program.Observational findings Phase 1During October 2011, golden trevally (1), snubnosed dart (2) and queenfish (2) were caughtin Rodds Bay while targeting barramundi. These fish were all considered to be in good healthwith no signs of redness or lesions observed. A giant trevally was also caught in the upperreaches of the Boyne River that had minor redness.During November 2011, while conducting a targeted prawn survey in the Harbour, onequeenfish was captured and displayed damage to the snout and lower jaw, with someassociated minor redness. Twenty small trevally (mixed species) were also caught during thetrawl survey, but all were described as being in good health with no discolouration or lesions.A single queenfish was also caught during netting activities with no signs of ill health.During December 2011, an oxeye herring caught in the Boyne River and a queenfish caughtin Rodds Bay showed no signs of ill health. Sampling in the Calliope River caught a singleoxeye herring observed to be in good health, however two of the six queenfish caught wereobserved to have skin discolouration with minor redness or red pin point marks on the bodyor fins.During net sampling in January 2012, three golden trevally were caught in Bundaberg, twoqueenfish in the Calliope River and two queenfish in the Boyne River. None of these fishshowed external signs of ill health. At the Spoil Ground, 42 queenfish were caught, of which67

four displayed signs of redness. These fish were also infected with calanoid copepods, with41 of the 42 queenfish having copepods present on the body surface.During February 2012, observations were conducted aboard commercial net fishingoperations at Rodds Bay and at Colosseum (halfway between Rodds Bay and the mouth ofthe Boyne River). Two of the 18 golden trevally caught showed minor redness ventrally,while all five giant trevally caught were in good condition. All 40 snubnosed dart and 32queenfish caught were in good health. No parasites were obvious on the skin of thequeenfish examined. These results were recorded as Rodds Bay and summarised withPhase 1 data in Figure 46.Conditions observed in pelagic fish species at each site Phase 1Skin Colour NormalSkin Minor RednessNumber of fish454035302520151050Oxeye Herring(n = 1)Queenfish(n = 8)Queenfish(n = 1)Trevally(n = 20)Giant Trevally(n = 1)Queenfish(n = 2)Oxeye Herring(n = 1)Queenfish(n = 42)Giant Trevally(n = 6)Golden Trevally(n = 17)Snubnosed Dart(n = 42)Queenfish(n = 35)Golden Trevally(n = 3)Queenfish(n = 1)Calliope River Gladstone Harbour upper Boyne River lowerBoyneRiverSpoilGroundRodds BayBundabergSpecies and LocationFigure 46 . Conditions observed in queenfish and other pelagic species at each site during Phase 1. Note,(n) refers to the total number of fish examined at each site.Histopathology and residue testing Phase 1Two trevally and two queenfish were submitted to the Biosecurity Queensland laboratoryduring Phase 1. The trevally were submitted due to their skin redness. One of these trevallywas provided in a private submission. Mild inflammation of the dermis and loss of epidermiswas observed, but no bacterial or fungal pathogen was identified in the skin lesions. Twoencysted parasites were found in the dermal layer of skin from one trevally.The queenfish were submitted in response to concerns about cyanide discharge from localindustry. The fish were effectively normal based on necropsy and histological examinations.68

Observational findings Phase 2April 2012: It was difficult to collect large numbers of the candidate species (queenfish) fromall sites during the April sampling period. Consequently, other pelagic species capturedduring sampling were also observed and selected for necropsy examination. Examples ofother pelagic species included grey mackerel, oxeye herring and giant trevally.All fish were observed to have normal skin appearance (no redness). A blue threadfin wasalso caught at the Spoil Ground. Although it might not be considered a pelagic species, itwas observed to have a laceration and was retained for necropsy examination.Four of the five queenfish caught at the Spoil Ground had two to three copepod parasiteseach. Copepods were also present on the queenfish caught offshore from the Fitzroy River.One queenfish caught at Bundaberg had a minor lesion consisting of a small area of scalesmissing.June/July 2012: Queenfish were observed at both the offshore Fitzroy River and Bundabergsites, but were not captured at the Spoil Grounds during the June/July sampling event.Queenfish were captured during netting activities in both the Calliope and Boyne Rivers.Observations of these fish were recorded and included in the results.A gold spot grouper, slate sweetlip and brown sweetlip were caught while sampling the spoilgrounds. These species are all demersal species and were assessed for external signs of illhealth according to the same protocols as all other species. The grouper and sweetlipsappeared to be in good health with no signs of skin discolouration or lesions observed.Queenfish with copepod parasites were observed from all areas except the Calliope River.The single queenfish caught in the Calliope River did have an injury thought to havehappened during the netting process and was classified as a physical injury. While theprevalence of copepods observed on fish from two of the four sites was as high as 100%,most fish only had a few visible copepods present. Other than the presence of the copepodparasites, all fish appeared to be in good condition.Phase 2 summary: Of the 83 pelagic fish observed during Phase 2, only two fish wereobserved with abnormalities including one queenfish with a small area of scales detachedand one queenfish with a physical injury thought to have occurred during capture. Numerousqueenfish were observed to be infected with calanoid copepods both from the reference sitesand Gladstone. It is thought the copepod is a normal parasite for this host, however thenormal parasite prevalence is not known. Results of observations made for pelagic speciesduring Phase 2 are illustrated in Figure 47.69

Conditions observed in pelagic fish species at each site during Phase 2Skin Colour Normal Small area of detached scales Physical damage Copepod parasites observed403530Number of Fish2520151050GreyMackerel(n = 1)Queenfish(n = 12)OxeyeHerring(n = 1)Queenfish(n = 14)Queenfish(n = 4)GiantTrevally(n = 5)Bonito(n = 1)GreyMackerel(n = 4)Queenfish(n = 5)Queenfish(n = 36)Fitzroy RiverCalliopeRiverlower BoyneRiverSpoil GroundsBundabergSpecies and LocationFigure 47. Conditions observed in queenfish and other pelagic species at each site during Phase 2. Note,(n) refers to the total number of fish examined at each site. Some fish may have been observed withmultiple conditions.Histopathology and residue testing Phase 2Tissue samples collected for histopathology and residue testing have been held in archive forfurther investigation if required.Other speciesDuring Phase 1 of the sampling program, a large number of other species were observed, assummarised in Table 2. This included finfish, crustaceans and molluscs. These observationswere made during routine observer trips aboard commercial fishing vessels, as well asduring sampling trips targeting other species using net, trawl and pot fishing apparatus. Theresults of observations and associated pathology on other species is summarised below.ScallopsObservational findingsTargeted sampling of scallops was conducted offshore from Gladstone with a commercialfisher on 29-30 October 2011. Sampling sites included areas adjacent to the Spoil Grounds.Some scallops identified by the fisher as appearing abnormal were preserved forhistopatholgy. Signs of abnormalities included discoloured digestive glands and brittle shells.A sample of scallops was also retained frozen for residue testing.In November 2011, scallops caught in waters offshore of Bundaberg by a commercial fisherwere collected and forwarded to Biosecurity Queensland. These scallops were apparentlyhealthy and were transported fresh for histopathology and residue testing.70

Further scallop samples were collected in February 2012 from both Bundaberg and the SpoilGrounds offshore from Gladstone. These scallops appeared healthy, were frozen andforwarded to Biosecurity Queensland for residue testing.Histopathology and chemical residue testingScallops collected from offshore of Gladstone and offshore from Bundaberg (referencesamples) in October and November 2011 were examined histologically and were normal inall organs and tissues, with some parasitic infestation internally and in the gills of scallopsfrom both sites. Composite samples of scallops from Bundaberg and Gladstone Harbourshowed very little difference in tissue metal concentrations. Lead, mercury and silver were allbelow detection limits and, of the remaining 10 metals, only iron differed to any extentbetween samples, being higher in the reference site. No organic chemicals were detected.Scallops collected in February 2012 were tested for relative concentrations of total andinorganic arsenic in tissues using four samples: two with pooled adductor muscles frommultiple individual scallops and two with pooled visceral tissues. No organic contaminantswere detected in any scallop samples. Metal concentrations were similar in the samplesfrom Gladstone and Bundaberg except for iron, aluminium, nickel and copper, all of whichwere higher in the Bundaberg (reference) samples.In testing for metal residue levels, arsenic levels were reported for total arsenic. However, insome instances, the appropriate standards are set for inorganic arsenic, and total arsenicshould not be directly compared with standards for inorganic arsenic. On the basis of the firsttest results, this matter was further investigated to determine if there was a risk to humanhealth. Results showed that the level for inorganic arsenic (As) in the two muscle sampleswas higher than the maximum level (ML) set in the Australia New Zealand Food StandardsCode. The current legal limit for inorganic arsenic in molluscs is 1 mg/kg (wet weight), whichis set out in the Table to Clause 2 of Standard 1.4.1 of the Code. The test results reported bythe Queensland Health Forensic and Scientific Services Laboratory that are above the MLare as follows: Bundaberg scallop muscle: 1.5 mg/kg as inorganic As (wet weight),Gladstone scallop muscle: 1.9 mg/kg as inorganic As (wet weight).Inorganic As is the combination of Arsenite (AsIII) as As (wet weight) and Arsenate (AsV) asAs (wet weight). For this particular scallop species, the viscera are not considered to be partof the edible portion of the animal and therefore the Code does not apply to these samples.Representatives from Queensland Health, Department of Agriculture, Fisheries and Forestryand Safe Food Queensland discussed the results. To inform their discussion, the grouprequested additional information including sampling information; history of results fromNational Residue Surveys; comparisons with As levels in other foods and associated riskassessments; importing country requirements for seafood sourced from Australia; and typicalprocessing procedures involving washing shucked scallops with filtered and sterilisedseawater, which may reduce the final level of As in market-ready scallops.The conclusion reached by these representatives on the significance of the test results wasthat they do not present an acute health risk. It is the chronic, cumulative effects of theexcess intake of As that could potentially be a concern. However, the frequency and level atwhich scallops are consumed indicates it is unlikely to be a public health issue.71

FinfishObservational findingsDuring September 2011, 52 fish were caught and observed in the Fitzroy River includingbeach salmon (2), blue threadfin (20), catfish (11), king threadfin (13), pikey bream (4) andtripletail (2). No skin discolouration, lesions or other signs of ill health were observed.During October 2011, blue threadfin (3) caught in the Narrows and Rodds Bay and a kingthreadfin caught in the Narrows were observed to be in good health, while a catfish and kingthreadfin caught in the Boyne River showed general redness on the skin and fins.During November 2011, 401 finfish were observed from the Fitzroy River (9), Hamilton Point(1), Rodds Bay (200) and Bundaberg (191). Of these fish, a single whiting caught inBundaberg was observed to have a lesion that was described as a recovering wound, and 60of the 200 whiting caught in Rodds Bay displayed some minor skin condition. These mild skinconditions consisted of areas of slight discolouration, often with a single red dot under ascale.During December 2011, observations were made of 87 finfish caught during both netting (52)and crabbing trips (35). During net fishing trips, other species of finfish were caught in theCalliope River (7), Hamilton Point (1), Rodds Bay (33) and the lower reaches of the BoyneRiver (11). Of these, eight fish showed signs of minor redness, while a black jew caught atHamilton Point had more general skin discolouration. During crabbing trips in the CalliopeRiver, the Narrows and South Trees Inlet, three fish were observed to have minor redness.These fish were a pikey bream from South Trees Inlet and two cod from the Narrows.In January 2012, observations were made of 42 fish from the Gladstone area. Of these, oneblubber lip (sweetlip) caught at Hamilton Point showed small areas of discolouration. Of 27catfish captured from the Calliope River and the upper reaches of the Boyne River, 19exhibited small areas of skin discolouration with an additional two showing discolourationmore broadly across the body. One of these fish also had an abnormal eye. Four catfishcaught in the upper reaches of the Boyne River had lesions present on the skin. A further 87fish were observed from the Fitzroy River and Bundaberg (reference sites). The specieswere dominated by catfish (27) and bony bream (42) with javelin fish (17), and a spotted scatwas also captured. All fish observed in the reference sites were considered to be in goodhealth with no abnormalities observed.During February 2012, observations were conducted aboard commercial net fishingoperations at Rodds Bay and at Colosseum (halfway between Rodds Bay and the mouth ofthe Boyne River). Blue salmon (11), catfish (7) and a single sweetlip were amongst thecatch. A single catfish was observed with redness to the pelvic fins, with the remaining fishshowing no signs of ill health.Histopathology and chemical residue testingTesting has also been carried out on a wide range of bony fish species including whiting,goldspot grouper, spangled emperor, flathead and scat. Many of these samples were privatesubmissions.Skin abnormalities from other finfish species were generally mild and not due toNeobenedenia sp. Findings included localised skin inflammation, skin erosion, fibrosis(scarring), reddening, dermal haemorrhage (bleeding of the skin), epidermal necrosis (celldeath) and oedema (swelling).72

No bacterial, parasitic or fungal pathogens were found that could explain the skin conditionson other finfish species.A range of bacteria were isolated from the fish, for example Moraxella spp, Micrococcus spp,Proteus vulgaris, and Pseudomonas spp. The lack of distinct invasion or proliferation on theskin and muscle lesions suggests that these bacteria were opportunistic and secondarycolonisers following the initial skin damage. These bacteria are not considered to be fishpathogens (i.e. they are not primary infectious agents that cause disease in fish). They arenormal microorganisms of the marine environment.73

ConclusionsIn 2011, a wide range of fish species in Gladstone were reported with health issuessupporting the hypothesis that the Gladstone Harbour ecosystem as a whole was understress. Barramundi provided the clearest and most consistent evidence for the cause of illhealth. In the other species, there appeared to be a range of observed symptoms. Anecdotalreports were received of increased levels of rust spot disease in mud crabs, firstcharacterised by Andersen and Norton (2001).An investigation began in September 2011 to determine the cause of the reported issues andspecifically to address public concern regarding the potential impact of industrial activityincluding dredging in Gladstone Harbour. The investigation, conducted by DAFF, was part ofa whole-of-government response.BarramundiIn September 2011, an external parasitic fluke Neobenedenia sp, was identified inbarramundi from the Gladstone Harbour. Neobenedenia was identified as the cause of theobserved eye problems and contributed to a number of the observed skin conditions inbarramundi.This parasite has previously been found in Queensland waters in the Hinchinbrook Channelbetween Hinchinbrook Island and mainland Queensland where barramundi are in highdensities (Deveney, et al, 2001), and had resulted in the loss of approximately 200 000cultured fish.An additional factor was the significant number of barramundi (estimated 30 000) washed outof the Awoonga Dam in the flooding during 2010/11. Many of these fish would have receivedphysical damage as a result of being washed over the spillway. This damage was reportedby recreational fishes as physical damage and was documented in necropsies of barramundiundertaken by a veterinarian with specialist skills in fish and aquatic animals in 2012 after asecond overtopping of Awoonga Dam. The fish showed damage to the jaw and head andusually presented with unilateral injuries to scales that appeared like a grazing along thebody.In the winter of 2011, the population of barramundi in the Boyne River was under extremestress. The densities of these large predatory barramundi in the Boyne River were at higherlevels than the system could naturally sustain, resulting in fish in poor condition and lowergrowth rates (Sawynok, et al, 2013). Their immune systems were further stressed by lowwinter water temperatures, resulting in an environment where the natural population of theparasite Neobenedenia could rapidly proliferate, and in turn resulting in observed eyeproblems and increases in fish exhibiting skin conditions (e.g. redness, loss of individualscales).As the water temperature increased, and the densities of barramundi were reduced byfishing (130 t of barra were caught by commercial fishers in the third quarter of 2011), theremaining barramundi became healthier. In 2012 and in 2013, other than barramundishowing relatively minor physical damage, no significant health issues have been reportedfor barramundi caught in the Gladstone region.These results are consistent with flooding and the significant number of barramundi beingwashed into the Boyne River. If the cause of the fish ill health continued beyond 2011, thehealth of the fish would not be improving and Neobenedenia infections would have beenseen in September 2012.74

Mullet and other finfishUnlike barramundi, there did not appear to be a single symptom that characterised what wasobserved in other fish including mullet in Gladstone. Symptoms observed included loss ofscales, small red pin point wounds, red colouration of fins and body areas, and increasedlevels of parasites.One catch of queenfish in Phase 1 had a high frequency of fish with a high intensityinfestation of Lepeophtheirus spinifer (sea lice). From the total catch of queenfish, 41 of the46 fish had sea lice. This is the same catch of queenfish that was examined by Matt Landosand presented in his report.In Phase 1, in addition to this catch, a total of 56 queenfish were observed in Gladstone. Ofthese fish, two fish were reported as showing no external signs of ill health, but parasiteswere not recorded as either being present or absent. Of the remaining fish, 13 were recordedas having no external visual parasites. The only time that parasites were recorded onqueenfish in Phase 1 in Gladstone was this particular catch.In Phase 2, of the 23 queenfish examined from the Gladstone region, only eight hadectoparasites. For the reference sites, 26 out of the 36 queenfish from Bundaberg examinedhad ectoparasites, while two of the 12 queenfish from the Fitzroy River had parasites.It appears from the sampling conducted by Fisheries Queensland over an extended period oftime that the percentage of queenfish with parasites is not unusual (64% in Gladstonecompared with 58% in the reference sites). It appears that the school of queenfish caught onthe Spoil Ground was unusual in both the percentage of fish infected with parasites and thelevel of parasitism. If high frequency and high intensity of parasitism was a continuing issuein Gladstone, Phase 2 sampling would have continued to show high levels of ectoparasiteson queenfish. This did not occur.It appears that the ecosystem in 2011 was different from periods before and after thereported fish health events.The introduction of a large number of barramundi and other fish into the system would havecaused a significant imbalance in the Gladstone Harbour in 2011. These large predatorswould have reduced the number of small fish and other prey items impacting on the foodsupply of other fish species. In addition, the lower salinity and turbidity caused by theflooding would have increased the stress on the ecosystem throughout Gladstone Harbour.The fact that the Phase 2 survey of fish health in Gladstone in 2012 did not find a continuingproblem in any of the candidate fish species supports the hypothesis that the observedissues were caused by a unique set of circumstances that occurred in 2011. This isconsistent with a stressed ecosystem resulting from the freshwater influx, turbidity caused bythe flooding and a large biomass of predators.SharksSkin redness and areas of detached scales associated with parasitic infection were observedon a number of shark species caught during the sampling program from Gladstone and thereference sites. It is unknown whether the numbers of parasites observed on sharks duringthis program are abnormal. However, it is evident that these parasites and associatedpathology are not unique to Gladstone. Severe skin redness was observed as a post-mortemeffect, and this is also not unique to Gladstone.75

Mud crabsFisheries Queensland did not observe any increase in the level of shell lesions in mud crabsin Gladstone Harbour compared with the reports in 1998 to 2001 (Andersen and Norton2001). Fisheries Queensland observed 2253 mud crabs in Gladstone in 2011/12 (Phase 1and 2 combined).SummaryIt is obvious from all the reports that something happened to aquatic life in GladstoneHarbour in 2011. The findings provided in this study indicate that the fish health in GladstoneHarbour had returned to a more normal situation in 2012. Other than the physical damage orrecovery from physical damage, this study found the fish in a health status similar to thatobserved in the Bundaberg and Fitzroy reference sites.All industrial activities in Gladstone that were operating in Gladstone in 2011 have continuedto operate, including dredging. The only factor different in 2011 compared to previous yearswas the significant rainfall, the consequent flooding and the introduction of a significantbiomass of fish from Lake Awoonga. Flooding has subsequently occurred in 2012 and 2013.However, other than continuing reports of barramundi showing physical damage as a resultof being washed over the Awoonga spillway in these flooding events, widespread reports ofsick fish have not been received by DAFF.The introduction of the large number of barramundi into the Gladstone ecosystem wouldhave caused the system to become imbalanced. There are a number of studies that showthe impact of introduced and invasive species on ecosystem, which in this situationbarramundi could be classified. Eby et al. (2006) found that stocking introduced fish resultedin predation, competition for food, and declines in animal abundance at different trophiclevels, which in turn drastically altered these ecosystems.The change in biomass of stocked fish is a gradual process commencing after the firstintroduction of hatchery reared fingerlings into new ecosystems. The ecosystem and speciescomposition change slowly over time. This was not the situation in the Boyne River in 2011where the biomass of barramundi was increased by 300 t in less than three months, havingstarted from a very low base. The changes identified by Eby et al. (2006) were compressedinto a short time frame, which did not allow the ecosystem to gradually change to a newequilibrium.The stocking of Lake Awoonga with 4.5 million fingerlings since 1996 meant that prior to the2010/11 flood, the Boyne River (including Lake Awoonga) was populated with more fish thanthe Boyne River would have naturally contained if Awoonga Dam was not present.Consequently, when these stocked fish moved downstream of Awoonga Dam, the biomassof barramundi was far in excess of the carrying capacity of the Boyne River. The populationof barramundi in Gladstone Harbour also increased to higher levels than normal, as indicatedby seven barramundi being caught while trawling for prawns. This population increasecaused stress to the whole ecosystem.The results in this study support that it was flooding combined with the introduction of a largenumber of barramundi that stressed the ecosystem in Gladstone. This study cannot rule outthe possibility that the activity of dredging and associated turbidity provided additional stressto the ecosystem, but it was not the primary stressor.76

ReferencesAdams, S.M; Brown, A.M. and Goede, R.W. 1993. A quantitative health assessment indexfor rapid assessment of fish condition in the field. Transactions of the American FisheriesSociety, 122: 63-73.Agius, C and Roberts, R.J. 1981. Effects of starvation on the melano-macrophage centres offish. Journal of Fish Biology, 19: 161-169.Agius, C and Roberts, R.J. 2003. Melanomacrophage centres and their role in fish pathology.Journal of Fish Diseases, 26: 499-509.Andersen, L. E., Norton, J. H. and Levy, N. H. 2000. A new shell disease in the mud crabScylla serrata from Port Curtis, Queensland (Australia). Diseases of Aquatic Organisms, 43:233-239.Andersen, L.E. and Norton, J.H. 2001. Port Curtis mud crab shell disease; nature,distribution and management. Centre for Environmental Management, Central QueenslandUniversity, Gladstone. 115pp.Andersen. L.E. 2003. A study into the epidemiology of mud crab (Scylla serrata) shelldisease. Thesis: Master of Applied Science in the Centre for Environmental Management,Faculty of Arts Health and Science, Central Queensland University, Gladstone, Australia.May, 2003.Anon 2011. Gladstone Area Water Board Annual Report 2011. 66pp.Anon 2012. Gladstone Area Water Board Annual Report 2012. 63pp.Dethloff, G.M and Schmitt, C.J. Condition factor and organo-somatic indices. In Schmitt, C.J. and G. M. Dethloff. (Eds). 2000. Biomonitoring of Environmental Status and Trends(BEST) Program: selected methods for monitoring chemical contaminants and their effects inaquatic ecosystems. U.S. Geological Survey, Biological Resources Division, Columbia,(MO): Information and Technology Report USGS/BRD-2000--0005. 81 pp.Deveney, M.R., Chisholm, L.A. and Whittington, I. A. 2001. First published record of thepathogenic monogenean parasite Neobenedenia melleni (Capsalidae) from Australia.Diseases of Aquatic Organisms, Vol. 46:79-82.Eby L.A., Roach; W. J. Crowder, L.B. and. Stanford, J.A. 2006. Stocking up freshwater foodwebs: effects on food web and ecosystem functioning. Trends in Ecology and Evolution,21:576-584.GenStat. 2011. GenStat for Windows, Release 14.1. VSN International Ltd, Oxford.Landos, M. 2012. Investigations of the cause of Aquatic Animal Health Problems in theGladstone Harbour and Nearshore Waters. A report commissioned by the Gladstone FishingResearch Fund. 201pp.McCullagh, P. and Nelder, J. A. 1989. Generalized Linear Models (2 nd ed.). Chapman andHall, London.77

Owens, L. 1983. Bopyrid isopod Epipenaeon ingens Nobili as a biological marker for thebanana prawn Penaeus merguiensis de Man. Australian Journal of Marine and FreshwaterResearch, 34: 477-81Owens, L. and Rothlisberg, P.C. 1985. Epidemiology of cryptonisci (Bopyridae: Isopoda) inthe Gulf of Carpentaria, Australia. Marine ecology progress series. Vol 122: 159-164Sawynok, B.; Platten, J.; Parsons, W and Sawynok, S. 2013. Gladstone RecreationalFishing Project Gladfish 2012: Assessing Trends in Recreational Fishing in GladstoneHarbour and Adjacent Waterways. 60pp.Whittington I.D. & Chisholm L.A. 2008. Diseases caused by Monogenea. In: Fish DiseasesVolume 2. Eiras J.C., Segner H., Wahlii T. & Kapoor B.G. (Eds.), pp. 683 – 816 (Chapter103). Science Publishers, Inc., New Hampshire, USA.78

Appendix A – Methods for Expanded Fish HealthSampling ProgramAnimal ethicsThe expanded fish health sampling program was conducted under the provision of AnimalEthics Approvals Ref# CA 2011/10/554 and CA 2012/04/605.Study areaThe principal study area was Gladstone Harbour and surrounding waters, as defined by theFisheries (Gladstone Harbour and Surrounding Waters) Emergency Disease and QuarantineDeclaration 2011. This area includes significant waterways, including the Narrows, GrahamCk (Curtis Island), Calliope River, Auckland Ck, South Trees Inlet, Boyne River, ColosseumInlet, Seven Mile Ck, Rodds Bay and Rodds Harbour. Sampling took place at a range of siteswithin the principal study area, and the locations of these sites are shown in Figure A-1. Thesites are the Narrows, Port Development Area, Hamilton Point (including waters east toEnfield creek), Calliope River (including Wiggins Island), Gladstone Harbour (trawl), SpoilGround, Upper Boyne River, Lower Boyne River and Rodds Bay.Figure A-1. Sampling sites (including reference sites) monitored during the program.The main reference sites (i.e. for comparison with the study area) for the scheduledmonitoring program include the nearby Fitzroy River (including Port Alma) to the north, andBundaberg with its adjacent coastal waters and estuaries to the south. Opportunistic datacollection also occurred in a variety of sites throughout the state, depending on routineactivity in the Fisheries Queensland Observer Program and other sampling programs.79

Candidate speciesThe monitoring program focussed on seven species of finfish, one species of shark, onespecies of prawn, and one species of crab. These species represent a range of different lifecycles (e.g. catadromous and estuarine) and trophic levels (e.g. predatory, omnivorousdetritivores and scavengers). They are species that have been encountered during regularsampling in the principal study area and reference areas to date (i.e. since October 2011),are caught using a range of different methods/fishing gear, and have displayed a variety ofabnormalities.Barramundi (Lates calcarifer)Barramundi is a highly predatory species and a principle target in the region for recreationalline and commercial net fishers. It is catadromous, in that fish live in fresh and marine waters,but must migrate to marine waters to spawn. Barramundi fingerlings are stocked intofreshwater impoundments throughout the Port Curtis and Fitzroy River catchments, includingAwoonga Dam.Mullet (Mugilidae)Mullet is an omnivorous detritivore with a catadromous lifecycle. The species is caughtmainly by commercial net fishers, although smaller numbers are also caught by recreationalfishers (mainly for bait) using cast nets. Sea mullet have been stocked into Lake Awoonga.Banana prawn (Fenneropenaeus merguiensis)Banana prawns are omnivorous detritivores and a principle target for the commercial trawlfishery in the region, as well as for recreational fishers using cast nets. Banana prawns usethe numerous intertidal mangrove lined creeks as nursery habitats, then move into morecoastal waters as they grow.Mud crab (Scylla serrata)Mud crab is the principle target species for recreational and commercial crabbers in theregion. The species is an active omnivorous scavenger that occurs in estuarine and coastalhabitats with mud substratum.Bull shark (Carcharhinus leucas)Bull shark is a predatory estuarine and coastal species caught frequently in the region bycommercial net fishers and recreational line fishers. The species is known to migrate intofreshwater, particularly as juveniles.Trawl species: Grinner (Saurida sp), Australian threadfin (Polydactylus multiradiatus)and Castelnau’s herring (Herklotsichthys castelnaui)These three taxa are small in size and common in the local demersal fish assemblages,which makes them common in bycatch of trawlers operating inside the study area. They arecaught occasionally by recreational anglers.Pelagic species: Queenfish (Scomberoides sp.)Queenfish is a pelagic species occurring throughout the region and commonly caught inbarramundi nets. In previous sampling it had shown some signs of redness and fish hadbeen found with a high prevalence of ectoparasites.80

Sampling regimeTimingThe sampling regime as outlined in Table1 commenced in April 2012 and was repeated inJune/July 2012 providing two separate sampling periods. During each sampling period, tenlocations were sampled (see Table 1). At each location candidate species have beenidentified for sampling. A third sampling was conducted in September 2012, focussing onbarramundi in the Boyne River, considering previously identified ectoparasite issues duringthe cooler months of 2011.Gear typeMultiple gear types were used during the sampling program depending on site and candidatespecies. The gear types were consistent with commercial fishing apparatus and included: gill net / haul net – barramundi, sea mullet, bull sharks and queenfish prawn trawl gear – banana prawn, grinner, Australian threadfin, Castelnau’s herring crab pots – mud crabs.81

Table 1: Candidate species for each sampling location.FitzroyTheNarrowsHamiltonPointPortDevelopmentAreaSampling locationsCalliope SpoilRiver groundsHarbourtrawlgroundsBundaberg(Reference(ReferenceSpeciesSite)Site)Mud crab × × × x × × ×UpperBoyneRiverLowerBoyneRiverBarramundi × × × × × × ×Bull shark × × × × × × ×Mullet × × × × × ×Rodd’sBayTrawl species –Banana prawn,Grinner, Australianthreadfin,Castelnau’s herringQueenfish× × ×× × ×(or other pelagicspecies. e.gmackerel)82

SamplingField observationsIn the field up to 100 individuals per site per trip, of each candidate species were observed.Their lengths and external health condition were recorded on field data sheets at the time ofcapture. From the observed individuals, 10 were retained for further processing. Theremaining individuals were returned to the water alive with the exception of some mud crabswhich were retained by the commercial fisher.Samples selected for further processingTen individuals per site per trip, of each candidate species (targeting five apparently healthyand five apparently unhealthy) were selected for further testing by Biosecurity Queensland.Fish/sharks were selected as unhealthy based on the presence of: lesions and skin discolouration as per Appendix 1 eye lesions as per Appendix 1 swelling or unusual mass on the abdomen signs of emaciation (i.e. large head, thin body, poor body condition) haemorrhage from gills or vent.Crabs were selected as unhealthy based on the presence of: rust spots shell erosion fouling of the shell lost appendages.Prawns were selected as unhealthy based on the presence of: presence of lesions/shell erosion parasitic infestation.Whilst 10 might seem a relatively low number, the overall level of replication is built upacross sites and times. Hence, the expectation of 10 fish/crabs/sharks/prawns per sampleper site per trip, should ultimately give adequate coverage across sexes, ages, lengths, andany other influential factors, so that these effects can then be taken into account in theoverall analysis of data.83

Sample processingSample processing involved two steps:Step 1: Observations and data recorded by field staff at the time of capture.For each species, the following data was recorded in the field on a Fish Health FieldDatasheet: site details time of capture specimen identification number on waterproof label where appropriate (see below).Mullet, barramundi, queenfish and sharks Length (forklength) was measured to the nearest 10 mm. Skin and eye condition was classified according to Appendix 1. All specimens selected for necropsy examination were allocated a specimenidentification number, photographed on both sides with close-up images captured ofabnormalities (specimen identification number label visible in all images), euthanized(blunt force trauma), individually bagged and transported on ice to the laboratory.Mud crabs Carapace width (notch to notch) was recorded to the nearest mm using calipers. Sex recorded An assessment of:o limbs missingo ‘rust’ spots number and grade as per criteria outlined in Table A-2Table A-2: Grading system used to classify shell lesions as described by Andersen (2003).GradeDescriptionGrade 1Non perforated < 5 mm diameterGrade 2Non perforated > or = 5 mm diameterGrade 3 Perforated cuticle (either partially or fully) or = 5mm diameter and < 20 mm diameterGrade 5 Perforated cuticle (either partially or fully) >or = 20 mm diameter84

o lesions present/absent (estimated % shell affected)o fouling of the shell present/absentAll specimens selected for further testing were allocated a specimen number andphotographed (specimen identification number label visible in all images).Live crabs were submitted for necropsy examination by Biosecurity Queensland.Banana prawns Carapace length (notch at base of rostrum to posterior margin of carapace) Shell lesions or erosion present/absent Visible parasites present/absent All specimens selected for further testing were allocated a specimen identificationnumber and photographed on both sides with close-up images captured ofabnormalities (specimen identification number label visible in all images). Thesespecimens were processed immediately after capture onboard the trawler by aveterinarian with aquatic animal health expertise.o Shell condition was rated and any lesions or irregularities photographed.Swabs were taken for bacteriology testing of any major lesions.o Cephalothorax/ hepatopancreas was injected with 1-2 ml 10% formalin.Prawns were cut through with a scalpel immediately posterior to thecephalothorax. Shell covering gills was removed on one side, heads fixedwhole in 50 ml 10% formalin fixative (see photo 1).o Tails were frozen for residue testing in separate containers.Photo 1: Injecting the cephalothorax of a banana prawn with 10% formalin, and removing the gill cover tooptimize fixation of tissues.85

Herring/grinner/Australian threadfin Length (forklength) was measured to the nearest 1 mm Skin and eye condition was classified according to Attachment 1 All specimens selected for further testing were allocated a specimen identificationnumber, photographed on both sides with close-up images captured of abnormalities(specimen identification number label visible in all images) and euthanized (severespinal cord behind head). These specimens were processed immediately aftercapture onboard the trawler by veterinarian with aquatic animal health expertise.o As much muscle as possible/practical was removed without rupturing the gutcavity and frozen for potential future chemical residue testing (see photo 2).This aimed to achieve of two aliquots each containing at least 10 g (wherepossible).o Operculum cover was removed on one side and the abdomen opened withscissors without rupturing any organs. The whole fish (minus muscle sample)was placed in formalin for potential histology, noting that the ratio of tissue :fixative should not exceed 1:10.Photo 2: Obtaining muscle samples for residue testing without disturbing the internal organs.86

Photo 3: Opening the abdomen of a herring to allow penetration of fixative to internal organs.Specimen identification numberWhere specimens had been selected for further processing, a specimen identificationnumber was assigned in the field. The number was written on waterproof paper andrecorded: Trip numbero e.g. T1, T2, T3 Site referenceo FIT – Fitzroyo NAR – Narrowso HAM – Hamilton Pointo CAL – Calliope Rivero SPG – Spoil groundso HTR – Harbour trawlo UPB – upper Boyne Rivero LOB – lower Boyne Rivero ROB – Rodds Bayo BUN – Bundaberg Specimen numbero 4 digits (e.g. 0001)87

For example: the third specimen, regardless of species collected from the upper Boyne onTrip 2, would be allocated a specimen number T2UPB0003. The specimen number wasrecorded on a waterproof label and remained with the specimen from capture to labprocessing.Step 2: Laboratory processingStep two involved the processing of samples that were kept for further analysis. Wherepossible, observations were quantified to allow proper comparison and analysis. All datawas recorded on a Fish Health Laboratory Datasheet by Biosecurity Queensland (Appendix2), and categories were used to develop fish health indices for crude overall comparison offish health (Appendix 3). Classification of necropsy criteria was based on Schmitt et al. 2000.Mullet, barramundi, queenfish and sharks Weight recorded Length recorded (forklength to nearest 10 mm) External photographs with close-up images captured of abnormalities Otoliths removed (for fisheries analysis, barramundi and mullet only) Analysis for skin ‘redness’ according to existing protocol (note this was done onboardwith live or freshly dead fish for post-mortem comparison) Lesion count and severity with samples fixed in formalin. If there was only one lesion,it was preserved for histology, taking a section of the lesion that spans theabnormality to normal skinEye condition, general condition, skin condition and abdomen were assessedaccording to the data sheet categoriesSwab taken from any skin lesionso A bacterial swab was inserted into a fresh scalpel cut edge of the skin lesionafter 1 ml of 70% ethanol was applied to disinfect the surface. The swabcollected blood or fluid from the wound and was replaced into its transportmedia, labelled and placed into a dry plastic zip-lock bag on ice orrefrigeration at 4 degrees celcius. These were sent to Biosecurity ScienceLaboratory in Coopers Plains for analysis within 72 hrsSkin sample taken – ensuring sample includes all layers of skin, epidermis throughdermis to start of muscleo Histolgy - ~2 cm sq in 50 ml 10% formalinGills observed for appearance, parasite load and sediment foulingo Sample taken for chemical residue testing - One side of gills, frozen. Thisconsisted of 2 aliquots each containing at least 10 g (where possible).o Sample for histology – 1 cm length of holobranch in 50 ml 10% formalin88

Open gut cavity. Photograph internal organs, noting any abnormalities Liver weighed, then divided into two samples:o Sample for chemical residue testing – frozen. This should consist of 2aliquots each containing at least 10 g (where possible)o Sample for histology - ~1 cm cubed in 50 ml 10% formalin Other tissues for histology: heart, spleen, kidney, hindgut - ~1 cm cubed fixed in100ml 10% formalin Gonad status, bile appearance, mesentery fat and parasite load were evaluatedaccording to the data sheet categories and any abnormalities noted Muscle sample taken –o Sample for residue testing – frozen. This should consist of 2 aliquots eachcontaining at least 10 g (where possible)o Sample for histology - ~1 cm cubed in 50 ml 10% formalin.Trawl Species: Banana prawns, herring/grinner/Australian threadfinThese species were processed onboard the trawl vessel according to processinginstructions outlined above (Step 1: Observations and data recorded by field staff at the timeof capture).Mud crabs Live crabs were held at 4°C for 15 minutes then transferred to -20°C until insensibleprior to necropsy Weight, carapace width and gender were recorded Photographs were taken of any shell lesions and internal organs Haemolymph was sampled for bacteriology. An aseptic sample of haemolymph wastaken from the heart using a syringe and sterile hypodermic needle through themembrane between thorax and abdomen (cleaned with alcohol first). This was thendischarged directly on marine agar with vitamin (MAV) plates for primary culture Gross evaluation with comment and scoring for:o Shell – lesions, noting ‘rust spot shell disease’ as outlined in the fieldobservations section above Gross evaluation with comment and scoring for:o Limbs – all present, one missing, more than one missingo Muscle – watery, translucent, opaque, contracted, multiple white spotso Gills – clean or fouled89

oooGonad – present/absentHepatopancreas – normal colouration or yellow/paleGut – normal/abnormalWeight of hepatopancreas recordedHistology - formalin fixing (1 cm x 1 cm in 50 ml 10% formalin):oooooolesionsgillhepatopancreasheartmusclegonad.Residue testing – Frozen: gill, hepatopancreas and claw muscle in separate containers eachcontaining at least 10 g (where possible).90

Appendix A–1: Eye and Skin lesion grading systemEye photo Description GradeThis section classifies eye condition according to gross observations.Normal lookingeyeE0Freshhaemorrhagesin or aroundnormal eye.(Damageassociated withthe captureprocess).E0aCloudy eye.NoteNeobenedeniasp. fluke(circled).E191

Cloudy eye,swollen,redness orhemorrhageE2Ruptured eyeballE392

Skin discolouration photo Description GradeThis section classifies skin based on colouration alone, even though some“discolouration” may technically represent “lesions”.No skindiscolouration(ignoring marksassociated withcapture)SK0Pale pink areasventrallySK193

Red pin pointmarks – nogeneral rednessand scalesattachedSK2Pin point areasskin/fins and/orgeneral rednessover large areaof body.SK394

Extensiveredness of theskin as a ‘rash’,may involve thefins and finbases.SK4No scalesdetached95

Lesion photos Description GradeThis section classifies lesions based on gross observations, considering location, size andseverity.No lesionsL0Individual scalesdetachedL1Adjoiningmultiple scalesdetachedL296

Shark: Adjoiningmultiple scalesdetachedL2(continued)Ulcerationthrough the skinwith scale loss,muscle damage,or bleeding.L3Thisclassificationdescribes whatis thought to beassociated withphysicaldamage. Forexample:L4jaw abrasions;97

L4continuedoperculumgrazing;cuts orlacerations98

Appendix A–2: Necropsy data sheetsNecropsy data sheet, modified from Biomonitoring of Environmental Status and Trends( BEST) program (from Goede 1989).Fish Species: Barramundi / Sea mullet / Bull shark / Threadfin / Grinner / Herring/ QueenfishAnimal ID (Fisheries):Submission number (BQ):Length:Date:Weight:Time since death:Organ/tissue Classification frozen fixed swab CommentSkinSK0; SK1; SK2; SK3; SK4 (redness)L0; L1; L2; L3; L4 (lesions)EyesE0; E0a; E1; E2; E3FinsF0; F1; F2GeneralconditionAbdomenGillsHeartLiverSpleenKidneyHindgut (0‐3)Bile (0‐3)Mesentery fat(0‐4)Parasites (0‐4)GonadMuscleGeneral Comment:Normal; emaciatedNormal; tucked; bloatedNormal; frayed; clubbed;marginate; pale; otherNormal or abnormalForm: Normal; fatty; nodules;Colour: Normal (red or light red);focal discolouration; generaldiscolouration; otherA) Black; red; otherB) Granular; nodular; enlarged;otherNormal; swollen; mottled;granular; urolithic; other.H0; H1; H2; H3(No, mild, moderate, severeinflammation)Empty/not visible; yellow;light green; dark greenNone; 50% covered;completely coveredP0 (none); P1 (single); P2 (few); P3(50). Monogenea;copepods; isopods; otherUnclear; M1 (immature); M2(gravid); F1; F2.Normal or abnormalMass:Otoliths taken:99

Appendix A–3: Fish health assessment variablesFish health assessment variables and substituted values, modified from Biomonitoring ofEnvironmental Status and Trends (BEST) program.Variable Variable condition Field SubstitutedDesignation ValueFins No active erosion F0 0Light active erosionF1 10Severe active erosion F220SpleenNormal: black, very dark red or red B 0Normal: granular, rough appearance G 0Nodular, containing fistulas or nodules D30EnlargedE30Other: aberration not fitting any above OT 30HindgutNormal, no inflammation or reddeningSlight inflammation or reddeningModerate inflammation or reddeningSevere inflammation or reddening0 01 10220330KidneyNormal: firm, dark, flatN0Swollen: enlarged or swollen S 30Mottled: gray discolourationM30Granular in appearance and texture G30Urolithiasis or nephrocalcinosisU30Other: aberration not fitting any above OT 30SkinNormal: no aberrationMild skin aberrationsModerate skin aberrationsSevere skin aberrationsExtensive redness as a rash. Scales intactSK0SK1SK2SK3SK4010203040Liver Normal: solid red or light red color A,B 0’Fatty’ liver, ’coffee with cream’ colour C 30Nodules or cysts in liver D 30Focal discolouration E 30General discolouration F 30Other: deviation not fitting any above OT30EyesGillsNo aberration, good, clea r eyesFresh haemorrhage (eg net damage)E0EOa00Opaque eye (one or bo th)Cloudy and swollen, re d or haemorrhagingE1E23030Ruptured (one or both)Normal: no apparent aberrationsE 3N300Frayed, ragged appearance F 30Clubbed, swelling of tipsC30Marginate: light discoloured margin M30Pale, very light colour P 30100

Other OT 30Parasites No observed parasites P0 0Few observed parasites P1 10Moderate parasite infestation P2 20Numerous parasites P3 30Reference: Schmitt, C. J. and G. M. Dethloff. editors. 2000. Biomonitoring of Environmental Status and Trends (BEST)Program: selected methods for monitoring chemical contaminants and their effects in aquatic ecosystems. U.S.Geological Survey, Biological Resources Division, Columbia, (MO): Information and Technology Report USGS/BRD‐2000‐‐0005.81 pp.101

Appendix B – Statistical report: Gladstone Harboursampling trips 2012OVERVIEWThis report describes the statistical analyses of the fisheries data collected on two sampling trips, fromtargeted sites around Gladstone Harbour as well as from two reference sites (Fitzroy River andBundaberg). Section 1 outlines the statistical methods employed. Section 2 covers the results for thespecies that were not submitted to the laboratory, and illustrates that there were no apparent problemsamongst these species. The comprehensive Section 3 outlines the results for the important and targetedfish species submitted to the laboratory (barramundi, bull sharks, mullet and queenfish), at both the trip /location level, and then combined into regional comparisons. Sections 4 to 6 then repeat thiscomprehensive presentation of results separately for three key species, namely barramundi, mullet andmudcrabs respectively. For bull sharks, there were insufficient numbers to conduct separate analyses. Forqueenfish, the observed percentages of diseased individuals were too low to warrant separate analyses.1. STATISTICAL METHODSAll analyses were conducted using GenStat (2011). General linear models (McCullagh and Nelder 1989)were used for the continuous variables. Generalised linear models (McCullagh and Nelder 1989) wereadopted for the discrete (categorical) data, using the Poisson distribution with a log link for counts, andthe Binomial distribution with the logit link function for binary (percentage) data.As there was incomplete coverage in the three‐way design matrix (fish by location by trip – see Table 1.1),it was important to account for this in the analyses, and to correctly adjust the resultant means. Figure 1.1demonstrates the large differences between fish species for the hepatosomatic index. This shows that any‘ raw’ averages (for each trip and location) will depend heavily on the balance of fish being caught, inparticular,the number of bull sharks.Table 1.1. Distri bution of catches by species and locations.Barra. Bull shark Grin ner Her ring Mullet Pra wn Queenfish ThreadfinFitzroy R 68 7 115 115 21 200 12 200Bundaberg 20 1 18 110 26 200 36 125Hamilton Pt 31 1 0 0 0 0 0 0Calliope R 23 3 0 0 18 0 14 0Harbour 7 0 46 120 0 174 0 220Spoil Grounds 0 0 0 0 0 0 5 0Upp. Boyne R 21 1 0 0 20 0 0 0Low Boyne R 32 10 0 0 20 0 4 0Rodds Bay 9 0 0 0 20 0 0 0Lake Awoonga 5 0 0 0 0 0 0 0102

Hepatosomatic index121086420BarramundiBull sharkMulletQueenfishFigure 1.1. Boxplot (showing the means, quartiles and ranges) of hepatosomatic index for the laboratoryassessedfish species.To avoid any apparent biases due to the varying combinations of fish caught at each location and trip, ‘fishtype’ was included as a main effect in all models, and the ‘location by trip’ tables were extracted using the‘full’ option. This effectively estimates missing (fish) values for those location/trip combinations wheneach fish type was not caught, and correctly adjusts each mean to the overall expectation.The deliberate laboratory sampling of both diseased and unaffected fish (the target was for five of each,for each location/trip combination) introduced further complexity into the analyses. In the field, each fishwas defined as ‘not diseased’ if its skin condition, lesions category and eye condition were all zero, and‘diseased’ otherwise. For the variables measured in the laboratory, any unweighted analyses for locationsand times would only reflect the proportions of diseased fish that were submitted to the laboratory, ratherthan the overall population (as best measured by all the fish obtained in the field). To correctly adjust forthis stratification, weighting factors were calculated as:Proportion (diseased or not diseased) in the whole sampleProportion (diseased or not diseased) in the laboratory subsampleThese weighting factors were determined at the location by trip by fish species level, and then used tocorrectly weight each of the respective observations (diseased or not) in the analyses.Fish size (length or weight) was trialed as an overall covariate, but showed inconsistent trends. Problemswere encountered due to the greatly varying size ranges of the species – there was evidence that this termwas merely fitting to the overall differences in size between the fish types. Adjusting all results back to acommon mean size also made little sense, given the large difference between the average lengths for bullsharks vs. the others. Hence size was omitted from all of the cross‐species analyses, leaving size effects tobe captured in with the ‘fish type’ factor. Length was included as a covariate in the separate barramundi103

analyses, but this was not done for mullet, as (within this group of fish) lengths were confounded with theactual species of mullet.‘Time since death’ was trialed as a covariate for those laboratory ratings for which there was a reasonablehypothesis that they would increase after death. This was tested at the individual species level, and for theinstances where this relationship was not significant, time since death was then dropped for the finalmodel.In addition to conducting the primary weighted analyses of location by trip by fish species (along withtheir interactions), second analyses were conducted for each variable including ‘disease status’ in themodel. This allowed the direct testing and reporting of the impact of disease status on each variable.The analyses of each of the variates (in turn) are followed by a multivariate summarisation. This adopts aprincipal components representation of all the data, taking the first two orthogonal dimensions, to obtainan overview as to which locations and times are most similar.2. RESULTS ­ TRAWL AND NET DATA (SPECIES NOT SUBMITTED TO THE LABORATORY)2.1 PrawnsOf the 574 prawns observed, there were only five with shell erosions – three from the Harbour and twofrom the Fitzroy River, with none from the Bundaberg samples. There was no statistical difference in thepercentages of prawns with erosions, with mean values (± s.e.) of 1.8% (± 1.0%) for Gladstone Harbour,1.0% (± 0.7%) for the Fitzroy River, and 0.0% (n/a) for Bundaberg.2.2 FishAll of the grinner (179), herring (345) and threadfin (545) were field‐classified as ‘not diseased’, having arating of zero for skin condition, lesion grade and eye condition. The grinners and herring also had noparasites. For threadfin, there were two found with parasites, one from Bundaberg (a reference site) andone from Gladstone Harbour. Similarly, all 71 Queenfish had ratings of zero for the skin and eye ratings,with two fish having lesions – one each from Bundaberg and the Calliope River. There were 36 queenfishwith externally‐observed parasites, with 28 from the reference sites and 8 from the Gladstone region.3. TRAWL AND NET DATA (SPECIES WHICH WERE SUBMITTED TO THE LABORATORY)Table 3.1 shows the distributions of the fish numbers that were submitted to the laboratory. In many ofthe location / trip combinations, these were the whole sample.Table 3.1. Distribution of numbers submitted to the laborato ry.Barra. Bull shark Mullet QueenfishTrip 1 Trip 2 Trip 1 Trip 2 Trip 1 Trip 2 Trip 1 Trip 2Fitzroy R 10 10 7 0 10 11 1 10Bundaberg 10 7 1 0 10 10 10 10Hamilton Pt 10 10 1 0 0 0 0 0Calliope R 10 10 3 0 7 10 0 0Spoil Grounds 0 0 0 0 0 0 5 0Upp. Boyne R 10 6 0 0 10 10 0 0Low Boyne R 10 10 0 0 10 10 0 1Rodds Bay 7 2 0 0 10 10 0 0Lake Awoonga 5 0 0 0 0 00 0Excluding net damage, there were only three fish with a recordable eye condition – one barramundi in theCalliope River, one barramundi at Bundaberg, and one sea mullet in the Fitzroy River. Notably, the lattertwo were from the reference sites. Due to this low incidence level, eye condition was not analysed.Similarly, for general condition, only one fish was classified as ‘emaciated’ – a barramundi from HamiltonPoint. All others were rated as ‘normal’.104

For heart condition, six fish were rated as ‘abnormal’, with four of these being barramundi. These sixappeared randomly spread across the trips (four vs. two, for trips 1 and 2 respectively), locations (onefrom a reference site and the other five from four different locations), and ‘disease status’ (three each).Similarly, only six fish had a hindgut rating other than ‘zero’ – three tiger mullet from the lower BoyneRiver on trip 1, and from trip 2 there were two barramundi from the Calliope River and one from theFitzroy River (r eference site). Five of these six fish were rated as ‘not diseased’.Nine fish had a kidney rating of ‘other’ (vs. ‘normal’). These were all mullet from trip 2 ‐ six coming fromthe upper Boyne River, two from the lower Boyne River, and one from Rodd’s Bay. Whilst being excludedfor the overall (cross‐species) analyses, this variable was included in the separate analyses conducted formullet.Most of the spleen categories were rated as red or black granular, which are considered normal. Therewas only one enlarged spleen, a barramundi from the lower Boyne River. There were no apparentpatterns in ratings of ‘nodular’ (four from the reference sites vs. three from the Gladstone sites), nor fort he ‘other’ category (eight from the reference sites vs. four from the Gladstone sites).Parasites (as determined in the field) had generally low incidence, with 30 from Bundaberg and 6 fromthe Fitzroy River (the reference sites) vs. a total of 11 from the seven sites around Gladstone.3.2 Comparing the laboratory and field assessmentsThe skin and lesion categories for each individual were re‐assessed in the laboratory. For skin categories,the respective counts are listed in Table 3.2. There was a significant (P < 0.01) degree of association, withperfect agreement in 63% of the cases (the light‐grey cells). There was no apparent bias, as the field ratingwas higher than the laboratory rating in 18% of the cases, and lower in 19%. This shows that the ratingsconducted in the field are unbiased, and as these were done on all 1504 fish (vs. the 294 that were alsorated in the laboratory), forms the best overall sample of skin category ratings.Table 3.2. C ounts for skin categories fo r fish (F = field assessment; L = laboratory assessment).F ‐ 0 1 2 3 4L ‐ 0 163 6 26 1 01 28 2 7 4 22 8 1 11 5 23 6 0 0 7 04 13 0 0 02As shown in Table 3.3, lesion categories tell a similar story. Again there was a significant (P < 0.01), buthigher, degree of association, with perfect agreement in 89% of the cases. The lack of bias is again evident,w ith 4.5% having a field rating higher than the rating in the laboratory, and 6.5% for vice versa.Table 3.3. C ounts for lesion categories for fish (F = field assessmen t; L = laborato ry assessment).F ‐ 0 1 2 3 4L ‐ 0 246 3 6 1 01 2 2 3 0 02 2 3 11 0 03 0 0 5 0 04 6 0 1 0 33.3 Overview of analysesOf the 13 variables that had sufficient data for analyses, ‘fish species’ was significant (P < 0.05) for 11,indicating the necessity to correctly adjust the other means for this important effect. Interestingly,‘disease status’ (as assessed in the field) only had a significant effect on four of the variables. Significantdifferences were found between locations in 11 analyses and between trips in four. Importantly, there105

was a significant (P < 0.05) location by trip interaction in seven of the 13 analyses (54%). As this isnotably higher than the random expectation of 5%, it was concluded that the changing patterns ofresponses over space and time are a real effect, so the data are reported on this two‐way basis (locationsby times).Time since death – This ranged from 0 to 17 hours. It significantly (P < 0.05) affected the laboratoryratings for skin and fin condition (% of fish affected), and there was a significant interaction with fishspecies. Of the 12 bull sharks submitted to the laboratory, most (8) had a recordable skin condition, yetfor all 12 the rated fin condition was zero. The analyses for laboratory‐rated skin and fin condition thusincluded time since death (along with the species interaction), and the resultant means are standardisedfor this.The observed patterns for the fish species are of interest. As binary (0 or 1) data cannot easily be visuallyinterpreted, we condensed these data into five levels of time since death, separately for the species (whichdid display different coverage), and also using the numbers of observations of each species in each class asa weighting factor for these analyses. For the percentage of fish with a recordable skin condition, therewas a relationship with time since death for both barramundi (R 2 = 58%; P = 0.010) and mullet (R 2 = 64%;P = 0.056; mullet had fewer observations over times). For both these species, there was no evidence of adifference in the relationship between the reference and the Gladstone sites (P > 0.17 for all four of thesignificance tests). Figures 3.1 and 3.2 show the mean values for the regions, along with the fitted lines.100Percentage with skin condition > 0806040200y = 33.8 + 3.92 * xReferenceGladstoneFitted line0 5 10 15Time since death (hours)Figure 3.1. Barramundi skin condition ratings vs. time since death, by regions.106

Percentage with skin condition > 0100806040200ReferenceGladstoneFitted liney=-30.9 + 8.55 * x0 5 10 15Time since death (hours)Figure 3.2. Mullet skin condition ratings vs. time since death, by regions.For the percentage of fish with a recordable fin condition, for barramundi there was a significant (P =0.006) difference between the regions in the intercepts. The data plus these lines are shown in Figure 3.3(overall R 2 = 70%; P = 0.016). Whilst this relationship was not significant for mullet (R 2 = 11%; P = 0.52),it is shown in Figure 3.4 for comparison.Percentage with fin condition > 050 ReferenceFitted lineGladstone40Fitted line302010y = 12.0 + 1.44 * xy=-8.1 + 1.44 * x00 5 10 15Time since death (hours)Figure 3.3. Barramundi fin condition ratings vs. time since death, by regions.107

Percentage with fin condition > 0151050ReferenceGladstoneFitted liney = 5.1 + 0.43 * x0 5 10 15Time since death (hours)Figure 3.4. Mullet fin condition ratings vs. time since death, by regions.Fish species – The adjusted means, the tests of significance between these, and their average standarderrors are listed in Table 3.4. These results (particularly the number of significant values) confirm theneed for the important ‘location by time’ means to be correctly adjusted for the numbers of fish taken ineach sample.Table 3.4. Effects of fish speci es on the response variables.Sig. level Barra. Bull shark Mullet Queenfish avg. s.e.Hepatosomatic index ** 1.1 8.9 1.4 0.9 0.1Condition score ** 1.3 1.5 1.5 1.1 0.0HAI score * 25.8 32.3 18.1 31.0 3.9% diseased (field) ** 28.7 85.2 23.7 4.3 4.8Skin % > zero (field) ** 21.5 81.6 29.3 0.0 3.9Lesions % > zero (field) ** 12.7 44.3 0.7 4.1 4.0Fins % abnormal 13.4 0.0 7.1 22.0 3.7Gills % abnormal 13.1 0.0 16.2 19.7 3.3Abdomen % tucked up ** 21.9 0.0 1.3 48.0 2.8Liver form % abnormal ** 16.1 35.9 9.7 43.0 7.1Liver colour % discoloured ** 17.0 40.1 14.3 55.0 7.2Parasites % > zero** 17.7 19.2 1.5 30.9 5.9Mesentary fat (%) ** 42.6 0.0 12.4 11.4 4.6* P < 0.05; ** P < 0.01Disease status (as assessed in the field) – The adjusted means, tests of significance and averagestandarderrors for the continuous variables follow in Table 3.5.108

Table 3.5. Effects of disease status on the response variables.Sig. level Not disea sed # Diseased avg. s.e.Hepatosomatic index 1.27 1.48 0.06Condition score 1.35 1.32 0.02HAI score ** 18.3 31.0 2.1Fins % abnormal ** 7.4 17.7 2.9Gills % abnormal 15.0 12.4 3.1Abdomen % tucked up 15.9 20.2 2.7Liver form % abnormal ** 10.9 21.1 2.7Liver colour % discoloured * 15.4 20.8 3.1Parasites % > zero 11.6 10.7 2.5Mesentary fat (%) 22.5 24.5 2.9# as assessed in the field; * P < 0.05; ** P < 0.01The tabulations and histograms of the full distributions of counts against field‐rated disease status havenot been presented, but as indicated by the general lack of significant results in Table 3.5, these were notinformative. In the analyses, the effect of disease on the resultant distributions of counts was significant (P< 0.05) only for spleen, bile and mesentary fat.3.4 Location by trip meansThese are the results that are of primary interest, and are listed in Tables 3.6a to 3.6c. For most of thecategorical measures, these tables show the binary contrast (‘normal’ vs. ‘other’), listed as percent notnormal.Mesentary fat (% cover) has been analysed as an approximately‐continuous variable. Of the fivecategories here, three are direct numbers (namely ‘none’ = 0% cover, ‘0.5’ = 50%, and ‘completelycovered’ = 100%), and the centre‐points were adopted for the remaining two categories: 25% for ‘50%’.It is notable that there were actually significant differences (P < 0.05) between the two reference sites(Fitzroy River and Bundaberg) for hepatosomatic index, condition factor, HAI, abdomen % tucked andliver colour (trip 1); and for condition factor, % diseased, % skin condition > 0, lesions % > 0, abdomen %t ucked, liver form and liver colour (trip 2).Ta ble 3.6a. Effects of locations an d trips on the response variables.LocationT ripHepat.indexConditionfa ctorHAIscore% #diseased% # skincond. > 0Fitzroy 1 0.82 1.34 11.8 7.6 7.8Bundy 1 1.74 1.46 32.8 7.9 6.6Hamilton 1 1.17 1.22 9.3 7.4 8.5Calliope 1 2.00 1.36 18.7 3.6 0.0Harbour 1 4.0 4.9Spoil 1 1.57 1.28 3.6 0.0 4.8UpBoyne 1 1.26 1.31 30.9 10.0 3.1LwBoyne 1 1.31 1.29 32.7 13.6 3.8Rodds 1 1.66 1.36 21.6 5.6 4.9LakeAwoonga 1 1.66 1.41 3.7 0.0 0.0Fitzroy 2 1.56 1.25 19.8 10.0 9.0Bundy 2 1.49 1.42 24.6 3.5 3.6Hamilton 2 1.49 1.23 19.9 12.4 13.6Calliope 2 1.79 1.39 22.2 13.4 11.6Harbour 2 7.8 7.3UpBoyne 2 1.80 1.24 38.8 7.6 5.2LwBoyne 2 2.08 1.36 28.2 10.8 10.2Rodds 2 1.57 1.35 14.9 5.7 4.4Sig. of ‐ Locations ** ** ** * *Trips ** **Interaction ** * **# as assessed in the field; ** P < 0.01; * P < 0.05109

Table 3.6b. Effects of locations and trips on the responsevariables.LesionsFins % Gills % AbdomenLocation T rip% # > 0 abnormalabnormal% tuckedFitzroy 1 0.3 16.6 9.4 0.0Bundy 1 2.3 1.8 16.4 9.8Hamilton 1 1.5 45.1 11.4 34.3Calliope 1 2.1 31.3 17.9 11.1Harbour 1 0.0Spoil 1 0.0 7.0 0.0 0.0UpBoyne 1 11.5 12.0 28.3 15.9LwBoyne 1 11.6 16.5 5.0 28.6Rodds 1 0.0 17.7 11.6 22.1LakeAwoonga 1 0.0 0.0 0.0 0.0Fitzroy 2 6.0 0.0 23.2 35.3Bundy 2 0.0 2.0 19.7 10.9Hamilton 2 3.7 0.0 0.0 33.4Calliope 2 5.0 19.8 24.4 16.8Harbour 2 3.0UpBoyne 2 12.0 12.8 12.2 39.0LwBoyne 2 4.1 11.1 4.7 13.8Rodds 2 14.1 61.9 0.0 0.0Sig. of ‐ Locations ** **TripsInteraction ** ** **# as assessed in the field; ** P < 0.01; * P < 0.05Table 3.6c. Effects of locations and trips on the response variables.Location T ripLiver form %abnormalLiver %discolo uredParasites% > zeroMesentaryfat (%)Fitzroy 1 7.2 15.3 13.5 42.7Bundy 1 12.9 37.0 2.3 42.3Hamilton 1 0.0 0.0 7.8 2.3Calliope 1 14.4 10.9 15.9 18.1Harbour 1Spoil 1 6.1 6.3 0.0 31.7UpBoyne 1 25.2 23.6 0.0 29.0LwBoyne 1 33.5 24.8 7.4 28.3Rodds 1 11.4 37.1 20.0 14.1LakeAwoonga 1 23.0 0.0 0.0 70.1Fitzroy 2 4.6 9.3 17.2 20.0Bundy 2 35.9 43.6 4.9 16.0Hamilton 2 0.0 10.0 51.6 0.0Calliope 2 0.0 0.0 27.4 1.5Harbour 2UpBoyne 2 47.7 18.7 12.3 30.3LwBoyne 2 39.5 23.6 24.0 9.4Rodds 2 0.0 0.0 25.3 38.7Sig. of ‐ Locations ** ** ** **T rips** **Interacti on**** P < 0.01; * P < 0.05110

3.5 Full categorical dataIn addition to just the binary contrasts (of ‘class zero’ vs. ‘the others’, in Table 3.6), we also looked at thedistributions of counts across all categories. However, these showed little of apparent interest, except thatmost categories were observed at most of the locations (including the reference sites). These overallcounts are summarised in Table 3.7a to 3.7f.Table 3.7a. Skin cat egories – counts by locations.SK0 SK1 SK2 SK3 SK4Fitzroy R 505 1 29 3 0Bundaberg 326 0 8 2 0Hamilton Pt 18 3 6 4 1Calliope R 47 4 5 2 0Harbour 392 1 0 0 0Spoil Grounds 5 0 0 0 0Upper Boyne R 36 0 0 6 0Lower Boyne R 47 1 15 0 3Rodds Bay 24 0 2 1 2Lake Awoonga 5 0 0 0 0Table 3.7b. Lesion categories – counts by locations.L0 L1 L2 L3 L4Fitzroy R 526 2 3 1 6Bundaberg 334 1 1 0 0Hamilton Pt 28 1 3 0 0Calliope R 52 1 4 0 1Harbour 393 0 0 0 0Spoil Grounds 5 0 0 0 0Upper Boyne R 28 3 9 0 2Lower Boyne R 41 0 24 0 1Rodds Bay 27 0 2 0 0Lake Awoonga 5 0 0 0 0Table 3.7c. Gonad c ategories – counts by locations.F1 F2 M1 M2 UnclearFitzroy R 13 3 5 2 36Bundaberg 4 22 3 6 23Hamilton Pt 4 13 0 1 3Calliope R 5 23 0 1 11Harbour 0 0 0 0 0Spoil Grounds 1 1 0 1 2Upper Boyne R 4 11 1 2 18Lower Boyne R 2 14 0 1 24Rodds Bay 8 4 3 0 14Lake Awoonga 0 0 0 0 5111

Table 3.7d. Spleen categories – counts by locations.BlackenlargedBla ckgranularBlac knodula r Othe rRedgranularRednodularFitzroy R 0 40 0 7 11 1Bundaberg 0 42 3 1 12 0Hamilton Pt 0 20 0 1 0 0Calliope R 0 27 0 3 9 1Harbour 0 0 0 0 0 0Spoil Grounds 0 5 0 0 0 0Upper Boyne R 0 26 0 0 10 0Lower Boyne R 1 25 0 0 14 1Rodds Bay 0 8 1 0 20 0Lake Awoonga 0 5 0 0 0 0Table 3.7e. Mesentary fat categories – counts by locations.None 50%CompletelycoveredFitzroy R 37 6 3 3 10Bundaberg 24 20 9 0 5Hamilton Pt 20 1 0 0 0Calliope R 29 5 5 0 1Harbour 0 0 0 0 0Spoil Grounds 3 1 1 0 0Upper Boyne R 22 1 2 0 11Lower Boyne R 29 3 0 1 8Rodds Bay 11 15 1 0 2Lake Awoonga 0 0 0 2 3Table 3.7f. Bile categories – counts by locations.DarkgreenEmpty o rnot visibleLightgreenYellowFitzroy R 12 3 26 18Bundaberg 11 13 19 15Hamilton Pt 1 0 7 13Calliope R 1 4 20 15Harbour 0 0 0 0Spoil Grounds 4 0 1 0Upper Boyne R 7 4 11 14Lower Boyne R 4 1 13 23Rodds Bay 0 0 3 26Lake Awoonga 0 5 0 03.6 Overall (multivariate patterns)Relationships amongst the continuous variables can be measured by the correlation coefficient, r. Table3.8 lists these values for the size measures and the variables in Table 3.6. Correlations between the binaryvariables are indicative only, but do show the direction and relative strength of these relationships.112

Table 3.8. Correlation matrix for the laboratory‐measured variables. Correlations are significant (P

32Bundy2Bundy1UpBoyne1UpBoyne2LwBoyne1Dimension 2 (22% variation)10-1Awoonga1Calliope1Fitzroy1Rodds1LwBoyne2Fitzroy2Calliope2-2Spoil1Rodds2Hamilton2Hamilton1-6 -5 -4 -3 -2 -1 0 1 2 3 4Dimension 1 (29% variation)Figure 3.5. Movements of the overall location and trip principal component means in multidimensionalspace. The reference sites are indicated by the squares with the dotted lines.Given that the four reference‐site points represent ‘normality’, it is pleasing to see that these scatteraround the mid‐top area of this figure, and that most of the Gladstone sites are in this proximity. Onecorner‐point here is the upper Boyne River at trip 2, which did have the highest HAI score, and the highestrates of abnormalities in the abdomen and liver‐form.Lake Awoonga does appear somewhat removed, and this is due to the ‘very good’ condition of thebarramundi (the only species sampled there) – these were all ‘undiseased’ and had low HAI scores.Similarly, the Spoil Grounds had generally healthy fish. Hamilton Point is also somewhat remote, but recallthat this sample consisted of 31 barramundi and one bull shark (hence see the separate ‘barramundi’analyses and discussion for this site).The ‘Harbour’ location is not plotted on this graph, as no samples were submitted to the laboratory fromthis site. Hence its status can only be judged by considering the percentages of diseased fish, and skin andlesion condition greater than zero. On the basis of these three variables, on trip 1 the Harbour sampleswere very similar to those from Rodds Bay. For trip 2, the Harbour samples were very similar to thesamplesfrom trip 1 at Bundaberg. Neither result appears a reason for concern.114

3.7 Summarising by regionsThe means for the individual locations can be combined into regional groupings, and also into the overallGladstone area (comprising Hamilton Point, the Calliope and Boyne Rivers, and Rodds Bay), as listed inTables 3.9a to 3.9c.Table 3.9a. Regional group means and significance‐difference testing.Hepat.ConditionHAI% % skinTrip 1 –index factor scor e diseasedco nd. > 0Fitzroy R / Bundabergb1.28a1.40a22.3 b b7.7a7.2Hamilton Pt / Calliope Ra1.58 1.29 b 14.0 b b5.5ab4.2Upper & Lower Boyne Rb1.29 1.30 b 31.8 a 11.8 3.4 bRodds Baya1.66 1.36 ab 21.6 5.6 bab4. 9Gladstone area (overall) # 1.48 1.31 22.7 8.0 4.1Trip 2 –Fitzroy R / Bundabergb1.53 1.33b22.26.7 b b6.3Hamilton Pt / Calliope Rab1.64 1.31b21.1a12.9a12.6Upper & Lower Boyne R 1.94 a 1.30a33.5ab9.27.7 bRodds Bay 1.57 ab 1.35 14.9 b 5.7 b 4.4 bGladstone area (overall) # 1.75 1.32 24.8 10.0 9.0a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).Table 3.9b. Regional group means and significance‐difference testing.L esionsF ins % Gills % AbdomenTrip 1 –#% > 0 abnormalabnormal% tuckedFitzroy R / Bundaberg 1.3 b 9.2 b 12.9 4.9 bHamilton Pt / Calliope Rb1.8a38.214.6 22.7 aUpper & Lower Boyne Ra11.6ab14.216.7 22.3 aRodds Bay 0.0 c 17.7 ab 11.6 22.1 abGladstone area (overall) 5.3 24.5 14.8 22.4Trip 2 –Fitzroy R / Bundabergb3.01.0 b a21.5a23.1Hamilton Pt / Calliope R 4.3 ab 9.9 b a12.2a25.1Upper & Lower Boyne R 8.0 a b12.08.5 ab 26.4 aRodds Bay 14.1 a 61.9 a 0.0 b 0.0 bGladstone area (overall) 7.8 21.1 8.3 20.6a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).# For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).115

Table 3.9c. Regional group means and significance‐difference testing.Trip 1 –Liver form %abnormalL iver %discolouredParasites% > zeroMesentaryfat (%)Fitzroy R / Bundabergb10.0a26.27.9 42.5 aHamilton Pt / Calliope R 7.2 b 5.4 b 11.8 10.2 cUpper & Lower Boyne R 29.4 a 24.2 a 3.7 28.7 bRodds Bayab11.4 37.1 a 20.0 14.1 bcGladstone area (overall) 16.9 19.3 10.2 18.4Trip 2 –Fitzroy R / Bundabergb20.226.4 a b11.118.0 bHamilton Pt / Calliope Rc0.05.0 bc a39.50.8 cUpper & Lower Boyne R 43.6 a b21.2 b18.119.8 bRodds Bay 0.0 c 0.0 c 25.3 ab 38.7 aGladstone area (overall) 17.4 10.5 28.1 16.0a,b,c Within columns and trips, means for the individual regions that do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).4. SEPARATE ANALYSES FOR BARRAMUNDI DATA4.1 Variables not analysedFor barramundi, there were too few ‘non‐normal’ ratings for the analyses of eye condition, generalcondition, heart condition, hindgut rating, kidney condition and spleen rating.4.2 Overview of analysesOf the thirteen continuous variables that had sufficient data for analyses, fish length was significant (P zero 20.0 14.9 3.7Mesentary fat (%) * 33.1 50.2 5.1# as assessed in the field; ** P < 0.01; * P < 0.054.3 Location by trip meansThe categorical measures have again been analysed and listed here as the binary contrast (per cent notnormal).Significant differences (P < 0.05) were found between locations in ten analyses, between trips infive, with a significant location by trip interaction in seven (54%). Hence, for most variables there areproven changing patterns over space and time, so again the data are reported on this two‐way basis, inTables 4.2a to 4.2c.116

Table 4.2a. Effects of locations an d trips on the response variables.Location T ripHepat.indexConditionfa ctorHAIscore diseasedskincond. > 0Fitzroy 1 1.05 1.36 16.5 16.6 14.4Bundy 1 1.24 1.36 22.1 21.5 22.8Hamilton 1 0.84 1.11 10.6 21.7 24.2Calliope 1 1.08 1.15 15.0 6.6 0.0Harbour 1 12.6 13.4UpBoyne 1 1.17 1.27 47.7 56.0 18.1LwBoyne 1 1.12 1.18 41.8 63.3 3.6Rodds 1 1.25 1.16 12.8 42.2 42.6LakeAwoonga 1 1.25 1.42 29.1 0.1 0.0Fitzroy 2 1.19 1.32 27.2 42.4 29.4Bundy 2 1.37 1.45 21.4 0.0 0.0Hamilton 2 1.08 1.12 20.2 44.8 47.2Calliope 2 0.98 1.30 13.1 32.9 38.8UpBoyne 2 1.43 1.32 48.5 63.0 39.4LwBoyne 2 1.18 1.17 30.0 42.1 41.6Rodds @ 2 1.06 0.93 19.5 99.9 100.0Sig. of ‐ Locations * ** ** **Trips * **Interaction * *# as assessed in the field; @ based on two (diseased) fish; ** P < 0.01; * P < 0.05Table 4.2b. Effects of locations and trips on the response variables.LesionsGills % Fins % AbdomenLocation T rip % # > 0 abnormalabnormal% tuckedFitzroy 1 0.0 0.0 19.0 0.0Bundy 1 0.0 30.0 0.0 8.9Hamilton 1 0.0 10.0 31.1 53.5Calliope 1 0.0 20.0 12.1 11.6Harbour 1 0.0UpBoyne 1 56.5 50.0 12.7 17.6LwBoyne 1 59.1 0.0 11.5 47.4Rodds 1 0.0 0.0 13.1 27.9LakeAwoonga 1 0.0 0.0 0.0 0.0Fitzroy 2 35.1 20.0 0.0 40.4Bundy 2 0.0 14.3 0.0 0.0Hamilton 2 12.6 0.0 0.0 51.4Calliope 2 9.8 0.0 0.0 23.8UpBoyne 2 61.6 16.7 28.3 41.6LwBoyne 2 19.9 0.0 9.0 17.2Rodds @ 2 100.0 0.0 100.0 0.0Sig. of ‐ Locations ** *Trips **Interaction ** ** * *# as assessed in the field; @ based on two (diseased) fish; ** P < 0.01; * P < 0.05% #% #117

Table 4.2c. Effects of locations and trips on the response variables.Location T ripLiver form %abnormalLiver %discolou redParasites% > zeroMesentaryfat (%)Fitzroy 1 0.0 16.8 0.0 94.9Bundy 1 1.9 19.5 11.1 56.4Hamilton 1 0.0 0.0 10.3 5.4Calliope 1 3.7 5.9 29.6 21.1Harbour 1UpBoyne 1 10.7 27.5 0.0 61.5LwBoyne 1 17.1 24.8 8.9 68.1Rodds 1 0.0 0.0 28.6 3.4LakeAwoonga 1 88.3 0.0 0.0 80.4Fitzroy 2 0.0 0.0 0.0 43.5Bundy 2 5.7 48.8 14.1 35.0Hamilton 2 0.0 5.1 71.2 2.5Calliope 2 0.0 0.0 39.3 7.9UpBoyne 2 93.0 48.0 0.0 73.5LwBoyne 2 8.0 8.9 40.2 20.6Rodds @ 2 0.0 0.0 49.6 3.4Sig. of ‐ Locations ** ** ** **Trips ** **Interaction *@ based on two (diseased) fish; ** P < 0.01; * P < 0.05118

4.4 Overall (multivariate patterns)The principal components spatial representation of the combined data is shown in Figure 4.1.4UpBoyne23Rodds2Dimension 2 (25% variation)210-1Hamilton2Rodds1LwBoyne2UpBoyne1LwBoyne1Fitzroy2Bundy1Awoonga1Bundy2Fitzroy1-2Hamilton1Calliope2Calliope1-6 -5 -4 -3 -2 -1 0 1 2 3Dimension 1 (38% variation)Figure 4.1. Movements of the overall location and trip principal component means for barramundi inmultidimensional space. The reference sites are indicated by the squares with the dottedlines.Most points are in the lower‐right region, reasonably close to the reference sites. The ‘slightly‐removed’group at the bottom (Rodds Bay Trip 1, and Calliope and Hamilton Trip 2) generally represent ‘good‐most abnormality quality’ fish, with low scores for HAI and rates.The most notable outlying points are –Rodd’s Bay (Trip 2) – this total sample consisted of only two fish, both of which were diseased, so thesegave ‘poor quality’ values across most of the variables.Upper Boyne (both trips) – generally had higher values for most variables, except this site did have zeroparasites.Lower Boyne, Trip 1 – had high HAI and lesions; however by trip 2 this site had ‘reverted’ back into thegeneral pack.Samples from the Harbour (Trip 1 only) were not submitted to the laboratory, hence for this locationthere are only values for field‐assessed disease, skin condition and lesion ratings. Based on these threevariables, the barramundi from the Harbour were very similar to those sampled from the Fitzroy River onTrip 1.119

4.5 Summarising by regionsThe means for the individual locations have again been combined into regional groupings, and also intothe overall Gladstone area. These are listed in Tables 4.3a to 4.3c.Table 4.3a. Regional group means and significance‐difference testing.Hepat.ConditionHAI% % skinTrip 1 –indexfactor scorediseased cond. > 0Fitzroy River / Bundaberg 1.15 a 1.36 a 19.3 b b19.118.6Hamilton Pt / Calliope River 0.96 b 1.13 b 12.8 b 14. 1 b 12.1Upper and Lower Boyne R 1.14 a 1.22 b 44.7 a 59. 6 a 10.8Rodds Bay 1.25 a 1.16 b 12.7 b 42.2 a,b 42.6Gladstone area (overall) # 1.09 1.17 25.6 37.9 17.7Trip 2 –Fitzroy River / Bundaberg 1.28 a 1.38 a 24.3 b c21.214.7 cHamilton Pt / Calliope River 1.03 b 1.21 b 16.6 b c38.9 43.0 bUpper and Lower Boyne River 1.31 a 1.24 b 39.3 a b52.540.5 Rodds Bay 1.06 a,b 0.93 c 19.5 a,b 99.9 a 99.9 aGladstone area (overall) # 1.15 1.17 26.3 56.5 53.4a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy River / Bundaberg).Table 4.3b. Regional group means and significance‐difference testing.LesionsGills % Fins % AbdomenTrip 1 –% # > 0 abno rmal abnormal % tuckedFitzroy River / Bundaberg 0.0 b 15.0 a,b 9.5b4.5Hamilton Pt / Calliope River 0.0 b b15.0 21.6a32.5Upper and Lower Boyne Rivera57.8 25.0 a 12.1a32.5Rodds Bay 0.0 b 0.0 b 13.1 27.9 a,bGladstone area (overall) 23.1 16.0 16.1 31.6Trip 2 –Fitzroy River / Bundabergc17.6 17.1 0.0 b a20.2Hamilton Pt / Calliope River 11.2 c 0.0 0.0 b a37.6Upper and Lower Boyne River 40.7 b 8.3 18.6 b 29.4 aRodds Bay 99.9 a 0.0 99.9 a 0.0 bGladstone area (overall) 40.77 3.3 27.4 26.8a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).# For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).120

Table 4.3c. Regional group means and significance‐difference testing.Trip 1 –Liver form %abnormalLiver %discolouredParasites% > zeroMesentary fat(%)Fitzroy River / Bundaberg 0.9b18.2 5.6 75.6 aHamilton Pt / Calliope River 1.9 3.0 b 19.9 13.2 bUpper and Lower Boyne River 13.9 26.1 a 4.5a64.8Rodds Bay 0.0 0.0 b 28.6 3.4 bGladstone area (overall) 6.3 11.6 15.5 31.9Trip 2 –Fitzroy River / Bundaberg 2.8 b b24.4 b7.139.3 aHamilton Pt / Calliope River 0.0 b 2.5 a,b 55. 2 a 5.2 bUpper & Lower Boyne Rivera50.5 28.4 a 20. 1 b 47.1 aRodds Bay 0.0 b 0.0 b 49.6 a,b 3.4 bGladstone area (overall) 20.2 12.4 40.1 21.6a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy River / Bundaberg).5. SEPARATE ANALYSES FOR MULLET DATA5.1 Variables not analysedFor mullet, there were too few ‘non‐normal’ ratings to allow analyses of eye condition, general condition,heart condition, hindgut rating, abdomen condition and spleen rating.5.2 Overview of analysesFish length was confounded with species of mullet, so the analyses were not adjusted for this variable.Time since death was only included for the analyses of HAI index, and the laboratory‐rated condition forskin. ‘Disease status’, as assessed in the field, showed no significant effects ( Table 5.1).Table 5.1. Effects of disease status (field‐rated) on th e response variables.Sig. level Not disea sed # Diseased avg. s.e.Hepatosomatic index 0.84 1.53 1.38 0.07Condition score 0.62 1.52 1.45 0.04HAI score 0.89 19.6 21.8 3.8Fins % abnormal 0.76 6.1 5.4 2.7Gills % abnormal 0.88 15.9 14.1 4.2Kidney % abnormal 0.60 9.5 14.7 2.6Liver form % abnormal 0.66 12.0 9.9 3.8Liver colour % discoloured 0.44 15.8 11.1 4.4Parasites % > zero 0.68 1.1 1.4 1.2Mesentary fat (%) 0.15 13.1 7.0 2.9# as assessed in the field5.3 Location by trip meansThe categorical measures have again been analysed and listed here as the binary contrasts (per cent notnormal).Significant differences were found between locations in seven of the 13 analyses, trips in five,with a significant (P < 0.05) location by trip interaction in seven (54%). Again, Tables 5.2a to 5.2c listthese means on the ‘location by trip’ basis.121

Ta ble 5.2a. Effects of locations an d trips on the response variables.LocationT ripHepat.indexConditionfa ctorHAIscore% #diseased%skincond. > 0Fitzroy 1 0.89 1.46 5.7 100.0 100.0Bundy 1 1.22 1.62 15.7 25.0 25.0Calliope 1 1.75 1.66 24.0 0.0 0.0UpBoyne 1 0.81 1.42 0.0 0.0 0.0LwBoyne 1 0.96 1.43 13.7 0.0 0.0Rodds 1 1.52 1.61 2.9 0.0 0.0Fitzroy 2 1.57 1.31 19.9 54.6 54.6Bundy 2 1.07 1.54 3.0 20.0 20.0Calliope 2 2.06 1.51 24.0 45.5 45.5UpBoyne 2 1.64 1.35 35.5 10.0 10.0LwBoyne 2 2.54 1.65 44.8 40.0 40.0Rodds 2 1.46 1.57 17.1 0.0 0.0Sig. of ‐ Locations ** ** * ** *Trips ** **Interaction ** ** ** **# as assessed in the field; ** P < 0.01; * P < 0.05n the Ta ble 5.2b. Effects of locationsand trips o response variables.LocationT rip % # > 0 abnormalabnormalabnormalLesionsGills % Fins % Kidney %Fitzroy 1 0.0 10.0 20.0 0.0Bundy 1 0.0 25.0 12.5 0.0Calliope 1 0.0 14.3 14.3 0.0UpBoyne 1 0.0 0.0 0.0 0.0LwBoyne 1 0.0 10.0 0.0 0.0Rodds 1 0.0 20.0 0.0 0.0Fitzroy 2 9.1 27.3 0.0 0.0Bundy 2 0.0 10.0 0.0 0.0Calliope 2 0.0 49.1 20.0 0.0UpBoyne 2 10.0 10.0 0.0 60.0LwBoyne 2 0.0 10.0 0.0 20.0Rodds 2 0.0 0.0 20.0 10.0Sig. of ‐ Locations **Trips **Interaction# as assessed in the field; ** P < 0.01; * P < 0.05Ta ble 5.2c. Effects of locatio ns and trips on the response variables.LocationT ripLiver form %abnormalLiver %discolou redParasites% > zeroMesentaryfat (%)Fitzroy 1 0.0 10.0 10.0 0.0Bundy 1 0.0 31.3 0.0 39.1Calliope 1 14.3 14.3 0.0 28.6UpBoyne 1 0.0 0.0 0.0 5.0LwBoyne 1 0.0 0.0 0.0 0.0Rodds 1 10.0 50.0 0.0 22.5Fitzroy 2 18.2 27.3 0.0 2.3Bundy 2 0.0 0.0 0.0 0.0Calliope 2 0.0 0.0 0.0 7.7UpBoyne 2 30.0 0.0 10.0 0.0LwBoyne 2 50.0 30.0 0.0 0.0Rodds 2 0.0 0.0 0.0 37.5Sig. of ‐ Locations **T rips* **Interaction * ** **** P < 0.01; * P < 0.05122

5.4 Overall (multivariate patterns)The principal components spatial representation is shown in Figure 5.1.3LwBoyne22Calliope1Rodds1Dimension 2 (23% variation)10-1-2Bundy1 Rodds2Calliope2LwBoyne1Bundy2UpBoyne1Fitzroy2UpBoyne2-3Fitzroy1-3 -2 -1 0 1 2 3 4Dimension 1 (29% variation)Figure 5.1. Movements of the overall location and trip principal component means for mullet inmultidimensional space. The reference sites are indicated by the squares with the dottedlines.Most points appear in the top‐left cluster, which contains both times for the Bundaberg reference site. TheFitzroy River, particularly at trip 1, is somewhat removed from the other locations. This was primarily dueto 100% of the sample being ‘diseased’ – all 10 mullet from trip 1 were rated as category 2 for skincondition. Notably, this was a reference site.The main other outlying points are the upper Boyne River at trip 2 (mainly due to the high value forkidney abnormalities – 6 out of the 10 mullet were rated as ‘other’), and the lower Boyne River also at trip2 (which had high values for the hepatosomatic index, HAI score and liver form abnormalities).123

5.5 Summarising by regionsThe means for the individual locations have been combined into regional groupings, as listed in Tables5 .3a to 5.3c.Table 5.3a. Regional group means and significance‐difference testing.Hepat.Co nditionHAI% % skinTrip 1 –indexfactorscorediseased cond. > 0Fitzroy R / Bundaberg 1.05 b b1.54 10.7a62.5 a62.5Calliope R 1.75 a 1.66 a 24.0b0.00.0 bUpper & Lower Boyne R 0.89 b 1.43 b 6.9b0.00.0 bRodds Bay 1.52 a 1.61 a 2.9 0.0 b 0.0 bGladstone area (overall) # 1.26 1.53 10.1 0.0 0.0Trip 2 –Fitzroy R / Bundaberg 1.32 b 1.42 11.5 b a37.337.3 aCalliope R 2.06 a 1.51 24.0 b a45.545.5 aUpper & Lower Boyne R 2.09 a 1.50 40.1 a 25.0 a 25.0 aRodds Bayb1.46 1.57b17.1b0.0b0.0Gladstone area (overall) # 1.93 1.52 30.3 23.9 23.9a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).Table 5.3b. Regional group means and significance‐difference testing.LesionsGills %Fins % Kidney %Trip 1 –% # > 0 abnormal abnormalabnormalFitzroy R / Bundaberg 0.0 17.5a16.30.0Calliope R 0.0 14.3b14.3 0.0Upper & Lower Boyne R 0.0 5.0 0.0 b 0.0Rodds Bay 0.0 20.0 0.0 b 0.0Gladstone area (overall) 0.0 11.1 3.6 0.0Trip 2 –Fitzroy R / Bundaberg 4.5 18.6 a,b 0.0 0.0 bCalliope R 0.0 49.1 a 20.0 0.0 bUpper & Lower Boyne R 5.0 10.0 b,c 0.0 40.0 aRodds Bay 0.0c0.0 20.0b10.0Gladstone area (overall) 2.5 17.3 10.0 22.5a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanforthe pooled reference sites (Fitzroy R / Bundaberg).124

Table 5.3c. Regional group means and significance‐difference testing.Trip 1 –Liver form %abnormalLiver %discolo uredParasites% > zeroMesentaryfat (%)Fitzroy R / Bundaberg 0.0 20.6 a 5.0 19.5 aCalliope R 14.3b14.3 0.0 28.6 aUpper & Lower Boyne R 0.0 0.0 b 0.0 2.5 bRodds Bay 10.0 50.0 a 0.0 22.5 aGladstone area (overall) 6.1 16.1 0.0 14.0Trip 2 –Fitzroy R / Bundaberg 9.1 b 13.6 a 0.0 1.1 bCalliope R 0.0 b 0.0 b 0.0 7.7 bUpper & Lower Boyne R 40.0 a 15.0 a 5.0 0.0 bRodds Bay 0.0 b 0.0 b 0.0 37.5 aGladstone area (overall) 20.0 7.50 2.50 11.3a,b,c Within columns and trips, means for the individual regions which do not have a common superscriptletter are significantly different (P < 0.05).#For the overall ‘Gladstone area’ means, those which are bolded are significantly different from the meanfor the pooled reference sites (Fitzroy R / Bundaberg).6. SEPARATE ANALYSES FOR MUDCRAB DATA6.1 OverviewShell condition (as measured at capture) had 1599 observations of which only 57 (3.6%) were classifiedas abnormal. These were then further classified according to their lesion grades. The sub‐samples (140individual crabs) were submitted to laboratory study for re‐checking the lesion gradings as well asmeasuring overall size (carapace width and weight) and hepatopancreas weight. Again, analyses of thelaboratory sample data were appropriately weighted to accurately reflect the observed damagedproportions in the overall sample.Crab class (defined as small females, large females, undersized males, and retained males) was included inall analyses, as this effect was always significant (P < 0.05). Carapace width and total weight were trialledseparately as a covariate for percentage damage, under the hypothesis of a positive relationship(increasing amounts of damage with increasing size). However, these relationship were nowhere nearsignificant (P > 0.20), and actually fitted as slightly negative, possibly due to ‘crab class’ capturing themajority of any effect here. Hence ‘crab class’ was retained in the final model but size was omitted.6.2 Comparing the laboratory and field assessmentsThe comparison between the lesion grade ratings showed a significant (P < 0.01) degree of association,with the counts listed in Table 6.1. There was agreement in 72% of these cases. Interestingly, the fieldassessment was higher than the laboratory assessment in 23% of these cases, whereas the reverse onlyoccurred in 5% of the crabs.Table 6.1. Highest lesion grade for each crab (Field = field assessment; Lab. = laboratory assessment).Field ‐ 0 1 2 3 4 5Lab. ‐ 0 88 2 7 6 7 01 3 1 2 0 2 12 1 1 2 0 1 13 0 0 0 1 0 04 0 0 1 0 1 25 0 0 0 0 1 5125

6.3 ResultsThere was no significant difference (P = 0.68) between the hepatosomatic index of normal and abnormalcrabs (0.057 ± 0.002 vs. 0.054 ± 0.002). Missing appendages (any vs. none) appeared to be randomoccurrences, with an overall mean rate of 16.9%, and not related to any of the design factors.The key variables analysed were the percentage abnormal (field‐assessed using all 1599 observations; aBinomial proportions model), the distribution of counts according to highest lesion class (a Poisson linearmodel), and hepatosomatic index (Normal model). In these analyses ‘crab category’ was always significant( P < 0.05), justifying its inclusion as a stratifying term, with the means shown in Table 6.2.Table 6.2. Mea n values and standard errors (s.e.) for the crab categories.Hepatosomatic index s.e. Percent abnorm al s.e.Female large 0.059 0.002 3.8 0.8Female small 0.064 0.004 1.8 0.7Male retained 0.043 0.003 7.1 1.7Male undersize 0.056 0.002 2.5 0.7Regarding locations and trips, the only significant (P < 0.01) effect was for the hepatosomatic index, wheretrip 2 (overall, 0.061 ± 0.002) was higher than trip 1 (0.050 ± 0.002). Despite the interaction betweenlocation and trip being non‐significant (P > 0.66) for all variables, it is the relationships amongst thesevalues that are of prime interest, and the two‐way means are presented here for consistency with theresults for the fish species. Table 6.3 lists these means, and the two‐dimensional spatial representation isshown in Figure 6.1. No principal components analysis was needed to summarise these results, as thereare only these two key continuous variates.Table 6.3. Mean values and average standard errors for locations and trips.Hepatosomatic index ‐ Percent abno rmal ‐Trip 1 Trip 2 Trip 1 Trip 2Fitzroy River 0.046 0.060 4.5 1.4Bundaberg 0.050 0.061 3.7 3.1The Narrows 0.055 0.059 3.2 5.2Port 0.057 0.068 8.8 4.4Hamilton Point 0.050 0.061 3.1 3.7Calliope River 0.049 0.063 5.8 2.7Rodds Bay 0.043 0.056 2.3 2.4(average s.e.) 0.004 (average s.e.) 2.0126

0.07Port 2Hepatosomatic index0.060.05Calliope R 2Bundaberg 2Hamilton Pt 2Fitzroy R 2Rodds Bay 2Narrows 1Hamilton Pt 1Bundaberg 1Narrows 2Calliope R 1Port 1Fitzroy R 1Rodds Bay 12 3 4 5 6 7 8 9Abnormal %Figure 6.1. Movements of the mudcrab location means over trips. The reference sites are indicated by thesquares with the dotted lines.For all sites, the hepatosomatic index was higher for trip 2. Trip 2 also sees similar or lower percentagesof abnormals than trip 1, for all sites except for the Narrows. Except for Port, all sites are in the samegeneral spatial vicinity (and hence not notably different from) the two reference sites, at each trip.Possibly due to the low number of abnormal crabs (57), the distributions of these counts across thehighest observed lesion class also showed no significant effects for location or trip (P = 0.14 and 0.13respectively). Again, all categories were observed both at the reference sites and across the Gladstoneregion. These counts are listed in Table 6.4 for completeness.Table 6.4. Lesion ca tegories – counts by locations.L1 L2 L3 L4 L5Fitzroy River 2 6 1 1 4Bundaberg 0 1 1 2 2The Narrows 1 0 2 2 1Port 2 5 1 1 1Hamilton Point 0 2 1 2 2Calliope River 1 1 2 1 0Rodds Bay 1 3 0 4 1127

Table 6.5 lists the mean values when combining the locations into their respective regions. The referencesites are Bundaberg and the Fitzroy River, and Gladstone includes the locations of the Narrows, Port,Hamilton Point, the Calliope River and Rodds Bay. There were no significant differences between thereference sites and Gladstone region, across both variables and both trips.T able 6.5. Mean values and average standard errors for the pooled regions.Hepatosomatic index ‐Percent abnormal ‐Trip 1 Trip 2 Trip 1 Trip 2Reference sites 0.048 0.060 4.1 2.3Gladstone region 0.051 0.061 4.6 3.7(average s.e.) 0.002 (average s.e.) 1.0ReferencesGenStat (2011). GenStat for Windows, Release 14.1. VSN International Ltd., Oxford.McCullagh, P. and Nelder, J. A. (1989). Generalized Linear Models (2 nd ed.). Chapman and Hall, London.128

Appendix C – Statistical report: Histology results1. METHODSThere were 45 barramundi, 10 bull sharks, 25 grinner and 40 mud crabs submitted for histologicalanalyses . As each species had a different sampling scheme over locations, these were analysed separately.The statistical methods have been previously detailed; in summary: General linear models (GLMs; McCullagh and Nelder 1989) were adopted for the analyses, usingGenStat (2011). Each fish or crab was taken as an independent experimental unit. Weighting factors were used to adjust for the stratified sampling of both diseased and unaffectedanimals. As the grinner were all rated as ‘normal’ (not diseased), no weightings were adopted fortheir analyses. Length was generally significant for barramundi, bull sharks and grinner, so was retained in allanalyses to adjust for size (length being approximately the same as age). The primary variable of importance is cumulative score. Analyses are also conducted for eachindividual component of this score. These were all considered as continuous variables, as thesignificance‐difference testing and interpretation are more understandable on this basis. Liver lipid (fish) and hepatopancreas lipid (mud crabs) are categorical variables, so wereanalysed as counts in these classes under a Poisson GLM with a log‐link. The means, and significant differences between these, are primarily reported at the location bytrip level. These values are also then compared at the regional level. The pooled standard error wascalculated according to the usual sum‐of‐variances method (Raj 1968), and these were used toconduct a t‐test for the degree of difference between these regional means. Bull sharks (with fewer observations) were only analysed and reported at the regional level. Secondary analyses were also conducted, including ‘disease status’ in the model (except forgrinner, where all fish were rated as undiseased).2. RESULTS – BARRAMUNDISkin lesions were missing for 14 of the 45 observations, unfortunately including all five from Bundabergon trip 1. Muscle lesions were missing for two observations. This also results in missing values for thecumulative score for these fish. As shown in Figure 2.1, the distribution of the residuals justifies theanalyses using the assumptions of untransformed continuous data and the Normal distribution.129

Standardized residualsStandardized residualsStandardized residualsCumulative_score102816420-1-20-2-1012345678910Standardized residualsFitted values22.010-1-21.51.00.50.0-2-1 0 120.00.5 1.0 1.52.0Normal plotHalf-Normal plotFigure 2.1. Distr ibution of residuals for the analysis of the barramundi cumulative score data.Disease status – None of the ratings in Table 2.1 showed a significant (P < 0.05) difference betweennormal and diseased fish. Whilst cumulative score was close, note that the diseased fish had lower averages cores here.Table 2.1. Effects of disease stat us on the his tology ratingsfor barramundi.NormalDiseasedAverage s.e. Sig. level (P)Cumulative score 7.99 7.04 0.42 0.07Gill hyperplasia 2.39 2.37 0.14 0.49Gill parasites 2.00 1.93 0.22 0.69Liver MMCs 1.76 1.71 0.07 0.46Skin lesions 0.83 0.63 0.23 0.42Muscle lesions 0.35 0.17 0.11 0.17There was no significant effect (P = 0.20) of disease rating on the classification of liver lipid. Table 2.2showsthat the counts are well distributed across the liver lipid categories.130

Table 2.2. Counts forliverlipid categ oriesby disease status.AbsentMild Moderate MarkedNormal 4 13 8 3Diseased 1 5 5 6Location by trip meansThese are the results that are of primary interest, and are listed in Tables 2.3a to 2.3f. For each rating,significant‐difference testing has been conducted amongst these means (locations by trips).Table 2.3a. Cumulative score.Lake FitzroyBundaberg HamiltonUpperAwoongaRive rPoint Boyne R. Avg. s.e.Trip 1 3.91 c 8.47 ab 5.66 bc 7.13 ab 0.91Trip 2 7.61 ab 6.38 bc 9.36 a 7.02 aba Means with a common superscript are not significantly different (P = 0.05).Table 2.3b. Gill hyperplasia.Lake F itzroy Bundaberg HamiltonUpperAwoongaRiverPoint Boyne R. Avg. s.e.Trip 1 1.14 b 2.64 a 2.49 a 1.34 b 2.64 a 0.27Trip 2 2.67 a 2.13 a 2.75 a 2.85 aTable 2.3c. Gill parasites.Lake FitzroyBundaberg Hamilton UpperAwoongaRiverPoint Boyne R. Avg. s.e.Trip 1 1.17 c 2.38 ab 1.68 abc 1 .81 abc 1.34 bc 0.43Trip 2 2.04 abc 2.02 abc 2.75 a 2.34 abcTable 2.3d. Liver MMCs.LakeFitzroyBundaberg HamiltonUpperA woongaRive rPoint Boyne R. Av g.s.e.Trip 1 1.76 abc 1.92 ab 1.40 c 2.04 a 1.71 abc 0.13Trip 2 1.58 bc 1.45 c 2.07 a 1.57 bcTable 2.3e. Skin lesions.Lake F itzroy Bundaberg HamiltonUpperAwoongaRiverPoin t Boyne R. Avg. s.e.Trip 1 0.00 b 1.25 a 0.24 ab 0.98 ab 0.48Trip 2 1.05 a 0.34 ab 0.84 ab 0.06 ab131

Table 2.3f. Muscle lesions.Lake FitzroyBundaberg Hamilton UpperAwoongaRive rPoint Boyne R. Avg. s.e.Trip 1 0.00 c 0.13 bc 0.54 ab 0 .31 abc 0.05 bc 0.22Trip 2 0.13 bc 0.30 abc 0.95 a 0.12 bcThere was no significant effect of location, trip or their interaction on the classification of liver lipid. Table2.4 shows that the counts by locations are well distributed across the liver lipid categories.Table 2.4. Counts for liver lipid categories by locations.LakeFitzroyBundaberg Hamilton UpperAwoonga RiverPoint Boyne R.Absent 0 0 0 2 3Mild 0 4 8 4 2Moderate 5 2 0 4 2Marked 0 4 2 0 3Summarising by regionsThe means for the locations can be combined into ‘Gladstone’ (Hamilton Point and Upper Boyne River;notably does not include Lake Awoonga) vs. ‘Reference’ (Bundaberg and Fitzroy River), as listed in Table2.5.Table 2.5. Regiona l m eans for the histology ratings. Within trips, significant (P < 0.05) differences arebolded.Trip 1 – T rip 2 –ReferenceGladstoneSig. levelReferenceGladstoneSig.levelCumulative score 8.47 6.39 0.08 6.99 8.19 0.16Gill hyperplasia 2.56 1.99 0.04 2.40 2.80 0.15Gill parasites 2.03 1.57 0.28 2.03 2.55 0.24Liver MMCs 1.66 1.87 0.12 1.52 1.82 0.03Skin lesions 1.25 0.61 0.30 0.69 0.45 0.58Muscle lesions 0.34 0.18 0.48 0.22 0.54 0.163. RESULTS – BULL SHARKSAll 10 sharks were from trip 1, so no trip contrast exists. The single shark from Bundaberg and the fivefrom the Fitzroy River were pooled as the sample for the reference region, and similarly the Gladstoneregion sample consists of the single one from Hamilton Point with the three from the Calliope River.Tables 3.1 and 3.2 list the histology ratings for disease status and regions, respectively. These contrastswereall non‐significant (P > 0.05).132

Table 3.1. Effects of diseas e status on the histology ratingsfor bull sharks.NormalDiseasedAv erage s.e. Sig. level (P)Cumulative score 1.45 0.76 0.54 0.43Gill hyperplasia 1.08 0.23 0.38 0 .19Gill parasites 0.05 0.11 0.19 0.83Liver MMCs 0.00 0.00 NA NASkin lesions 0.16 0.33 0.24 0.65Muscle lesions 0.00 0.13 0.17 0.64Table 3.2. Effects of regi ons on the histology ratings for bull sharks.GladstoneReferenceAverage s.e. Sig. level (P)Cumulative score 0.46 1.19 0.45 0.30Gill hyperplasia 0.00 0.70 0.31 0.14Gill parasites 0.23 0.01 0.15 0.35Liver MMCs 0.00 0.00 NA NASkin lesions 0.40 0.23 0.20 0.58Muscle lesions 0.00 0.17 0.15 0.45For liver lipid, all ten bull sharks were rated as ‘marked’.4. RESULTS – GRINNERLocation by trip meansThese results are listed in Tables 4.1a to 4.1f. These proved to have higher variability, so few significantdifferences were found. Significant‐difference testing is only indicated in Table 4.1c (gill parasites), wheresome differences were found. In the other tables, the lack of superscripts indicates that the means are notproven different. As the locations here are similar to regions, no follow‐up regional comparisons havebeen conducted.Table 4. 1a. Cumulative score.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.60 0.74 0.45Trip 2 1.02 0.89 1.87Table 4. 1b. Gill hyperplasia.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.00 0.38 0.28Trip 2 0.41 0.43 0.18Table 4. 1c. Gill parasites.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.60 ab 0.21 b 0.31Trip 2 0.59 ab 0.00 b 1.21 aa Means with a common superscript are not significantly different (P = 0.05).Table 4. 1d. Liver MMCs.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.00 0.00 0.15Trip 2 0.03 0.12 0.25Table 4. 1e. Skin le sions.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.00 0.02 0.22Trip 2 0.00 0.37 0.22133

Table 4. 1f. Muscle lesions.Fitzroy River BundabergGladstone Harbour A vg.s.e.Trip 1 0.00 0.20 0.10Trip 2 0.00 0.00 0.00For grinners, there was no significant effect of location (P = 0.58) or trip (P = 0.64) on the classification ofliver lipid. Table 4.2 shows that the counts by locations are well distributed across the liver lipidc ategories.Table 4.2. Counts for liver lipid categories b y locations.Fitzroy River Bund aberg Gladstone HarbourAbsent 3 2 2Mild 4 3 4Moderate 2 0 4Marked 1 0 05. RESULTS – MUD CRABSGill hyperplasia and gill parasites were measured for all 40 mud crabs. However, the data forhepatopancreas, hepatopancreas lipid, heart, cuticle lesions and muscle lesions all contained missingobservations. This is unfortunate, as the cumulative score can only be calculated for crabs which have allscores measured, and there are only 19 observations here.Carapace width, gender and crab category were all trialed as covariates. From 21 analyses only two weresignificant (P < 0.05), approximately matching the random expectation of 1.05, so no covariate was usedfor the final analyses. Again, the distributions of the residuals justified the statistical assumption ofuntransformed continuous data under a Normal distribution.Disease status – As listed in Table 5.1, only cuticle lesions showed a significant (P < 0.05) differencebetween abnormal and normal crabs.Table 5.1. Effects of disease status on the histologyratings for m ud crabs.Abnormal NormalAverages.e. Sig. level (P)Cumulative score 2.33 1.95 0.63 0.67Gill hyperplasia 0.64 0.70 0.16 0.77Gill parasites 0.78 0.68 0.15 0.66Hepatopancreas 0.13 0.00 0.08 0.27Heart 0.13 0.13 0.11 1.00Cuticle lesions 0.41 0.00 0.11 0.01Muscle lesions 0.27 0.25 0.17 0.92There was no significant effect (P = 0.40) of disease rating on the classification of hepatopancreas lipid.Table 5.2 shows that the counts are well distributed across the hepatopancreas lipid categories.Table 5.2. Counts for he patopancreas lipidcategories by disease status.AbsentM ild ModerateAbnormal 2 9 4Normal 1 9 9Location by trip meansThese results are listed in Tables 5.3a to 5.3g. Again, significant‐difference testing has only been indicatedon the four tables that had some proven differences amongst the means.Table 5. 3a. Cumulativescore.Bundaberg Fitzroy RiverThe NarrowsPort AreaAvg . s.e.Trip 1 3.43 a 2.94 a 0.81 b 2.00 ab 1.29Trip 2 3.00 a 0.11 b 0.75 b 3.00 aa Means with a common superscript are not significantly different (P = 0.05).134

Table 5. 3b. Gill hy perplasia.Bundaberg FitzroyRiverThe NarrowsP ort AreaAvg . s.e.Trip 1 1.00 ab 0.01 d 0.30 bcd 0 .83 abc 0.26Trip 2 1.00 ab 0.32 bcd 0.68 abcd 1.24 aTable 5. 3c. Gill parasites.Bundaberg Fitzroy RiverThe NarrowsPort AreaAvg . s.e.Trip 1 1.30 a 1 .02 ab 0.31 b 0.69 ab 0.27Trip 2 0.57 ab 0.34 b 0.50 ab 0.72 abTable 5. 3d. Hepatopancreas.Bundaberg Fitzroy River The NarrowsPort AreaAvg . s.e.Trip 1 0.00 0.00 0.08 0.00 0.08Trip 2 0.00 0.00 0.00 0.00Table 5. 3e. Heart.Bundaberg Fitzroy River The NarrowsPort AreaAvg . s.e.Trip 1 0.00 0.00 0.08 0.00 0.21Trip 2 0.00 0.33 0.00 0.46 0.00Table 5. 3f. Cuticle lesions.Bundaberg Fitzroy RiverThe NarrowsPort AreaA vg.s.e.Trip 1 0.025 0.020 0.000 0.000 0.083Trip 2 0.058 0.055 0.000 0.000Table 5. 3g. Mu scle l esions.Bundaberg Fitzroy RiverThe NarrowsPort AreaAvg . s.e.Trip 1 0.00 c 1.85 a 0.04 c 0.00 c 0.25Trip 2 0.00 c 0.95 b 0.00 c 0.00 cThere was no significant effect (P = 0.18) of location on the classification of hepatopancreas lipid. Table5.4 shows that the counts by locations are well distributed across the lipid categories.T able 5.4. Counts for hepatop ancreas lipid categories by locations.Bundaberg Fitzroy RiverThe Narrows Port AreaAbsent 0 0 1 2Mild 6 3 6 3Moderate 3 6 2 2As shown in Table 5.5, there was a significant effect (P = 0.006) of trip on the classification ofhepatopancreas lipid, with a shift of numbers from moderate to mild. These patterns are more fullyexplored below, at the regional level.T able 5.5. Counts for hepatopa ncreaslipid categories by trips.Trip 1 Trip 2Absent 3 0Mild 5 13Moderate 10 3Summarising by regionsThe means for the locations can be combined into ‘Gladstone’ (the Narrows plus the Port area) vs.‘ Reference’ (Bundaberg and Fitzroy River), as listed in Table 5.6. The three significant differences all occurontrip 1, and interestingly all show higher values from the reference sites.135

Table 5.6. Regiona l m eans for the histology ratings. Within trips, significant (P < 0.05) differences arebolded.Trip 1 – T rip 2 –ReferenceGladstoneSig. levelReferenceGladstoneSig.levelCumulative score 3.18 1.40 0.015 1.55 1.88 0.887Gill hyperplasia 0.51 0.56 0.819 0.66 0.96 0.287Gill parasites 1.16 0.50 0.015 0.46 0.61 0.597Hepatopancreas 0.00 0.04 0.596 0.00 0.00 1.000Heart 0.00 0.04 0.792 0.16 0.23 0.840Cuticle lesions 0.02 0.00 0.778 0.06 0.00 0.523Muscle lesions 0.92 0.02 0.000 0.47 0.00 0.105Considering the hepatopancreas lipid categories over trips separately for each region, Table 5.7 showsthere was no significant change (P = 0.13) for the Gladstone region, with only some interchange in thenumbers between the absent and mild categories. Overall, only 25% were rated as moderate forhepatopancreas lipid.T able 5.7. Gladsto ne region ‐ counts for hepatopancreas lipid categories by trips.Trip 1 Trip 2Absent 3 0Mild 3 6Moderate 2 2Table 5.8 shows there was a significant change (P = 0.017) between the trips for the reference region. Intrip 1, eight out of ten crabs were rated as moderate, but this dropped to one from eight for trip 2. Overall,50% of the mudcrabs from the reference region were rated as moderate for hepatopancreas lipid.T able 5.8. Referen ce region ‐ counts for hepatopancreas lipid categories by trips.Trip 1 Trip 2Absent 0 0Mild 2 7Moderate 8 1ReferencesGenStat (2011). GenStat for Windows, Release 14.1. VSN International Ltd., Oxford.McCullagh, P. and Nelder, J. A. (1989). Generalized Linear Models (2 nd ed.). Chapman and Hall, London.Raj, D. (1968). Sampling theory. McGraw‐Hill, New York.136

Appendix D – Statistical analysis chemical residues137

Contaminants in Tissue ofFish and Crabs Collected inthe Gladstone AreaExpanded Gladstone Harbour Fish Health SurveyWater Quality and Investigations, Environmental Monitoring and Assessment Sciences,Science Delivery, DSITIA.29 April, 2013

Department of Science, Information Technology, Innovation and the ArtsPrepared bySuzanne Vardy, David Mayer and Michael Warne.Water Qualityand Investigations, Environmental Monitoring and Assessment SciencesScience Delivery DivisionDepartment of Science, Information Technology, Innovation and theArtsPO Box 5078Brisbane QLD 4001© The State of Queensland(Department of Science, Information Technology, Innovation and the Arts) 2013The Queensland Government supports and encourages the dissemination and exchange of itss information. Thecopyright in this publicationis licensed under a Creative Commons Attribution 3.00 Australia (CCC BY) licenceUnder this licence you are free, without having to seekk permission from DSITIA, to t use this publication in accordancewith the licence terms.You must keep intact the copyright notice and attribute the State ofQueensland, Department of Science, InformationTechnology, Innovation and the Arts as the source of the publication.For more information on this licence visit http://creativecommons.org/licenses/by//3.0/au/deed.enDisclaimerThis document has been prepared with all due diligence and care, based on the best availablee information at the time ofpublication. The department holds no responsibility for any errors or omissions within w this document. Any decisions madeby other parties based on this documentt are solely thee responsibility of those parties. Information contained in thisdocument is from a number of sources and, as such, does not necessarily represent government or departmental policy.If you need toaccess this document in a language other than English, please call the Translating and InterpretingService (TIS National) on 131 450 and ask them to telephone Library Services onn +61 7 3170 5725CitationVardy S, Mayer D and Warne M.St.J. (2013) Contaminants in tissueof fish and crabs c collectedd in the Gladstone area.Department of Science, Information Technology, Innovation and theArts. Brisbane, Queensland.Acknowledgements DSITIA would like to acknowledge the staff from Fisheries Queensland who undertookk thesampling andstaff from Biosecurity Queensland who prepared the samples for analysis.This report has been prepared by the Department of Science, Information Technology, Innovation and the Arts.May 2013

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaExecutive SummaryIn August 2011, the Queensland Government received reports of visibly sick barramundi occurringin the Boyne River, Gladstone. Subsequently a variety of other aquatic species were allegedlyaffected by red skin and other conditions in the Gladstone region. In January 2012 the GladstoneFish Health Scientific Advisory Panel recommended that the Queensland Government enhance theexisting investigation in order to understand the causes of fish health issues in Gladstone Harbour.Based on this recommendation the Gladstone Integrated Aquatic Investigation Program wasdeveloped by the Queensland Government and commenced in February 2012.This report outlines the results of tissue residue testing undertaken as part of the GladstoneIntegrated Aquatic Investigation Program (specifically Phase 2 of the Gladstone Fish HealthInvestigation). In order to assess any potential association between contaminants and disease, fishand crab samples were collected between late April and early May 2012, and mid-June and earlyJuly 2012. Analysis for 13 metals and metalloids was undertaken in mud crab (Scylla serrata)hepatopancreas, barramundi (Lates calcarifer) gills and liver, and grinner (Saurida sp.) muscle. Inaddition, a suite of 17 polynuclear aromatic hydrocarbons (PAHs), 57 pesticides, 8 miscellaneousorganic chemicals (including synthetic fragrances) and 32 polychlorinated biphenyls (PCBs) wereanalysed in mud crab hepatopancreas. All results and conclusions in the current study apply onlyto the period when samples were collected.Comparison of metal and metalloid concentrations in barramundi with visible signs of disease(diseased) with visibly normal barramundi (healthy) showed no significant difference in gill or livertissue. A similar analysis of diseased crab hepatopancreas found significantly higher aluminiumand selenium levels in crabs with visible lesions. These two metals were not at elevatedconcentrations in mud crabs from Gladstone sites compared to reference sites. All grinner wereassessed as being healthy.Comparisons with historical data revealed that mud crab hepatopancreas concentrations of metalsand metalloids in the Gladstone area were similar to those found other estuaries and near-coastalsites along the east coast of Queensland. No organic contaminants were detected in mud crabhepatopancreas in Gladstone Harbour.Statistical analysis of metal and metalloid concentrations in barramundi livers from Gladstone sitescompared to reference sites showed significantly higher arsenic, cadmium, iron and zincconcentrations in the Gladstone area from one sampling trip for each metal, indicating possibleelevated exposure to these metals and metalloids. However, as the vast majority of samples hadarsenic, cadmium and zinc concentrations that were below the concentrations associated withtoxicological effects (no data were available for iron) and as there was a lack of statisticalsignificance between fish disease and metal concentration, it is unlikely that metals and metalloidsare associated with fish disease in Gladstone Harbour.Grinner tissue from Bundaberg (a reference site) had significantly higher aluminium, iron, copperand zinc concentrations compared to other sites, but there were no other significant trends.Based on the samples collected during Phase 2 of the fish health investigation, there was nostrong evidence of a link between fish health and tissue residue concentrations. There was asignificant association with crab disease and elevated levels of selenium and aluminium in both thereference sites and Gladstone sites.i

Department of Science, Information Technology, Innovation and the ArtsContentsExecutivee Summary................................... .................. .................. .................. ................................... iList of tables .................................................................................................................................... iiii1. Introduction ........................................................................................................................... 12. Sampling Sites ...................................................................................................................... 33. Methods ................................................................................................................................. 53.1.3.2.3.3.3.4.3.5.Sample CollectionSample PreparationLaboratory Analysis.Statistical AnalysisToxicological Effects Assessment566774. Is there a link between the concentration of metals and metalloidss and organiccontaminants in tissue and animal health? .................................................................................. 84.1.4.2.4.3.4.4.Mud CrabBarramundiGrinnerSummary81011115. Arethe concentrations of contaminants in crab and fishh tissue from Gladstonesimilar to those fromother Queensland estuarine and near-coastal locations? .................... 125.1.5.2.5.3.5.4.Mud CrabBarramundiGrinnerSummary122124246. Arethe levels of contaminants in fish and crab tissue known k to bee associated withreported ecotoxicological effects? .............................................................................................. 256.1.6.2.6.3.6.4.IntroductionMetal Residues in the Hepatopancreas of Mud CrabsMetal Residues in the Gills andd Liver of BarramunddiSummary252525297. Conclusions ......................................................................................................................... 308. References ........................................................................................................................... 31Appendix A ..................................................................................................................................... 34ii

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaAppendix B ..................................................................................................................................... 40Appendix C ..................................................................................................................................... 52List of tablesTable 1 – Sites where species used for tissue residue analysis were collected ......................................... 3Table 2 – Number of samples from each site used for tissue residue analysis .......................................... 5Table 3 – The limit of reporting for each metal and metalloid (on a wet weight basis). .............................. 6Table 4 – Disease status and mean concentrations of metals and metalloids (mg/kg) in the mudcrab hepatopancreas (dry-weight basis). Metals and metalloids that have significantlydifferent concentrations (P≤0.05) between healthy and diseased crabs are bolded andshaded in light blue. ...................................................................................................................... 9Table 5 – Aluminium concentrations (mg/kg) in the hepatopancreas of diseased and healthy crabsby site (dry-weight basis, n=5). Sites that have a significant difference (P≤0.05) betweenhealthy and diseased crabs are bolded and shaded in light blue. ................................................ 9Table 6 – Selenium concentrations (mg/kg) in the hepatopancreas of diseased and healthy crabsby site (dry-weight basis, n=5). ................................................................................................... 10Table 7 – Disease status and gill metal and metalloid concentrations (mg/kg dry-weight basis). Nosignificant difference (P>0.05) between healthy and diseased fish were found. Measuredconcentrations of chromium, nickel, silver and lead were generally below the limit ofreporting and were not included in this analysis. ........................................................................ 10Table 8 Disease status and mean measured concentrations of metal contaminants in liver (mg/kgdry-weight basis). No significant difference (P>0.05) between healthy and diseased fishwere found. Measured concentrations of chromium, nickel, silver and lead were generallybelow the limit of reporting and were not included in this analysis. ............................................. 11Table 9 – Regional means for the metal and metalloid concentrations in mud crab hepatopancreas(mg/kg dry weight basis). ............................................................................................................ 13Table 10 – Mean measured concentrations of metals and metalloids in barramundi gills (mg/kg dryweight) from Reference and Gladstone Harbour sites. ............................................................... 21Table 11 - Mean measured concentrations of metals and metalloids in barramundi liver (mg/kg dryweight) from Reference and Gladstone Harbour sites. ............................................................... 22Table 12 – Summary of effects concentrations associated with arsenic concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight) .................... 26Table 13 – Summary of effects concentrations associated with cadmium concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight). ................... 27Table 14 – Summary of effects concentrations associated with zinc concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight). ................... 28iii

Department of Science, Information Technology, Innovation and the ArtsList of figuresFigure 1 – Sampling sites for the tissue residue program. Note that this figure shows the locationsof all sites sampled for the Expanded Gladstone Fish Health Survey. Note: only thosesites included in Table 1 were part of the tissue residue study. .................................................... 4Figure 2 – Mean measured aluminium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 14Figure 3 – Mean measured arsenic concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 14Figure 4 Mean measured barium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 15Figure 5 – Mean measured cadmium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 15Figure 6 – Mean measured chromium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study Note: Gladstone Harbour samples from1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 16Figure 7 – Mean measured copper concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 16Figure 8 – Mean measured iron concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and data from this study. Note: Gladstone Harbour samples from 1996 includeddata from the Harbour and the Narrows, Gladstone Port 2012 samples were collectedfrom the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data fromNegri et al (2009). ....................................................................................................................... 17Figure 9 – Mean measured mercury concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofiv

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 17Figure 10 – Mean measured nickel concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 18Figure 11 – Mean measured selenium concentrations(±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 18Figure 12 – Mean measured zinc concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast ofQueensland (1996-2006) and data from this study. Note: Gladstone Harbour samplesfrom 1996 included data from the Harbour and the Narrows, Gladstone Port 2012 sampleswere collected from the Port Development Area. 1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009). .............................................................................................. 19Figure 13 – A principal component analysis of mean measured concentrations of metals andmetalloids in mud crab (Scylla serrata) hepatopancreas (mg/kg dry weight) from historicalsurveys along the east coast of Queensland (1996-2006) and data from the 2012 FishHealth Survey. Aluminium and barium were not included as they were not measured at allsites in all years. ‘Gl_Port12’ refers to samples collected in the Port Development Areaduring the current study and ‘Gl_Narr12’ refers to samples collected in the Narrows in thecurrent study. .............................................................................................................................. 20Figure 14 - Mean measured concentrations of iron in barramundi liver (mg/kg dry weight) by siteand by trip ................................................................................................................................... 22Figure 15 - Mean measured concentrations of cadmium in barramundi liver (mg/kg dry weight) bysite and by trip ............................................................................................................................. 23Figure 16 - Mean measured concentrations of arsenic in barramundi liver (mg/kg dry weight) bysite and by trip ............................................................................................................................. 23Figure 17 - Mean measured concentrations of zinc in barramundi liver (mg/kg dry weight) by siteand by trip ................................................................................................................................... 24v

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area1. IntroductionIn August 2011, the Queensland Government received reports of ill barramundi occurring in theBoyne River, Gladstone. Subsequently a variety of other aquatic species were allegedly affectedby illnesses in the Gladstone region. In response to this, Fisheries Queensland closed GladstoneHarbour on 16 September 2011 for a period of 21 days. The Queensland Government set up amulti-faceted investigation program into potential causes of the fish illness that included fish healthsampling and testing (DAFF 2012), water and sediment quality testing (DERM 2011a, b; EHP2012a, b, c, d, e, f, g, h, i; DSITIA 2013a, b) and the commissioning of a review on the possible linkbetween metals and metalloids and immunosuppression and increased disease susceptibility(Poulsen and Escher 2012).Initial investigations found that two conditions were apparent in diseased barramundi: red spotdisease (in one barramundi from Port Alma) and external parasitism by the fluke Neobenedeniasp., which was affecting the eye and skin of the barramundi (DAFF 2012). Also noted were a rangeof generally mild skin conditions, although no bacterial, parasitic or fungal pathogens wereidentified which could explain the skin conditions (DAFF 2012). Monthly water quality monitoring ofdissolved metals and metalloids for a year from September 2011 to September 2012 and early in2013 indicated levels of contaminants were generally below the Australian and New Zealand waterquality guidelines (ANZECC and ARMCANZ (2000), although some areas of the harbour did haveconcentrations of metals and metalloids that exceeded water quality guidelines consistently. Thesewere for dissolved copper in the Gladstone Marina and dissolved arsenic, molybdenum andaluminium in South Trees Inlet (DERM 2011, DERM 2012a, b, EHP 2012a, b, d, e, f, g, h, i;DSITIA 2013a, b). An extensive sediment survey undertaken in the harbour in February and March2012 did not find any contaminants at levels of concern (EHP 2012c).In October 2011 the Gladstone Fish Health Scientific Advisory Panel was established to provideindependent scientific advice to the Queensland Government on the fish health investigation inGladstone Harbour. In response to the recommendations of the Scientific Advisory Panel(Gladstone Fish Health Scientific Advisory Panel 2012), the Queensland Government commencedthe Gladstone Integrated Aquatic Investigation Program in order to understand the causes of fishhealth issues in Gladstone Harbour. This program commenced in late April 2012.This report outlines the results of tissue residue testing undertaken as part of the extendedGladstone Integrated Aquatic Investigation Program. The focus of the tissue residue work was totest the validity of the public perception of a link between dredging activities in Gladstone Harbourand fish and crab health. In particular, concern was raised that dredging caused the re-suspensionof sediments and the subsequent release of metals contaminants that then affected fish and crabhealth. A review of immunosuppression literature (Poulsen and Escher 2012) indicated thatexposure to dissolved metals at sufficiently high concentrations in water may increasesusceptibility to pathogens and diseases.An association between metals and crab disease has been raised previously in GladstoneHarbour. Crab shell disease has been reported in the harbour as early as 1994. Work by Andersen(2003) indicated there may be some association between metal contamination and crab shelldisease in mud crabs in Gladstone Harbour. Although this disease is found in other areas of thestate, Andersen (2003) reported higher prevalence rates of the disease in Gladstone Harbourcompared to reference sites.In order to assess any potential association between contaminants and disease, analysis for 13metals and metalloids was undertaken in mud crab (Scylla serrata) hepatopancreas, barramundiPage 1

Department of Science, Information Technology, Innovation and the Arts(Lates calcarifer) gills and liver, and grinner (Saurida sp.) muscle. In addition, a suite of 17polynuclear aromatic hydrocarbons (PAHs), 57 pesticides, 8 miscellaneous organic chemicals(including synthetic fragrances) and 32 polychlorinated biphenyls (PCBs) were analysed in mudcrab hepatopancreas. Historically, mud crabs have been used as biomonitors of long term waterquality trends for metals and metalloids and organics along the Queensland Coast (Mortimer 2000;Negri et al 2009) as they bioaccumulate these contaminants from the surrounding environment.These past studies provide data for comparison with results of the current study.In order to assess potential links between exposure to contaminants and disease in fish and crabsfrom the Gladstone region, three research questions were investigated using analysis of tissueresidues:1. Is there a link between the concentrations of contaminants in fish and crab tissue andvisible signs of animal health (diseased versus healthy)?2. Are the concentrations of contaminants in crab and fish tissue from Gladstone similar tothose from other Queensland estuarine and near-coastal locations?3. Are the levels of residues measured in fish and crab tissue known to be associated withecotoxicological effects reported in other studies?Page 2

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area2. Sampling SitesSampling was undertaken at eight sites during two sampling trips. The first sampling trip occurredbetween late April and early May 2012 and the second occurred between mid-June and early July2012. Three species were collected for tissue residue analysis and the sites they were collected atare summarised in Table 1 and the sites are illustrated in Figure 1. Sites that are considered to bereference sites in terms of dredging were Fitzroy River and Bundaberg. Awoonga Dam wasincluded for comparison.Table 1 – Sites where species used for tissue residue analysis were collectedSiteSpecies collectedMud crabs(Scylla serrata)Barramundi(Latescalcarifer)Grinner(Saurida sp.)Awoonga Dam •Upper Boyne River 1 •Port Development Area 1 •Hamilton Point 1 •The Narrows 1 •Gladstone Trawl 1, 3•Fitzroy River 2 • • •Bundaberg 2 • • •Note: Black circles indicate the species collected at each site. 1 Denotes Gladstone sites, 2 denotes referencesites. Awoonga Dam was excluded from of the comparison of reference vs. Gladstone sites as it is considered aseparate reference site (i.e. is a freshwater system and it will not be impacted by water quality from GladstoneHarbour), 3. This denotes samples collected by trawling within Gladstone Harbour.Page 3

Department of Science, Information Technology, Innovation and the ArtsFigure 1 – Sampling sites for the tissue t residue program. Note that this figure shows the locations ofall sites sampled for the Expanded Gladstonee Fish Health Survey. Note: N only those sites included inTable 1 were part of the tissue residue study. .Page 4

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone Area3. Methods3.1.Sample CollectionSamples were collected using methods outlined in DAFF (2012). Although A many specimens werecollected at each site,only five from each sampling round were sent for tissuee residue analysis(Table 2). Apart from Awoonga Dam where sampling only occurred once, andd Bundaberg whereno grinner were caught in Trip 1, 10 samples were submitted from each site (over the two trips).Fish and crabs were classed as healthy or diseased based on externally visible criteria.Fish/sharks were selected as unhealthy based on the presence of:• lesions andskin discoloration;• eye lesions;• Swelling orunusual mass on the abdomen;• Signs of emaciation (i.e. large head, thin body, poor body condition);• Haemorrhage from gills or vent.Crabs weree selected as unhealthy based onn the presence of:••••Rust spots;Shell erosion;Foulingof the shell;Lost appendages.Diseased fish were deliberately selected for, to be roughly 50 per cent of the sample populationwhere possible. This bias was addressed in the statistical analysiss (Appendices A and B).Table 2 – Number of samples fromeach site used for tissue residuee analysisSiteSpeciescollected per tripMud CrabBarramundiGrinnerTrip 1 Tripp 2Trip 1Trip T 2Trip 1Trip 2AwoongaDamUpper Boyne River1Port Development5Area 1HamiltonPoint 1The Narrows 15Gladstone Trawl 1Fitzroy River 25Bundaberg 25555555555555555555Note: 1 Denotes Gladstone sites, 2 denotes reference sites.Page 5

Department of Science, Information Technology, Innovation and the Arts3.2.Sample PreparationFish and crab specimens were taken to a laboratory in Gladstone and dissected as soonaspossible. Scalpel blades or knives were used for dissection and were cleanedd using soapy water,then wipedwith 100%ethanol onpaper towels between specimens. Sampless were frozenimmediately after dissection, andwere stored until theywere sent for analysiss at QueenslandHealth Forensic and Scientific Services (when all sampling was completed). Subsampless weresent to theQueensland Government Science Delivery Chemistry Centre C for dry weight analysis.3.3.LaboratoryAnalysis.Analysis for 13 metalsand metalloids (Tablee 3) was undertaken onn mud crab b hepatopancreas,barramundi gills and liver, and grinner muscle. The samples were macerated and sub-sampleswere takenfor analysis. The concentrationsof metals and metalloids were analysed by ICP-MS Iafter microwave digestion. The limit of reporting for each metal andd metalloid is presented in Table3.A suite of 17 polynuclear aromatic hydrocarbons (PAHs), 57 pesticides, 8 miscellaneouss organicchemicals (including synthetic fragrances) and 32 polychlorinatedbiphenyls (PCBs) were analysedin mud crab hepatopancreas. Organic chemicals were extracted by macerating the tissuein acetone and hexane, followed by Gel Permeation Chromatography to remove interferences.This was followed by a Florisil Macro Column Clean-up. Extracts were then concentrated byevaporation under vacuum and nitrogen andd analysis was undertaken using Gas ChromatographyMass Spectrometry.Dry weights were obtained by drying the tissue sample to a constant weight at 105°C, according tothe drying procedure outlined in ASTM method D4638-03 (ASTM International 2010) for moisturecontent.Table 3 – The limit of reporting for each metal and metalloid (on a wet w weight basis).Limit of reportingMetal(mg/kg)Aluminiumm

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone Area3.4. Statistical AnalysisVery few organics were detectedd in the mudd crab hepatopancreas(the only mud crab tissueanalysed) and so these results were not analysed statistically. Statistical analysis was undertakenusing a general linearmodel (GLM) in orderr to investigate whether there weree statisticallysignificant associations between observed signs of fishhealth (seee DAFF 2012 for methodology)and measured metal concentrations in the organisms sampled. In addition, analysis of themeasured concentrations of metals and metalloids at eference sites and Gladstone harbour sitesin each tissue group sampled were undertaken. Appendix A outlines the detailed statisticalmethods and results from the statistical analysis of metal concentrations measured in themud crabhepatopancreas and Appendix B outlines the methods and resultss for the assessment off themeasured metals andmetalloids in barramundi gill andliver tissuee samples and grinner muscletissue. A sensitivity analysis was undertakenn where samples had measured m concentrations at orjust above the limit of reporting. This analysis is presented in Appendix C.Data from historical analyses of measured metal concentrations inn mud crab hepatopancreastissue expressed as dry weights were collected from the literature (Mortimer 2000; Negri et al2009) and combined with data from the current survey. A principall component analysis (PCA)(wasundertakenn on the combined dataa in order too assess whether the current c tissue samples fromGladstone had unusually high contaminantss concentrations compared to historical data from fGladstone Harbour orother catchments along the east coast of Queensland.Principal componentsanalysis (PCA) (Stevens 1986) isa multivariate statistical method. . It constructs linear combinationsof all the variates thatt maximize the t variationn containedwithin them, thereby summarising theoverall patterns and displaying most of the original variability in a smaller s number of dimensions(typically two, visualised as a scatterplot). No historical concentration data could be located forbarramundi gill and liver or grinner flesh andd so analysis of these samples s was limited tocomparison between reference and Gladstone Harbourr sites, undertaken using the GLM. The levelof statistical inferencee for all statistical analyses was taken to be 0.05.3.5.Toxicological Effects AssessmentTissue residue data from the current study were compared to dataa retrieved from theEnvironmental Residue and Effects Database (US Army Corp Engineers & USEPA 2011). Thisdatabase is sourced from the literature. It contains dataa from studies where biological effects andtissue contaminant concentrations were simultaneouslymeasuredd in the same organism. Datarelating to mortality/survival, growth and reproduction were used for the comparison.Page 7

Department of Science, Information Technology, Innovation and the Arts4. Is there a link betweenthe concentration of metalsand metalloids and organic contaminantsin tissueand animal health?4.1.4.1.1.Mud CrabOrganicsNone of the 114 organic contaminants that were analysed for weree detected in the crabhepatopancreas from samples collected in the Gladstone area, but some pesticides and somepolychlorinated biphenyls were found in samples from the reference sites. This indicatesthat theseorganic contaminantsare not contributing too the mud crab illness in the Gladstone area.4.1.2.Metals and metalloidsAnalysis ofthe concentrations off metals andd metalloidsin mud crab hepatopancreas indicated asignificantly (P≤0.05) higher mean concentration of aluminium andd selenium in diseasedd crabsthan in healthy crabs (Table 4). The mean concentration of arsenic, barium, cadmium, chromium,copper, iron, nickel, mercury andsilver weree higher in diseased crab hepatopancreas, although thedifference was not significant (P> >0.05) (Table 4). The mean concentration of zinc and lead in thediseased crab hepatopancreas were lower than that found in the healthy h crabs, althoughh again thedifferenceswere not significant ( P>0.05). There was nosignificantt difference between mud crabshell disease prevalence when the mud crabb shell disease data were pooled into Gladstone sitesand reference sites (P>0.05).Andersen (2003) investigated thepotential link betweencrab lesions (shell disease) andmetalcontamination. She found no significant differences (P> >0.05) between metalss and metalloids in thehepatopancreas of diseased andhealthy mud crabs although she found a general, but notstatisticallysignificant(P>0.05) trend of elevated metal and metalloids concentrations in thehepatopancreas of diseased crabs. It shouldd be noted she did not analyse forr aluminiumm orselenium inmud crab tissue. Copper exposure experiments undertaken by Andersen (2003)indicated that copper inhibited calcium uptake into the carapace off soft shell juvenile crabs,affecting the production of the hard carapace.Page 8

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 4 – Disease status and mean concentrations of metals and metalloids (mg/kg) in the mud crabhepatopancreas (dry-weight basis). Metals and metalloids that have significantly differentconcentrations (P≤0.05) between healthy and diseased crabs are bolded and shaded in light blue.Metal MeanMeanAverage Significance Ratio of meanconcentration concentration Standard level (P) metalin healthy crabs in diseasedcrabsErrorconcentrationdiseased/healthyAluminium 12.9 21.6 2.5 0.021 1.67Arsenic 21.4 26.4 2.2 0.123 1.23Barium 1.26 1.47 0.25 0.554 1.17Cadmium 2.60 3.51 0.47 0.181 1.35Chromium 0.13 0.19 0.02 0.072 1.50Copper 630 670 125 0.826 1.06Iron 256 323 27 0.097 1.26Lead 0.12 0.09 0.02 0.282 0.75Nickel 2.57 2.77 0.39 0.727 1.08Mercury 0.27 0.29 0.03 0.665 1.06Selenium 6.30 7.69 0.38 0.016 1.22Silver 4.03 4.71 0.82 0.575 1.17Zinc 186 177 20 0.761 0.95Assessment was undertaken to determine whether there was a link between the significantlyhigher concentrations of aluminium and selenium in diseased crabs and location of the site. Thediseased crabs collected from the Fitzroy River had significantly (P≤0.05) higher aluminiumconcentrations in their hepatopancreas compared to diseased crabs at other sites (Table 5).However, this result should be interpreted with caution due to the low replication of diseased crabsper site (n=5). No significant relationship (P>0.05) was found between site and seleniumconcentrations in diseased crabs (Table 6).Table 5 – Aluminium concentrations (mg/kg) in the hepatopancreas of diseased and healthy crabs bysite (dry-weight basis, n=5). Sites that have a significant difference (P≤0.05) between healthy anddiseased crabs are bolded and shaded in light blue.MetalMeanconcentration indiseased crabsStandardErrorMeanconcentration inhealthy crabsStandardErrorBundaberg 14.61 4.49 13.70 4.49Fitzroy River 40.52 4.49 13.85 4.49The Narrows 14.60 4.49 10.24 4.49Port Development Area 16.55 6.06 12.96 3.76Page 9

Department of Science, Information Technology, Innovation and the ArtsTable 6 – Selenium concentrations (mg/kg) inn the hepatopancreas of o diseased and healthycrabs bysite (dry-weight basis, n=5).MetalMean StandardMean StandardconcentrationErrorconcentrationErrorin diseasedin healthyycrabscrabsBundabergFitzroy RiverThe NarrowsPort DevelopmentArea6.3997.3448.0009.5000.800.800.801.075.145.647.517.160.800.800.800.674.2.BarramundiDiseased fish were classified as any fish with a skin abnormality (including redness or lesions), eyelesions, swelling or unusual masss on the abdomen, signs of emaciation, or haemorrhagee from gillsor vent. There was nosignificant (P>0.05) associationbetween fish disease and measured metalconcentrations in barramundi gills (Table 7) or barramundi liver (Table 8). This indicates thatexposure to metals and metalloids overall iss unlikely to be contributing to immunosuppressiveeffects in Gladstone Harbour barramundi.Table 7 – Disease status and gill metal and metalloid concentrationss (mg/kg dry-weight basis). Nosignificant difference (P>0.05) between healthy and diseased fish were found. Measuredconcentrations of chromium, nickel, silver and lead weree generally below the limit of reporting andwere not included in this analysis.Metal Mean Meanconcentrationconcentrattionin healthy fish in diseasedfish fAluminium ArsenicBarium13.51.731.6418.13.341.26AverageSignificanceRatio of meanStandardlevel (P)Errormetalconcentrationdiseased/ /healthy2.60.3241.340.610.0981.930.580.6150.77Cadmiumm CopperIronMercurySeleniumZinc0. .0453.411890. .2862.9056.20.0755 4.953740.4000 3.5965.70. 0150.80800. 0410.364.00.1220.2080.1250.0520.2430.1311.661.451.981.401.241. 17Page 10

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone AreaTable 8 Disease status and meanmeasured concentrations of metal contaminants in liverr (mg/kgdry-weight basis). No significant difference (P>0.05) between healthy and diseased fish were found.Measured concentrations of chromium, nickel, silver andlead were generally below the limit ofreporting and were not included in this analysis.MetalMeanconcentrationin healthy fishMeanconcentrationin diseasedfishAverageStandardErrorSignificancScelevel l (P)Ratio of averageemetalconcentrationdiseased/healthyAluminiumm 2.263.1880.530.4891.41Arsenic4.405.8000.800.4291.32Cadmium0. 1060.1377 0.0190.5621.28Copper44.818.5510.20.0650.41Iron1420190112820.4311.34Mercury0. 5970.6244 0.1450.8431.05Selenium5.364.5220.860.3730.84Silver0. 2600.0800 0.0630.0550.31Zinc67.966.007.70.7010.974.3.GrinnerAll grinner were assessed as being healthy and therefore no statistical analysis betweentissueconcentrations and health status were conducted.4.4.Summary• There were significantly higher concentrations of aluminiumm and selenium in diseased mudcrab hepatopancreas compared to healthy mud crabs. As this t result occurred forr all sites,these elevatedmetal concentrationsmay be contributing too the prevalence of mud crabdisease in general, not just Gladstone Harbour.• There was no significant associationn between fish disease and measured metalconcentrationsin barramundi gills orr liver and therefore no evidence that metals andmetalloids were related toany illnesss of this species.• There is no evidence thatt organic contaminantsanalysed for f are contributing to mud crabillness in the Gladstone sites.Page 11

Department of Science, Information Technology, Innovation and the Arts5. Are the concentrationsof contaminants inn crab andfish tissuefrom Gladsttone similar to thosee from otherQueensland estuarine and near-coastal locations?Measured concentration of contaminants in tissues collected from aquatic organisms can be usedto assess overall environmental exposure too contaminants over time, and may indicate exposure oforganisms to contaminants that are not identified from discrete grab sampling. Comparisons ofcontaminants in tissues of the same speciess collected from different areas may indicate variation inthe contaminant exposure. Five contaminannts (aluminium, arsenic, copper, molybdenum, and zinc)were found, during thewater quality monitoring program(DERM 2011; 2 DERMM 2012 a, b, c EHP2012a, b, d, e, f, g, h, i; DSITIA 2013a, b), too have exceeded the Australian A and New Zealandwater quality guidelines (ANZECC and ARMCANZ 2000).5.1.Mud CrabComparison of the pooled metal and metalloid concentrations in mud m crab hepatopancreas fromeach trip inthis study indicated significant differences between trips but theree was generally noconsistent pattern between trips apart from concentrations of barium and selenium. Barium wassignificantly higher in the mud crab hepatopancreas at reference sites s compared to GladstoneHarbour sites for bothtrips and selenium was significantly higher at a Gladstone sites for both trips(Table 9). Copper andzinc were significantlyy higher at the Gladstone Harbour sites in Trip 1, butthere was no significant difference in Trip 2. Iron was significantlyhigher in Trip 2 at the Gladstonesites. Chromium and lead were significantlyhigher at the reference sites in Trip 1, and cadmiumand mercury were significantly higher in Tripp 2. This may reflect slightly different areas from whichthe crabs were collected. In order to put these results inperspective at a larger scale, comparisonswere madewith historical data collected in estuarine and near coastal locations.Comparison of concentrations off metals andd metalloidsin the hepatopancreas of mud crabscollected from the Gladstone Port Development Area (Gladstone Port) P and the Narrows (GladstoneNarrows) in2012 (Figure 1) showed that crabs from these sites did not have higher level of metalsand metalloids in theirhepatopancreas compared to mud crabs collected fromm other estuarine andnear-coastal regions of eastern Queenslandd (Figure 2 to Figure 12). Mean measuredconcentrations of aluminium, arsenic, barium, cadmium, chromium, iron, mercury, nickell andselenium ( Figures 2 to6 and Figures 8 to 11) in the Port Development Area tended to liein thelower half of the measured concentrations of samples collected along the eastern coast ofQueensland. Mean measured concentrations of copperand zinc ( Figure 7 and 12) in thePortDevelopment Area were in the upper half of f the historical data. Andersen (2003) also found thesetwo metalsto be elevated in crabhepatopancreas in Gladstone Harbour compared to a referencesite at Ayr.A direct comparison of the hepatopancreasmetal concentrations from the current studycannot be made with data presented in Andersen (2003) as in the latter the data were reported aswet weight. The concentration off zinc measured in the crab hepatopancreas in the PortDevelopment Area in 2012 was lower than that reported by Mortimer (2000) in the Gladstone area,although the concentration of copper in the hepatopancreas was higher. h Angel et al. (2012) foundthe copperconcentration in water in Gladstone Harbour had increased from previous studies andsuggested this was caused by increased shipping. Measured concentrationsof mercury in the PortDevelopment Area were in the higher half off the historical dataset, , but were much lower thanmeasured concentrations in mudcrabs fromthe Tully, Herbert, and Gordon catchments(Figure 9).Silver was not measured in mud crab hepatopancreasin the previous studiess and so historicalPage 12

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Areacomparisons are not possible. However, comparison between the reference and Gladstone sitesfound no significant difference in concentrations in silver (Table 9).Overall, these results indicate that the water and sediment quality that mud crabs are exposed to inGladstone is similar to that of other estuarine and near-coastal regions of eastern Queensland.Table 9 – Regional means for the metal and metalloid concentrations in mud crab hepatopancreas(mg/kg dry weight basis).Trip 1 Trip 2ReferenceGladstone SignificancelevelReferenceGladstone SignificancelevelAluminium 16.7 14.7 0.646 12.2 9.0 0.513Chromium 0.186 0.098 0.031 0.120 0.166 0.288Iron 273 188 0.057 187 418 0.000Nickel 3.22 3.16 0.938 1.48 2.43 0.306Copper 526 1127 0.004 621 294 0.139Zinc 103 195 0.002 209 242 0.275Arsenic 17.3 20.2 0.419 23.2 27.4 0.285Selenium 5.89 7.46 0.027 5.03 7.36 0.004Silver 4.80 3.67 0.401 4.91 3.18 0.251Cadmium 2.47 1.55 0.268 4.45 2.41 0.030Barium 1.55 0.68 0.021 2.25 0.97 0.003Mercury 0.283 0.245 0.349 0.416 0.155 0.000Lead 0.155 0.069 0.046 0.072 0.180 0.024Note: Blue cells indicate a significant difference between the metal or metalloid concentrations at the referenceand Gladstone sites, bolded cells indicate measured concentrations at pooled Gladstone sites were significantlyhigher.Page 13

Department of Science, Information Technology, Innovation and the Arts200150Aluminium1996200520062012100500NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 2 – Mean measured aluminium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast of Queensland(1996-2006) and data from this study. Note: Gladstone Harbour samples from 1996 included data from theHarbour and the Narrows, Gladstone Port 2012 samples were collected from the Port Development Area.1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009).12010080Arsenic199419962005200620126040200NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 3 – Mean measured arsenic concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).Page 14

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area54Barium19962005200620123210NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 4 Mean measured barium concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).2015Cadmium199419962005200620121050NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 5 – Mean measured cadmium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast of Queensland(1996-2006) and data from this study. Note: Gladstone Harbour samples from 1996 included data from theHarbour and the Narrows, Gladstone Port 2012 samples were collected from the Port Development Area.1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009).Page 15

Department of Science, Information Technology, Innovation and the Arts43Chromium199419962005200620121994210NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 6 – Mean measured chromium concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast of Queensland(1996-2006) and data from this study Note: Gladstone Harbour samples from 1996 included data from theHarbour and the Narrows, Gladstone Port 2012 samples were collected from the Port Development Area.1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009).15001000Copper199419962005200620125000NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 7 – Mean measured copper concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).Page 16

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area1000800600Iron199419962005200620124002000NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 8 – Mean measured iron concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).32.52Mercury199419962005200620121.510.50NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 9 – Mean measured mercury concentrations (±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast of Queensland(1996-2006) and data from this study. Note: Gladstone Harbour samples from 1996 included data from theHarbour and the Narrows, Gladstone Port 2012 samples were collected from the Port Development Area.1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009).Page 17

Department of Science, Information Technology, Innovation and the Arts1510Nickel1994199620052006201250NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 10 – Mean measured nickel concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).15Selenium199419962005200620121050NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 11 – Mean measured selenium concentrations(±SE) in mud crab (Scylla serrata)hepatopancreas (mg/kg dry weight) from historical surveys along the east coast of Queensland(1996-2006) and data from this study. Note: Gladstone Harbour samples from 1996 included data from theHarbour and the Narrows, Gladstone Port 2012 samples were collected from the Port Development Area.1996 data from Mortimer (2000); 2005-2006 data from Negri et al (2009).Page 18

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area400350300250Zinc19941996200520062012200150100500NormanbyBarronJohnstoneTullyHerbertGordonBurdekinO'ConnellPioneerFitzroyGladstone HarbourGladstone PortGladstone NarrowsBurnettMaroochyBrisbaneFigure 12 – Mean measured zinc concentrations (±SE) in mud crab (Scylla serrata) hepatopancreas(mg/kg dry weight) from historical surveys along the east coast of Queensland (1996-2006) and datafrom this study. Note: Gladstone Harbour samples from 1996 included data from the Harbour and theNarrows, Gladstone Port 2012 samples were collected from the Port Development Area. 1996 data fromMortimer (2000); 2005-2006 data from Negri et al (2009).A PCA of the same data also indicated that the measured concentrations of metals and metalloidsin crab hepatopancreas were similar to other catchments and historical data (Figure 13).If tissue metal concentrations are linked to organism health, one might expect that theconcentrations at Gladstone would be higher than those from other areas. The fact that the tissueconcentrations from Gladstone are fairly typical indicates that metal and metalloids in tissues arenot responsible for the organism health issues at Gladstone.Page 19

Department of Science, Information Technology, Innovation and the Arts4Fitzroy'06Fitzroy'05Normanby'052O'Connell'06Normanby'06Burdekin'06Gordon'05Dimension 2 (24% of variation)0Gl_Harbour'96Brisbane'96Fitzroy'12Pioneer'06Herbert'06Johnstone'05Gordon'06O'Connell'05Gl_Port'12Barron'06Gl_Narr'12Burnett'06Burnett'05Johnstone'06Barron'05Burdekin'05Herbert'05Tully'05Tully'06-2Burnett'12Maroochy'94Maroochy'96-4-4 -2 0 2 4Dimension 1 (26% of variation)Figure 13 – A principal component analysis of mean measured concentrations of metals andmetalloids in mud crab (Scylla serrata) hepatopancreas (mg/kg dry weight) from historical surveysalong the east coast of Queensland (1996-2006) and data from the 2012 Fish Health Survey.Aluminium and barium were not included as they were not measured at all sites in all years.‘Gl_Port12’ refers to samples collected in the Port Development Area during the current study and‘Gl_Narr12’ refers to samples collected in the Narrows in the current study.Page 20

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone Area5.2. BarramundiThere are no historical data for similar Queensland estuarine and near-coastal sites available tocompare with the results from thecurrent study. The measured concentrations of metalsinbarramundi gills from Gladstone Harbour were similar or less than those fromm the reference sites(Table 10) . The one exception was arsenic which was twice as high in the Gladstone sites than thereference sites, however, this was not a statistically significant difference (P>0.05).The mean measured concentrations of iron and cadmium in barramundi liverss collected during thefirst sampling round were significantly (P≤0. 05) higher at the Gladstone sites (Hamilton Point andUpper Boyne) compared to reference sites (Table 11), and arsenicc and zinc were significantlyhigher in the second sampling round (Table 9). The significantly higher levelss of these metals andmetalloids were drivenby resultss from specimens sampled at Hamilton Point (Figure 14 to Figure17). However, again the small number of samples taken at each site and trip (n=5) means thatthese results should be interpreted with caution.Table 10 – Mean measured concentrations off metals andmetalloidss in barramundi gills (mg/kg dryweight) from Reference and Gladstone Harbour sites.MetalTrip 1Trip 2ReferencesitesGladstonesitessSignificancelevel (P)ReferencesitesGladstonesitesSignificancelevel (P)Aluminiumm ArsenicBariumCadmiumCopperIronMercurySeleniumZinc30.33 1.821.650.0946.044210.3813.633 67.26.13.790.340.0783.922490.2523.3262.70.0000.1190.2330.6250.2200.3070.1250.6790.60014.5 16.02.87 2. 681.97 1. 530.0511 0.0214.62 3. 11393 1730.3933 0.3663.56 3. 1462.1 55.90.7670.8750.6860.3510.3870.2020.7510.5710.482Note: Chromium, lead, nickel and silver had manyy measurements below thee limit of reporting and were notincluded in the analysis. Blue bolded cells indicate a significant difference between the metal concentrations atthe reference and Gladstone sites.Page 21

Department of Science, Information Technology, Innovation and the ArtsTable 11 - Mean measured concentrations of metals and metalloids in barramundi liver (mg/kg dryweight) from Reference and Gladstone Harbour sites.Metal Trip 1 Trip 2ReferenceSitesGladstoneSitesSignificancelevel (P)ReferencesitesGladstoneSitesSignificancelevel (P)Aluminium 2.44 3.30 0.390 2.24 3.97 0.093Arsenic 5.37 7.68 0.140 2.52 6.13 0.026Iron 1695 2782 0.046 969 1550 0.284Cadmium 0.118 0.196 0.035 0.068 0.119 0.165Copper 45.7 37.4 0.726 34.8 19.7 0.529Mercury 0.748 0.779 0.911 0.396 0.586 0.482Selenium 4.69 6.47 0.287 4.41 4.96 0.744Silver 0.163 0.141 0.867 0.245 0.151 0.496Zinc 67.9 89.8 0.124 42.7 71.5 0.048Note: Barium, chromium, lead, nickel and silver had many measurements below the limit of reporting and thuswere not included in the analysis. Blue bolded cells indicate a significant difference occurred between metalconcentrations in barramundi livers collected from Gladstone and Reference sites.4000IronTrip 1Trip 23000200010000Awonga DamBundabergFitzroyHamilton PointUpper BoyneFigure 14 - Mean measured concentrations of iron in barramundi liver (mg/kg dry weight) by site andby tripPage 22

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area0.350.3CadmiumTrip 1Trip 20.250.20.150.10.050Awonga DamBundabergFitzroyHamilton PointUpper BoyneFigure 15 - Mean measured concentrations of cadmium in barramundi liver (mg/kg dry weight) by siteand by trip15ArsenicTrip 1Trip 21050Awonga DamBundabergFitzroyHamilton PointUpper BoyneFigure 16 - Mean measured concentrations of arsenic in barramundi liver (mg/kg dry weight) by siteand by tripPage 23

Department of Science, Information Technology, Innovation and the Arts150ZincTrip 1Trip 2100500Awonga DamBundabergFitzroyHamilton Pon PointUpper BoyneFigure 17 -by tripMean measured concentrations of zinc in barramundi liver (mg/kg dry weight) by site and5.3.GrinnerAluminium, iron, copper and zincwere significantly higher (P≤0.05) in grinnerr muscle collected atBundaberg(a reference site) compared to other sites, but there were no other trends (Appendix B).5.4.SummaryComparisons revealedthat mud crab hepatopancreasconcentrations of metals and metalloidswere similar to other estuaries and near-coastal sites along the east coast of Queensland.Barramundi livers fromGladstonesometimes had higher concentrations of arsenic, cadmium, ironand zinc than the reference sites, but there was no relationship between the concentrations ofthese metals and the health of this species. The presence of elevated concentrations of arsenic,iron and zinc in barramundi liversis consistent with the fact that these metals s were detected inwaters of the Gladstone area. The elevated concentration of cadmium in barramundi liver isunusual given that it was rarely quantified inn the watersof the Gladstone area.Page 24

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone Area6. Are the levels of contaminants in fish andknown to be associatedwith reportedecotoxicologicaleffects?crab tissue6.1.IntroductionWhere residues were identified as being significantly elevated at Gladstone G sites compared toreference sites or were significantly associated with fishor crab health issues, results werecompared to data retrieved from the Environmental Residue and Effects E Database (ERED,http://el.erdc.usace.army.mil/ered/). If there were no significant differences between measuredmetal concentrationsof diseasedd and healthy fish, or Gladstone Harbour H andd reference sites thenthese metals and metalloids could be discounted as contributing too the fish disease in GladstoneHarbour.6.2.Metal Residues in the t Hepatopancreas of Mud CrabsThe concentration of aluminium and selenium was found to be significantly higher in diseasedcrabs than healthy crabs (Section4.1). No data for toxicological effects basedd on aluminium orselenium concentrations in crab hepatopancreas was found in the Environmental Residue andEffects Database, so the elevated concentrations of these metals and metalloids could not beassessed in terms of their potential effects.6.3.Metal Residues in the t Gillsand Liver of BarramundiMeasured concentrations of arsenic, cadmium, iron andzinc in thee liver of barramundi were foundto be significantly higher in the group of samples collected for at least one tripp in the GladstoneHarbour samples, particularly in the t Port DevelopmentArea (Figure 17).6.3.1.ArsenicArsenic was detected in liver tissue at concentrations ranging between 0.29 and 6.37 mg/kg wetweight at Gladstone and reference sites, with the highest level of arsenic a (6.37 mg/kg) beingmeasured in a samplefrom the Bundaberg(reference) site. LD50ss (concentrations lethal to 50% ofthe test organisms) were associated with liver concentrations of arsenic between 5.69 mg/kg wetweight (Mozambiquetilapia - Oreochromis mossambicus) and 122.8 mg/kg (Green sunfish -Lepomis cyanellus) (US Army Corp Engineers & USEPA, 2011) (Table 12). Of the 45 barramundiliver samples collected in the program, only two had arsenic concentrations above the lowestrecorded effects concentrations and the vast majority were significantly lower. The only chroniceffect noted for arsenic concentrations was at a No Observed Effect Doss (NOED) of 11.6 mg/kgwet weightin bluegill Lepomis macrochirus.This is nearly double the t maximum concentration ofliver arsenic measured in this study, therefore it is unlikely that arsenic residues in barramundilivers would cause harmful effects to fish.There weree no data on immunosuppressioneffects in fish due to arsenic a in the Poulsen andEscher (2012) report.Page 25

Department of Science, Information Technology, Innovation and the ArtsTable 12 – Summary of effects concentrations associated with arsenic concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight)SpeciescommonnameBluegillBluegillGreen sunfishGreen sunfishGreen sunfishGreen sunfishGreen sunfishMozambiquetilapiaMozambiquetilapiaKillfishKillfishSpeciesscientificnameLepomismacrochirusLepomismacrochirusLepomiscyanellusLepomiscyanellusLepomiscyanellusLepomiscyanellusLepomiscyanellusOreochromismossambicusOreochromismossambicusFundulusheteroclitusFundulusheteroclitusConcentrationwet weight(mg/kg)EffectclassToxicitymeasure 1Exposureroute11.6 Growth NOED Not defined11.6 Mortality NOED Not defined27.4 Mortality LD50 Water77.2 Mortality LD50 Water77.8 Mortality LD50 Water82 Mortality LD50 Water122.8 Mortality LD50 Water5.69 Mortality LD50 Water9.55 Mortality LD10 Water17.98 Mortality LD20 Water29.22 Mortality LD70 Water1LD10, LD20 and LD70 are the concentrations of arsenic in liver tissue that correspond to 10, 20 and 70 per centlethality of the test organism, respectively.6.3.2. CadmiumCadmium was rarely detected by the Gladstone Integrated Aquatic Investigation Program (DERM2011a, b; EHP 2012a, b, c, d, e, f, g, h, i; DSITIA 2013a, b). By comparing the concentrations ofmetals in Gladstone Harbour and the effect concentrations reported by Poulsen and Escher(2012), DISITIA (in prep) found that it was unlikely cadmium was causing immunosuppression orincreasing disease susceptibility in fish.The measured range of concentrations of cadmium in the liver of the barramundi sampled from theGladstone area was 0.01 to 0.1 mg/kg wet weight, whereas the range from the reference sites was

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaNOED values for growth ranged from 2.5x10 -5 to 3.8 mg/kg (US Army Corp Engineers & USEPA,2011). The lowest concentrations reported for mortality were 1.37 x10 -4 mg/kg in juvenile brooktrout (Salvelinus fontinalis) a freshwater species (Hamilton, Mehrle and Jones 1987). That studyreported particularly low concentrations of cadmium in liver associated with mortality, whereas allother studies reported concentrations of cadmium in the liver where effects were measured at0.175 mg/kg wet weight or higher, higher than that found in the Gladstone Harbour residue study.Overall, the concentrations of cadmium measured during the 2012 study indicate that it is possiblethat the cadmium measured in the barramundi tissue could be having an effect. However, themajority of effects retrieved from ERED were much higher than the concentrations detected in thebarramundi collected from the Gladstone area, and therefore it was unlikely that the measuredconcentrations of cadmium in Gladstone Harbour were related to immune-suppression. Analysis ofthe cadmium concentrations in healthy and diseased fish indicated no significant associationbetween fish illness and cadmium concentrations in the liver.Table 13 – Summary of effects concentrations associated with cadmium concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight).Effect classToxicitymeasureRange ofconcentrationswet weight(mg/kg) fromEREDRange ofconcentrationswet weight(mg/kg) atGladstone sitesRange ofconcentrationswet weight(mg/kg) atReference sitesGrowth NOED 2.5x10 -5 - 3.8 0.01 - 0.1

Department of Science, Information Technology, Innovation and the Artsliver concentration of 34.44 mg/kg in yellow perch (Perca flavescens) (i.e. the ED4 is the dose thatexerts a 4% reduction in growth). Thus it is possible for such a minor effect (4% reduction ingrowth) to occur in both Gladstone and the reference sites. The next lowest was a growth ED26(the dose that exerts a 26% reduction in growth) at 34.44 mg/kg in the liver tissue of yellow perch,which is the higher end of the concentration of zinc measured in the livers of barramundi. Only 3samples of 45 barramundi samples collected had zinc levels higher than 34.44 mg/kg wet weight,two of these were from Hamilton Point in the Gladstone Area and one from a reference site.Zinc concentrations measured during the Gladstone Integrated Aquatic Investigation Programwere, on occasion, higher than the concentrations observed to cause chemical inducedimmunosuppression effects in the most sensitive species. However, the low frequency and limitedspatial distribution of the measured zinc concentrations are not consistent with zinc causing sucheffects (see DSITIA in prep).Table 14 – Summary of effects concentrations associated with zinc concentrations in fish liverreported in the Environmental Residue and Effects Database (mg/kg wet weight).SpeciescommonnameSpecies scientificnameConcentrationwet weight(mg/kg)Effect class ToxicitymeasureExposurerouteBrook Trout Salvelinus fontinalis 68 Growth NOED WaterBrook Trout Salvelinus fontinalis 60 Growth NOED AbsorptionYellowPerchYellowPerchBluegillPerca flavescens 34.44 Growth ED4 NotStatedPerca flavescens 34.44 Growth ED26 NotStatedLepomismacrochirus207 Growth ED89 NotdefinedBrook Trout Salvelinus fontinalis 68 Reproduction NOED WaterBrook Trout Salvelinus fontinalis 60 Reproduction NOED AbsorptionBrook Trout Salvelinus fontinalis 60 Mortality NOED AbsorptionThreespinedSticklebackGasterosteusaculeatus4.05 Mortality NOED WaterGoldfish Carassius auratus 3.36 Mortality NOED WaterGoldfish Carassius auratus 10.51 Mortality NOED WaterGoldfish Carassius auratus 5.34 Mortality NOED WaterGoldfish Carassius auratus 15.06 Mortality NOED WaterGoldfish Carassius auratus 10.69 Mortality NOED WaterGoldfish Carassius auratus 21.49 Mortality NOED WaterTrout -BrookSalvelinus fontinalis 68 Survival NOED WaterPage 28

Contaminants in Tissue of o Fish and Crabs Collectedin the Gladstone Area6.4. SummaryArsenic, cadmium, iron and zinc were identified as being sometimes significantly elevated in thebarramundi livers collected at Gladstone sites compared to references sites. A comparison withdata retrieved from the Environmental Residue and Effects Database(http://el.erdc.usace.army.mil/ered/) revealed that:• Thevast majority of samples had arsenic, cadmium and zinc concentrations which werebelow the concentrationsassociatedd with toxicological effects.• This combinedwith the lack of statistical significance between fish disease and metalconcentrationindicated that it is unlikely that metals and metalloids are associated with fishdisease in Gladstone Harbour.Comparison with immunosuppression data indicated that it is unlikely that cadmium or zinc wererelated to fish diseasee in Gladstone Harbourr (DISITIA in prep). No data were available for arsenicand iron.Page 29

Department of Science, Information Technology, Innovation and the Arts7. ConclusionsThere was no indication of elevated metals and metalloids in the tissues of diseased barramundicompared to healthy barramundi collected in the Gladstone area. Significantly higher arsenic,cadmium, iron and zinc concentrations were found in barramundi livers from at least one samplingtrip in the Gladstone area, indicating possible elevated exposure to these metals and metalloids.The measurement of elevated concentrations of these metals and metalloids does not necessarilyindicate a toxicological problem, but does indicate exposure. Comparison of concentrations ofthese metals and metalloids in barramundi livers with toxicological effects data did not, overall,indicate that the concentrations in the barramundi liver were of concern. In addition, as there wereno significant associations between these metals and metalloids and fish health, it is unlikely thatthey are associated with the fish health in Gladstone Harbour.Analysis of crabs from both Gladstone Harbour and reference sites found a significant associationof aluminium and selenium concentrations and disease. Concentrations of metals and metalloids inmud crab hepatopancreas detected in the Gladstone area were similar to sites along the eastcoast of Queensland. The available data also indicated that the suite of organic chemicals testedfor did not contribute to the crab disease.All grinner were assessed as being healthy so the existence relationships between metalconcentrations and observed fish health could not be assessed for this species.Overall, there is no strong evidence of a link between fish health at the time of sampling and tissueresidue concentrations.Page 30

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area8. ReferencesAndersen. L.E. (2003). A study into the epidemiology of mud crab (Scylla serrata) shell disease.Thesis: Master of Applied Science in the Centre for Environmental Management, Faculty of ArtsHealth and Science, Central Queensland University, Gladstone, Australia. May, 2003.Angel BM, Jarolimek CV, King JJ, Hales LT, Simpson SL, Jung RF and Apte SC (2012). Metalconcentrations in the waters and sediments of Port Curtis, Queensland. CSIRO Wealth fromOceans Flagship Technical Report. Sydney, 75p. Available from .ASTM International (2010), Annual book of ASTM Standards: section 11, water and environmentaltechnology, American Society for Testing and Materials International, West Conshohocken, PA.DAFF (Department of Agriculture, Fisheries and Forestry. Fisheries Queensland. (In prep)Gladstone fish health investigation: DAFF final report.DAFF (Department of Agriculture, Fisheries and Forestry). 2012. Fisheries Queensland. Expandedfish health survey. Interim report June/July 2012. 17p. Available from .DERM (Department of Environment and Resource Management). 2011. Water Quality of PortCurtis and Tributaries. Supplementary Report Based on Data Collected in the Week of 26 thSeptember. Environment and Resource Science, Department of Environment and ResourceManagement, Brisbane, Qld, 38p. Available from.DERM (Department of Environment and Resource Management). 2012a. Second Update on theWater Quality of Port Curtis and Tributaries Including Data Collected in the Week of 24 October2011. Environment and Resource Science, Department of Environment and ResourceManagement, Brisbane, Qld, 47p. Available from .DERM (Department of Environment and Resource Management). 2012b. Third Update on theWater Quality of Port Curtis and Tributaries Including Data Collected in the Weeks of 12 Novemberand 21 December 2011. Environment and Resource Science, Department of Environment andResource Management, Brisbane, Qld, 52p. Available from.DERM (Department of Environment and Resource Management). 2012c. Fourth Update on theWater Quality of Port Curtis and Tributaries Including Data Collected in the Week of 9 January2012. Environment and Resource Science, Department of Environment and ResourceManagement, Brisbane, Qld, 52p. Available from .DSITIA (Department of Science, Information Technology, Innovation and the Arts). 2013a. Postfloodwater quality monitoring in Gladstone Harbour and Waterways – January 2013. 10p.Available from .DSITIA (Department of Science, Information Technology, Innovation and the Arts). 2013b. Postfloodwater quality monitoring in Gladstone Harbour and Waterways – February 2013. 10p.Available from www.ehp.qld.gov.au/gladstone/pdf/monitoring-report-post-flood-feb2013.pdf.Page 31

Department of Science, Information Technology, Innovation and the ArtsDSITIA (Department of Science, Information Technology, Innovation and the Arts). In prep. Ananalysis of water quality in Gladstone Harbour and waterways between September 2011 andSeptember 2012. Water Sciences Technical Report Volume 2013 Number 3, ISSN 1834-3910,ISBN 978-1-7423-0987.EHP (Department of Environment and Heritage Protection). 2012a. Fifth Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 6 February 2012.Environment and Resource Science, Department of Environment and Heritage ProtectionBrisbane, Qld, 66p. Available from .EHP (Department of Environment and Heritage Protection). 2012b. Sixth Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 5 March 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 60p. Available from .EHP (Department of Environment and Heritage Protection). 2012c. Update on the Quality ofSediment from Port Curtis and Tributaries. Department of Environment and Heritage ProtectionBrisbane, Qld, 26p. Available from .EHP (Department of Environment and Heritage Protection). 2012d. Seventh Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 2 April 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 69p. Available from .EHP (Department of Environment and Heritage Protection). 2012e. Eighth Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 8 May 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 59p. Available from EHP (Department of Environment and Heritage Protection). 2012f. Ninth Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 5 June 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 55p. Available from EHP (Department of Environment and Heritage Protection). 2012g. Tenth Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 3 July 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 57p. Available from EHP (Department of Environment and Heritage Protection). 2012h. Eleventh Update on the WaterQuality of Port Curtis and Tributaries Including Data Collected in the Week of 1 August 2012.Environment and Resource Science, Department of Environment and Heritage Protection,Brisbane, Qld, 57p. Available from Page 32

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaEHP (Department of Environment and Heritage Protection). 2012i. September 2012 Test of WaterQuality in Port Curtis. Available from .Gladstone Fish Health Scientific Advisory Panel 2012. Gladstone Fish Health Scientific AdvisoryPanel – Final Report 5 January 2012. 47p. Available from.Hamilton SJ, Mehrle PM, and Jones JR. (1987) Cadmium-saturation technique for measuringmetallothionein in brook trout: Transactions of American Fisheries Society 116:541-550.Mortimer MR. (2000) Pesticide and trace metal concentrations in Queensland estuarine crabs.Marine Pollution Bulletin 41: 359-366.Negri AP, Mortimer M, Carter S and Mueller JF. (2009) Persistent organochlorines and metals inestuarine mud crabs of the Great Barrier Reef. Marine Pollution Bulletin. 58: 765-786.Poulsen AH and Escher BI 2012. Chemically induced immunosuppression and diseasesusceptibility in marine wildlife: A literature review. National Research Centre for environmentalToxicology (Entox), The University of Queensland, Brisbane, Queensland. 110p.Stevens J. (1986). Applied multivariate statistics for the social sciences. Hillsdale, NJ: LawrenceErlbaum Associates.US Army Corp and USEPA. 2011. The Environmental Residue and Effects Database. Availablefrom: http://el.erdc.usace.army.mil/ered/. Last updated October 2011. Downloaded: 12 April 2013.Page 33

Department of Science, Information Technology, Innovation and the ArtsAppendix AStatistical Report – Metals, Crabs1. STATISTICAL METHODSGeneral linear models (McCullagh and Nelder 1989) were used for the analyses of the metalconcentrations (on a dry-weight basis), under GenStat (2011). There were 40 crabs in total - 5samples (containing both normal and abnormal crabs) from each combination of the four locationsby two trips. Each crab was taken as an independent experimental unit. For samples reported as‘below the detection limit’, the usual convention of substituting half the detection level was used.This was done for the wet-weight data (upon which the test had been performed), and then thesevalues were corrected for moisture per cent back to a dry-weight basis (which were used for theanalyses).Again, the deliberate sampling of both diseased and unaffected crabs introduced further complexityinto the analyses. To adjust for this stratification, weighting factors were calculated as –Proportion (normal or abnormal) in the whole sample /Proportion (normal or abnormal) in the laboratory subsampleThese weighting factors were determined at the location by trip level, and then used to correctlyweight each of the respective observations (normal or abnormal) in the analyses for trips bylocations.Crab weight was trailed as a covariate, but for the 13 metals analysed was only significant (P =0.038) for barium, and this fitted as a negative. There were no other ‘near-significant’ fits; nor werethe fitted slopes predominantly positive (six out of 13 were negative). Similarly, carapace width hadno association with increased metal levels over time, with none being significant and seven fittingas negative slopes. Hence the expected bio-accumulation effect could not be estimated and adjustedfor, and so was dropped for all analyses.In addition to conducting the primary weighted analyses of location by trip (along with theinteraction), second analyses were conducted for each variable including ‘disease status’ in themodel. This allowed the direct testing and reporting of the impact of disease status on eachconcentration.The univariate analyses of each of the metals are followed by a multivariate summarisation. Thisfirst considers the relationships amongst the metals (via a correlation matrix), and then adopts aprincipal components representation of all the data (taking the first two orthogonal dimensions) toobtain an overview as to which locations and times are most similar.2. RESULTSWhilst there were a few high values for some of the metals, in general these data were not skewed,so no transformations were required. Of the 13 analyses, significant differences were found betweenlocations in seven and between trips in five. Importantly, there was a significant (P < 0.05) locationby trip interaction in four of the 13 analyses, with this effect being near-significant (0.05 < P

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Areathe changing patterns of responses over space and time are a real effect, so the data are reported onthis two-way basis (locations by trips).Disease status – Two of the 13 metals showed a significant (P < 0.05) difference between normaland abnormal crabs, as listed in Table 2.1. It is notable that the abnormal crabs generally had higherlevels than the normal crabs.Table 2.1. Effects of disease status on the metal concentrations (dry-weight basis). Significant (P

Department of Science, Information Technology, Innovation and the Artsd. Nickel.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 0.83 c 5.62 a 3.60 ab 2.72 bc 0.87Trip 2 1.04 bc 1.92 bc 1.63 bc 3.23 abe. Copper.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 241 cd 811 abc 1016 a 1237 a 205Trip 2 274 cd 968 a 198 d 390 bcdf. Zinc.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 95 d 111 cd 207 ab 183 bc 28Trip 2 265 ab 153 bc 276 a 208 abg. Arsenic.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 19.1 ab 15.5 b 25.0 ab 15.4 b 3.7Trip 2 21.1 ab 25.3 ab 26.0 a 28.9 ah. Selenium.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 5.58 bc 6.20 ab 7.82 a 7.11 ab 0.71Trip 2 4.81 c 5.25 bc 7.29 ab 7.44 abi. Silver.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 3.43 ab 6.18 a 2.81 ab 4.52 ab 1.40Trip 2 4.52 ab 5.29 ab 1.89 b 4.48 abj. Cadmium.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 1.03 c 3.91 ab 1.13 c 1.96 bc 0.86Trip 2 5.30 a 3.60 ab 1.62 bc 3.19 abck. Barium.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 1.00 cd 2.10 ab 0.81 cd 0.55 d 0.38Trip 2 2.56 a 1.93 abc 1.19 bcd 0.75 cdl. Mercury.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 0.371 b 0.195 cd 0.147 cd 0.343 b 0.042Trip 2 0.642 a 0.191 cd 0.078 d 0.232 cPage 36

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Aream. Lead.Bundaberg Fitzroy R. The Narrows Port area Avg. s.e.Trip 1 0.166 ab 0.143 ab 0.073 b 0.064 b 0.043Trip 2 0.068 b 0.076 b 0.231 a 0.129 abThe results for lead need to be taken with some caution, as there were only four values recorded as‘> the level of resolution’ (on a wet-weight basis). For the remaining 36 values which were ‘< 0.05’(hence taken as 0.025), conversion from wet to dry-weight basis introduced reasonable variation, asthe moisture content varied quite widely - between 55% (where 0.025 converts to 0.056) and 82%(converts to 0.139). So whilst this data set appears to contain the usual amounts of variability,amongst which the analysis found some significant differences, in reality these are based on fewactual values. The observed distribution of these values also results in significant regionaldifferences (see Table 2.4). The lead values measured as ‘> the level of resolution’ were one eachfor Bundaberg and the Fitzroy River (both on trip 1), and one each for the Narrows and Port area(both on trip 2). All four of these crabs were rated as ‘normal’.This was not a problem with any other metal. Chromium was the only other one with any ‘< thelevel of resolution’ values, and had 20 of these along with 20 actual values, so constituted a validanalysis.Overall (multivariate patterns)Relationships amongst the metals can be measured by the correlation coefficient, r. Table 2.3 liststhese values, which show the direction and relative strength of these relationships.Table 2.3. Correlation matrix for the metals. Significant (P < 0.05) correlations have an absolutevalue > 0.310, and are indicated in bold.Aluminium 1.000Chromium 0.090 1.000Iron 0.022 0.295 1.000Nickel -0.057 0.157 -0.032 1.000Copper 0.138 -0.011 -0.158 0.319 1.000Zinc -0.128 -0.083 0.371 0.144 0.160 1.000Arsenic -0.022 0.211 0.436 0.100 0.055 0.443 1.000Selenium 0.102 0.303 0.367 0.453 0.425 0.361 0.602Silver 0.236 0.182 0.233 0.234 0.683 0.180 0.250Cadmium 0.202 0.116 0.320 0.018 0.228 0.266 0.312Barium 0.135 0.458 0.260 -0.156 0.111 0.118 0.193Mercury -0.044 -0.059 -0.021 -0.262 -0.065 -0.217 0.032Lead -0.042 0.402 0.363 -0.062 -0.097 -0.024 0.080Aluminium Chromium Iron Nickel Copper Zinc ArsenicSelenium 1.000Silver 0.391 1.000Cadmium 0.071 0.686 1.000Barium -0.066 0.381 0.583 1.000Mercury -0.038 0.214 0.339 0.176 1.000Lead 0.121 0.107 -0.086 0.138 0.000 1.000Selenium Silver Cadmium Barium Mercury LeadPage 37

Department of Science, Information Technology, Innovation and the ArtsThe principal components overall spatial representation for the 13 metals (Table 2.2) is shown inFigure 2.1.43Bundy2Dimension 2 (23% variation)210-1Narrows2Port2Bundy1Fitzroy2Fitzroy1-2Narrows1Port1-4 -3 -2 -1 0 1 2 3Dimension 1 (32% variation)Figure 2.1. Movements of the overall location and trip principal component means inmultidimensional space. The Gladstone sites are indicated by the circles with the solid lines, and thereference sites by the squares with the dotted lines.Summarising by regionsThe means for the individual locations can be combined into ‘Gladstone’ (Port area and theNarrows) vs. ‘Reference’ (Bundaberg and Fitzroy River), as listed in Table 2.4. The pooledstandard error for each mean was calculated according to the usual sum-of-variances method (Raj1968), and these were used to conduct a t-test for the degree of difference between these regionalmeans. In each of the trips, three of the six significant differences showed a higher level for thereference sites.Page 38

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 2.4. Regional means for the metal concentrations (dry-weight basis). Within trips, significant(P < 0.05) differences are bolded.Trip 1 – Trip 2 –Reference Gladstone Sig. level Reference Gladstone Sig. levelAluminium 16.7 14.7 0.646 12.2 9.0 0.513Chromium 0.186 0.098 0.031 0.120 0.166 0.288Iron 273 188 0.057 187 418 0.000Nickel 3.22 3.16 0.938 1.48 2.43 0.306Copper 526 1127 0.004 621 294 0.139Zinc 103 195 0.002 209 242 0.275Arsenic 17.3 20.2 0.419 23.2 27.4 0.285Selenium 5.89 7.46 0.027 5.03 7.36 0.004Silver 4.80 3.67 0.401 4.91 3.18 0.251Cadmium 2.47 1.55 0.268 4.45 2.41 0.030Barium 1.55 0.68 0.021 2.25 0.97 0.003Mercury 0.283 0.245 0.349 0.416 0.155 0.000Lead 0.155 0.069 0.046 0.072 0.180 0.024Again, the means for lead are based on relatively few actual values, so should be considered withsome caution.ReferencesGenStat (2011). GenStat for Windows, Release 14.1. VSN International Ltd., Oxford.McCullagh, P. and Nelder, J. A. (1989). Generalized Linear Models (2 nd ed.). Chapman and Hall,London.Raj, D. (1968). Sampling theory. McGraw-Hill, New York.Page 39

Department of Science, Information Technology, Innovation and the ArtsAppendix BStatistical Report – Metals, Barramundi and Grinner1. STATISTICAL METHODSGeneral linear models (McCullagh and Nelder 1989) were used for the analyses of the metalconcentrations in fish (on a dry-weight basis), using GenStat (2011). These were done separately forthe gill and liver samples of the barramundi, and also for the muscle samples from the grinner. Eachfish was taken as an independent experimental unit. For samples reported as ‘below the detectionlimit’, the usual convention of substituting half the detection level was used. This was done for thewet-weight data (upon which the test had been performed), and then these values were corrected formoisture per cent back to a dry-weight basis (which were used for the analyses).Again, the deliberate sampling of both diseased and unaffected barramundi introduced furthercomplexity into the analyses. To adjust for this stratification, weighting factors were calculated as –Proportion (normal or diseased) in the whole sample /Proportion (normal or diseased) in the laboratory subsampleThese weighting factors were determined at the location by trip level, and then used to correctlyweight each of the respective observations (normal or diseased) in the analyses for trips bylocations. As none of the grinner were rated as diseased, no weightings were adopted for theiranalyses.For some metals, most of the observations were classified as LOR) observations. For these no valid analysis is possible – evennonparametric tests show no result when there are only a few (possibly-random) actual values. Forexample, the ‘most extreme’ regional comparison in Table 1.1 is two positives for Gladstone vs.zero for the reference sites, and this Chi-square test gives P = 0.15. For the metals which hadsufficient data to be assumed as continuous there were a few high values, however these data werenot notably skewed so no transformations were required.Table 1.1a. For the barramundi gill samples, those metals with few observations >LOR, listingtheir counts across sites and trips (blanks = zero).Chromium Nickel Silver LeadTRIP 1 - Awoonga DamUpper Boyne River 1Hamilton PointBundabergFitzroy RiverTRIP 2 - Awoonga DamUpper Boyne River 1Hamilton Point 1 1 1Bundaberg 1Fitzroy RiverPage 40

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 1.1b. For the barramundi liver samples, those metals with few observations >LOR, listingtheir counts across sites and trips (blanks = zero).Chromium Nickel Barium LeadTRIP 1 - Awoonga DamUpper Boyne RiverHamilton PointBundabergFitzroy RiverTRIP 2 - Awoonga DamUpper Boyne RiverHamilton Point 1 1Bundaberg 1Fitzroy RiverTable 1.1c. For the grinner muscle samples, those metals with few observations >LOR, listing theircounts across sites and trips (blanks = zero).Chromium Nickel Silver Cadmium Barium LeadTRIP 1 - Gladstone 2BundabergFitzroy River 1TRIP 2 - GladstoneBundaberg 2 2 1Fitzroy RiverFor barramundi, both length and weight was trailed separately as covariates. In general these werenegative or non-significant. The only significant positive fits were for length vs. Aluminium forliver (P = 0.045), and for Mercury for both gills and liver, where length was more highly significant(P < 0.001 for both tissues). It was concluded that this strong result for Mercury was evidence ofbio-accumulation, so length was retained as a covariate for this metal only. The analyses for allother metals (including Aluminium) used no covariate.Across the metals by tissues combinations of analyses, sites, trips and their interaction weresignificant (P < 0.05) far more often than expected by random chance. Hence there were provenpatterns in the data, and for full information (and consistency with the reporting of other data) themeans are listed on the two-way (sites by trips) basis.For barramundi, second analyses were also conducted for each variable including ‘disease status’ inthe model. This allowed the direct testing and reporting of the impact of disease status on theconcentrations of each metal.The univariate analyses of each of the metals are followed by a multivariate summarisation. Thisfirst considers the relationships amongst the metals (via a correlation matrix), and then adopts aprincipal components representation of all the data (taking the first two orthogonal dimensions) toobtain an overview as to which locations and times are most similar.Page 41

Department of Science, Information Technology, Innovation and the Arts2. RESULTS – BARRAMUNDI, GILLSDisease status – None of the metals in Table 2.1 showed a significant (P < 0.05) differencebetween normal and diseased fish, although mercury was very close. It is notable that the diseasedfish generally had higher levels than the normal fish.Table 2.1. Effects of disease status on the metal concentrations (dry-weight basis), for the gilltissues of barramundi.Normal DiseasedAverages.e.Significancelevel (P)Ratio (Diseased/ Normal)Aluminium 13.5 18.1 2.6 0.324 1.34Iron 189 374 80 0.125 1.98Copper 3.41 4.95 0.80 0.208 1.45Zinc 56.2 65.7 4.0 0.131 1.17Arsenic 1.73 3.34 0.61 0.098 1.93Selenium 2.90 3.59 0.36 0.243 1.24Cadmium 0.045 0.075 0.015 0.122 1.66Mercury 0.286 0.400 0.041 0.052 1.40Barium 1.64 1.26 0.58 0.615 0.77Location by trip meansThese are the results that are of primary interest, and are listed in Tables 2.2a to 2.2i. For eachmetal, significant-difference testing has been conducted amongst these means (locations by trips).Table 2.2a. Aluminium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 2.12 d 22.01 b 38.6 a 6.18 cd 6.08 cd 5.21Trip 2 17.23 bc 11.7 bc 12.78 bc 19.3 bca Means with a common superscript are not significantly different (P = 0.05).Table 2.2b. Iron.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 225 a 279 a 564 a 272 a 226 a 168Trip 2 522 a 264 a 163 a 184 aTable 2.2c. Copper.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 3.18 b 3.93 ab 8.14 a 3.89 ab 3.96 ab 1.70Trip 2 4.58 ab 4.67 ab 3.41 ab 2.82 bTable 2.2d. Zinc.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 54.2 a 61.6 a 72.8 a 65.1 a 60.2 a 8.6Trip 2 54.9 a 69.3 a 53.3 a 58.6 aPage 42

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 2.2e. Arsenic.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.74 cd 2.28 bc 1.35 bc 6.87 a 0.70 cd 1.24Trip 2 5.28 ab 0.47 d 4.28 abc 1.07 bcTable 2.2f. Selenium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 2.21 b 3.49 ab 3.76 ab 4.70 a 1.94 b 0.74Trip 2 4.55 a 2.58 ab 3.78 ab 2.49 bTable 2.2g. Cadmium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.124 a 0.091 ab 0.096 ab 0.071 ab 0.086 ab 0.031Trip 2 0.044 ab 0.058 ab 0.021 b 0.021 bTable 2.2h. Mercury.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.443 ab 0.393 ab 0.368 abc 0.144 c 0.361 abc 0.083Trip 2 0.484 a 0.302 abc 0.230 bc 0.503 aTable 2.2i. Barium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.74 a 2.84 a 0.47 a 0.25 a 0.43 a 1.09Trip 2 1.95 a 2.00 a 0.54 a 2.51 aOverall (multivariate patterns)Relationships amongst the metals can be measured by the correlation coefficient, r. Table 2.3 liststhese values, which show the direction and relative strength of these relationships.Table 2.3. Correlation matrix for the metals. Significant (P < 0.05) correlations have an absolutevalue > 0.294, and are indicated in bold.Aluminium 1.000Iron 0.202 1.000Copper 0.337 0.862 1.000Zinc 0.347 0.697 0.840 1.000Arsenic -0.129 0.574 0.263 0.206 1.000Selenium 0.107 0.793 0.663 0.525 0.798 1.000Cadmium -0.041 0.558 0.607 0.509 0.163 0.365 1.000Mercury 0.251 0.462 0.322 0.366 0.375 0.337 0.300 1.000Barium 0.117 -0.068 -0.027 0.109 -0.186 -0.038 -0.201 -0.144 1.00Aluminium Iron Copper Zinc Arsenic Selenium Cadmium Mercury BariumPage 43

Department of Science, Information Technology, Innovation and the ArtsThe relatively high degree of correlations amongst these nine metals means that the first twoprincipal components capture 62% of the total variation. This overall spatial representation is shownin Figure 2.1.21Awonga1Boyne1Boyne2Fitzroy2Bundy1Fitzroy1Dimension 2 (27% variation)0-1-2Hamilton2Bundy2-3Hamilton1-3 -2 -1 0 1 2 3 4 5Dimension 1 (35% variation)Figure 2.1. Movements of the overall location and trip principal component means inmultidimensional space for the barramundi gill samples. The Gladstone sites are indicated by thecircles with the solid lines, the reference sites by the squares with the dotted lines, and AwoongaDam (trip 1 only) by the star.Summarising by regionsThe means for the individual locations can be combined into ‘Gladstone’ (Hamilton Point andUpper Boyne River; notably does not include Awoonga Dam) vs. ‘Reference’ (Bundaberg andFitzroy River), as listed in Table 2.4. The pooled standard error for each mean was calculatedaccording to the usual sum-of-variances method (Raj 1968), and these were used to conduct a t-testfor the degree of difference between these regional means.Page 44

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 2.4. Regional means for the metal concentrations (dry-weight basis). Within trips, significant(P < 0.05) differences are bolded.Trip 1 – Trip 2 –Reference Gladstone Sig. level Reference Gladstone Sig. levelAluminium 30.3 6.1 0.000 14.5 16.0 0.767Iron 421 249 0.307 393 173 0.202Copper 6.04 3.92 0.220 4.62 3.11 0.387Zinc 67.2 62.7 0.600 62.1 55.9 0.482Arsenic 1.82 3.79 0.119 2.87 2.68 0.875Selenium 3.63 3.32 0.679 3.56 3.14 0.571Cadmium 0.094 0.078 0.625 0.051 0.021 0.351Mercury 0.381 0.252 0.125 0.393 0.366 0.751Barium 1.65 0.34 0.233 1.97 1.53 0.6863. RESULTS – BARRAMUNDI, LIVERDisease status – None of the metals in Table 3.1 showed a significant (P < 0.05) differencebetween normal and diseased fish. Silver and copper were close to significant, with diseased fishhaving lower levels of these metals than the normal fish.Table 3.1. Effects of disease status on the metal concentrations (dry-weight basis), for the livers ofbarramundi.Normal DiseasedAverages.e.Significancelevel (P)Ratio (Diseased/ Normal)Aluminium 2.26 3.18 0.53 0.489 1.41Iron 1420 1901 282 0.431 1.34Copper 44.8 18.5 10.2 0.065 0.41Zinc 67.9 66.0 7.7 0.701 0.97Arsenic 4.40 5.80 0.80 0.429 1.32Selenium 5.36 4.52 0.86 0.373 0.84Cadmium 0.106 0.137 0.019 0.562 1.28Mercury 0.597 0.624 0.145 0.843 1.05Silver 0.260 0.080 0.063 0.055 0.31Location by trip meansThese are the results that are of primary interest, and are listed in Tables 3.2a to 3.2i. Again,significant-difference testing has been conducted amongst the nine means for each metal.Table 3.2a. Aluminium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.57 c 3.92 ab 0.97 bc 2.96 abc 3.64 ab 0.99Trip 2 3.23 abc 1.25 bc 5.42 a 2.52 abca Means with a common superscript are not significantly different (P = 0.05).Page 45

Department of Science, Information Technology, Innovation and the ArtsTable 3.2b. Iron.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 663 c 1993 bc 1397 bc 3665 a 1899 bc 530Trip 2 845 bc 1092 bc 2297 ab 803 bcTable 3.2c. Copper.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 5.6 b 60.0 ab 31.5 ab 67.7 a 7.2 ab 23.6Trip 2 61.5 ab 8.2 ab 26.8 ab 12.6 abTable 3.2d. Zinc.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 32.5 c 67.7 bc 68.1 bc 120.6 a 59.0 c 14.0Trip 2 45.3 c 40.1 c 107.7 ab 35.3 cTable 3.2e. Arsenic.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 3.48 cd 7.75 bc 2.99 d 13.22 a 2.15 d 1.54Trip 2 4.11 cd 0.93 d 10.16 ab 2.10 dTable 3.2f. Selenium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 2.32 bc 5.81 abc 3.57 bc 10.14 a 2.80 bc 1.66Trip 2 6.96 ab 1.85 c 7.04 ab 2.88 bcTable 3.2g. Cadmium.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.027 d 0.147 bc 0.089 cd 0.275 a 0.117 bc 0.036Trip 2 0.050 cd 0.086 cd 0.182 ab 0.056 cdTable 3.2h. Mercury.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.659 a 1.036 a 0.460 a 0.619 a 0.939 a 0.273Trip 2 0.428 a 0.365 a 0.596 a 0.575 aTable 3.2i. Silver.AwoongaDam BundabergFitzroyRiverHamiltonPointUpperBoyne R. Avg. s.e.Trip 1 0.057 b 0.229 ab 0.098 ab 0.201 ab 0.080 ab 0.136Trip 2 0.433 a 0.058 b 0.105 ab 0.196 abPage 46

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaOverall (multivariate patterns)Table 3.3 lists the correlation coefficients, r, which show the direction and relative strength of therelationships.Table 3.3. Correlation matrix for the metals. Significant (P < 0.05) correlations have an absolutevalue > 0.294, and are indicated in bold.Aluminium 1.000Iron 0.639 1.000Copper 0.004 0.201 1.000Zinc 0.610 0.724 0.413 1.000Arsenic 0.375 0.634 0.307 0.656 1.000Selenium 0.346 0.569 0.791 0.659 0.670 1.000Cadmium 0.668 0.889 0.228 0.840 0.641 0.587 1.000Mercury 0.796 0.725 0.074 0.571 0.305 0.330 0.730Silver -0.058 -0.015 0.890 0.176 0.075 0.701 0.040Aluminium Iron Copper Zinc Arsenic Selenium CadmiumMercury 1.000Silver -0.028 1.000Mercury SilverThe relatively high degree of correlations amongst these nine metals means that the first twoprincipal components capture 80% of the total variation. This overall spatial representation is shownin Figure 3.1.Page 47

Department of Science, Information Technology, Innovation and the Arts2Boyne1Dimension 2 (19% variation)10-1-2Awonga1Fitzroy2Fitzroy1Boyne2Bundy1Hamilton2Hamilton1-3Bundy2-3 -2 -1 0 1 2 3 4 5Dimension 1 (61% variation)Figure 3.1. Movements of the overall location and trip principal component means inmultidimensional space for the barramundi liver samples. The Gladstone sites are indicated by thecircles with the solid lines, the reference sites by the squares with the dotted lines, and AwoongaDam (trip 1 only) by the star.Summarising by regionsAgain, the means for the individual locations can be combined into ‘Gladstone’ (Hamilton Pointand Upper Boyne River) vs. ‘Reference’ (Bundaberg and Fitzroy River), as listed in Table 3.4. Thedegree of difference between these regional means was determined with t-test.Page 48

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaTable 3.4. Regional means for the metal concentrations (dry-weight basis). Within trips, significant(P < 0.05) differences are bolded.Trip 1 – Trip 2 –Reference Gladstone Sig. level Reference Gladstone Sig. levelAluminium 2.44 3.30 0.390 2.24 3.97 0.093Iron 1695 2782 0.046 969 1550 0.284Copper 45.7 37.4 0.726 34.8 19.7 0.529Zinc 67.9 89.8 0.124 42.7 71.5 0.048Arsenic 5.37 7.68 0.140 2.52 6.13 0.026Selenium 4.69 6.47 0.287 4.41 4.96 0.744Cadmium 0.118 0.196 0.035 0.068 0.119 0.165Mercury 0.748 0.779 0.911 0.396 0.586 0.482Silver 0.163 0.141 0.867 0.245 0.151 0.4964. RESULTS – GRINNER, MUSCLELocation by trip meansThese are listed in Tables 4.1a to 4.1g. For each metal, significant-difference testing has beenconducted amongst these means (locations by trips). Due to the unbalanced nature of the samplingfor grinner, with only one Gladstone and one reference site for trip 1, and one Gladstone and tworeference sites for trip 2, no ‘regional comparisons’ have been done. Any such differences can bejudged by the significant differences between the sites (within each trip). It is notable that for thefirst four metals the reference site of Bundaberg (which was only sampled on trip 2) hadsignificantly higher levels than all other sites, across both trips. Within trips, there were no instanceswhere the levels at Gladstone were significantly higher than the levels at either of the referencesites.Table 4.1a. Aluminium.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 3.8 b 1.9 b 3.9Trip 2 23.2 a 5.1 b 5.4 ba,b means with a common superscript are not significantly different (P = 0.05).Table 4.1b. Iron.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 12.9 b 8.8 b 6.0Trip 2 44.5 a 8.6 b 13.7 bTable 4.1c. Copper.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 0.89 b 0.68 b 0.15Trip 2 1.81 a 0.92 b 1.07 bTable 4.1d. Zinc.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 21.9 b 17.8 b 3.1Trip 2 36.8 a 20.9 b 25.4 bPage 49

Department of Science, Information Technology, Innovation and the ArtsTable 4.1e. Arsenic.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 6.83 ab 10.27 a 1.88Trip 2 7.63 ab 3.34 b 7.49 abTable 4.1f. Selenium.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 2.35 ab 2.41 a 0.13Trip 2 2.03 b 2.44 a 2.37 abTable 4.1g. Mercury.Bundaberg Fitzroy River Gladstone Harbour Average s.e.Trip 1 0.147 a 0.189 a 0.050Trip 2 0.121 a 0.225 a 0.250 aOverall (multivariate patterns)Relationships amongst the metals are measured by the correlation coefficient, r, as listed in Table4.2.Table 4.2. Correlation matrix for the metals. Significant (P < 0.05) correlations have an absolutevalue > 0.398, and are indicated in bold.Aluminium 1.000Iron 0.665 1.000Copper 0.735 0.693 1.000Zinc 0.491 0.785 0.811 1.000Arsenic -0.036 0.180 0.106 0.145 1.000Selenium -0.192 -0.169 -0.203 -0.282 0.014 1.000Mercury -0.285 -0.355 -0.347 -0.351 -0.260 0.126 1.000Aluminium Iron Copper Zinc Arsenic Selenium MercuryGiven there are only seven metals and that some high correlations exist amongst these, the first twoprincipal components capture 92% of the total variation. This overall spatial representation is shownFigure 4.1. As was apparent in Table 4.1, the major difference appears to be between Bundabergand the other sites.Page 50

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone Area2Fitzroy21Dimension 2 (16% variation)0Gladstone2Fitzroy1Bundy2-1Gladstone1-2 -1 0 1 2 3 4 5Dimension 1 (76% variation)Figure 4.1. Movements of the overall location and trip principal component means inmultidimensional space. Gladstone is indicated by the circles with the solid line, and the referencesites by the squares with the dotted line.ReferencesGenStat (2011). GenStat for Windows, Release 14.1. VSN International Ltd., Oxford.McCullagh, P. and Nelder, J. A. (1989). Generalized Linear Models (2 nd ed.). Chapman and Hall,London.Raj, D. (1968). Sampling theory. McGraw-Hill, New York.Page 51

Department of Science, Information Technology, Innovation and the ArtsAppendix CSensitivity analysis1. Pb, crabsAs discussed in the revised stats report, there were too few ‘actual’ values (four) to have any realconfidence in these differences, so either say this, or (preferably) omit these means altogether &state why. Note that for the fish there were a number of metals not analysed for this reason.2. Cr, crabsGiven that 20 (out of 40) values were 0.10 for all tests).Table 1 – Original ResultsBundaberg Fitzroy R. The Narrows Port area AverageStandard ErrorTrip 1 0.136 ab 0.235 a 0.110 b 0.086 b 0.041Trip 2 0.116 ab 0.123 ab 0.190 ab 0.142 aba,b Means with a common superscript are not significantly different (P = 0.05).Table 2 – Changing all

Contaminants in Tissue of Fish and Crabs Collected in the Gladstone AreaThe original differences were that (on trip one) Fitzroy was significantly higher than both theNarrows and the Port area. The first scenario loses one of these (Fitzroy is no longer different tothe Narrows). The second scenario, whilst having a higher standard error, picks up an extradifference – here Bundaberg is also significantly lower than the Fitzroy mean. The basic result, thatFitzroy on trip one is notably higher than the other sites, remains the same in all these scenarios.For the overall regional means, the original results had higher average Cr on trip one for thereference sites vs. the Gladstone region. The comparable results are now –P-value Reference mean Cr Gladstone mean Cr PercentageOriginal analyses 0.031 0.186 0.098 190Scenario 1 -

Appendix E – Time line BarraBy Johnny Mitchell;Johnny Mitchell's Fishing Charters17 November 2011My name is Johnny Mitchell, and the following article 'Time Line Barra ' was constructedduring my own time. My entire life has been based in or around the ocean and the land. Myexperiences are well spread, having spent time as part of a blue water charter fishingoperation and as an estuary and coastal commercial fisherman. I have been a professionalbarramundi fishing guide in Gladstone for the last five years and a small scale film maker. Ihave completed two films on barramundi, one on mud crabs and one on Spanish mackerel.My number one study fish is the barramundi, a species which I have observed carefully andthoroughly in the local harbour for over two decades and for over 14 years in Lake Awoonga.For about the last 14 years the species has taken up a huge portion of my life, with round theclock study sessions in both fresh and salt to refine and advance my knowledge with thespecies. I have an observant nature which has allowed the opportunity to record informationabout the time line of events from the Awoonga spill of 2010. I enjoy micro scoping any fishysituation.Having lived in Gladstone for almost 37 years and extensively fished Lake Awoonga forbarramundi during the last 14.5 years, the December 12, 2010 spill over event brought with itboth positive and negative vibes. From watching some fish die from the free fall, to ponderingthe success of the mature fish transition there certainly was speculation as to the outcome ofthe event.For the last 20 years, and especially the last 14, barramundi have become my number onestudy priority, personally spending between 250-350 days a year on the water and up to 3sessions a day between salt and fresh water. From true wild fisheries in the CentralQueensland region to the land locked waters of Awoonga I have relished in the challengeand variations between salt water and freshwater barramundi; both an a personal level andas a professional fishing guide.My experiences between the two fisheries are vast.My goal from the flood event was to follow the progression of the lake fish into the saltwaterand to document their movements far and wide, why, because now we had two fisheriesblending as one, 20 years of study blending together in a new chapter on barra- important tome. I witnessed their change in behaviour, and watched their state of health throughout; bothgood and bad.Below is a timeline of information and photographs covering almost 12 months of events andobservations during my time on the water from Rockhampton to Point Richards but usinghome waters of Gladstone Harbour as the central point of study.138

December, 2010- The OverflowPicture A. Shows a healthy fish from Lake Awoonga, photographed during the start of theflood event in Lake Awoonga- Mint Condition!December 12, 2010, the much anticipated flood event arrives and with it many barramundiexit Lake Awoonga. Most of these fish were in the 90-130cm bracket and approximately 4 -12yrs old. Thousands of fish went over in the first week or so. Around 30 fish in 30 secondswas the highest count I witnessed in my many visits to the wall lookout. It was rare to sit for aminute or two and not see a fish go over; it was usually multiple fish per minute. At just 1 fishper minute, the rate is equal to around 1440 a day. That's over 10,000 in a week. The rate offish escaping varied, and was basically impossible to record once the water reached the 4m+mark over the wall as the water was a raging white water torrent, but the rate of escapeslowed considerably as the water level subsided. Printed figures of only 10,000 fish escapingthe lake were neither accurate nor supported- those making the calls having reason to downplay the situation rather than allowing reality to come forth. Even the modern figures of20,000 - 30,000 seem unsupported considering the variables that were in play. Night timecounts were impossible with no lights to assist and considering barramundi are instinctivelyactive at night time it stands to reason to question what actually did escape in the dead ofnight, especially when the various lunar movements and wind change trigger periods aremicroscoped; the environmental stimulants that often dictate barra movement. Threeprofessional fishing guides worked Awoonga daily up until the spill event. Two have nowmoved on to other work ( myself included), the third has not landed a barra from the lake innearly 12 months- a huge portion of Awoonga's adult barra have vanished. Personally I haveseen only two documented barra captures from Awoonga in the last 11 months since the spillevent. The remaining fish are laying low.Around 1000 dead fish were counted along the shores of the Boyne from the dam wall andout to the harbour- obviously fish that died from wounds caused by the fall- some had headwounds, others had lost huge amounts of scales and skin and had red areas where therewas no slime, skin or scales. Dead fish up to 135cm were found and measured by my close139

friends. Rapidly rising flood waters made a proper count impossible as dead and dying fishended up 20 - 50km from the dam wall and out into the harbour and open ocean in no time.Dying barra were picked up on the surface in blue water locations.Within just one week of the initial spill over I witnessed many new fresh and lively arrivals(escapee lake fish), detected with the use of high definition depth sounder in areas within theharbour up to 43km from the dam wall. (over 20 km from Boyne River mouth) These fishseemed to rest in small schools on the bottom in deeper water- possibly adjusting to thechange. Whilst none of these new arrivals were captured on a lure in this early period, theirpresence was noted. (February's information will highlight their immense behaviouralchange)Photo B. Awoonga Dam Overflow event 2010.January, February, March- Mystery and Intrigue.100% capacity was reached around December 12, 2010, and 140.25% and 4.094m abovespillway was reached on 28th December ( GAWB INFO) and the dam was still flowing overon Sunday 23rd January, 2011, about 0.42m above. (About 40 days of spill and counting). Itwasn't until late June 2011 that Awoonga actually stopped spilling over the wall. DuringJanuary there were mixed reports from anglers of these escapees starting to feed and strikelures in the Boyne River system. First reports shared stories of weak fish that didn't fight veryhard and many that floated away and died after release obviously showing signs of ill healthfrom the overflow ordeal. Anglers stopped fishing for a period and when the fish showedgreater health signs so began a period in the Pike's Crossing region (upper Boyne) known asthe 'Turkey Shoot of 2011' where literally thousands of barra were captured by hundreds ofanglers in narrow waters. (boats, kayaks and off the shore) This radical fishing below thedam wall included the whole month of February and trickled into March until a large exodusof fish that coincided with another mini flood and water rise from Awoonga during very earlyMarch.140

In the mid to lower saltwater reaches of the Boyne, massive schools exceeding a thousandfish per school were evident (large barra schools found with modern side imaging sonartechnology) with anglers catching double figure catches in a fishing session. It was evidentthat a leap frog of schools was happening; big numbers would exit the Boyne only to bereplaced by schools from higher up the river. In the harbour and waters approximately 50kmto the north and south the escapee barra also started to bite with a vengeance ( lateJanuary) and typical daily catch rates for us doubled and tripled that of typical years withmostly wild stocks available. Sessions of 12-15 escapee barra averaging around 100cmwere easily possible in just 1-2 hrs fishing. The eating quality of these harbour fish thatescaped the lake and the river was first class with reduced fat levels evident in the flesh. Thetaste was so good that many locals commented- softer flesh than wild barra, flavour high andon par with any reef fish. It was noticeable by now (March) that many of the fish were startingto thin down (lose condition). Almost every escapee fish had some sign of damage, whetherit be scale loss, open cuts (healing) or damage to the jaw or head; some wounds weresevere although advanced healing was obvious, others had very insignificant wounds, butnoticeable if you quickly searched. It was like nearly every fish had a reminder mark from theordeal (see images below).Figure C Nearly every fish had a reminder mark from the ordeal.Figure D. A skinny fish from post flood event at Boyne Mouth.141

To make note – condition loss was expected since the lifestyle lived by the barra in Awoongawas fairly easy, food on tap, little water flow to contend with). To contrast the negative therewas some very good looking barramundi being caught by anglers – everything looked fineand positive at this stage, healing, hungry fish being very encouraging.Figure E. Some very good looking barramundi.To make further note, turtle deaths became obvious in the Harbour in early March as well as anescalating commercial barramundi net fishing effort taking place at the mouth of the Boyne River justoutside the river closure zone.Figure F. Initially the commercial sector was blamed for the turtle deaths but thefollowing months drew focus to a much bigger picture.During late March escapee fish captured in the harbour that were kept for the table had started roedevelopment for a possible late spawning due to high water temperatures in March. Barramundiframes including these roe developments were handed to Bill Sawynok for further analysis. Theweather cooled immediately after these findings and no further roe developments were found. (Thecoming three months Nov, Dec, 2010 and Jan, 2011 will offer better information on escapee barraspawn possibilities).142

During late March escapee fish captured in the harbour that were kept for the table hadstarted roe development for a possible late spawning due to high water temperatures inMarch. Barramundi frames including these roe developments were handed to Bill Sawynokfor further analysis. The weather cooled immediately after these findings and no further roedevelopments were found. (The coming three months Nov, Dec, 2010 and Jan, 2011 willoffer better information on escapee barra spawn possibilities).April- BoomingBy the 1st April I had personally caught tagged barramundi over 51km from the AwoongaDam wall (south to Bird Island, Rodd's Bay) Tag data showed place of origin was from LakeAwoonga. (See image A 1).Charter and personal trips produced numerous saltwatercaught, tagged barra, originally from Awoonga. Vastschools of barra had moved away from the Boyne and welure caught hundreds of fish in this 50km southern zone andmy predictions had some of these fish ending up in theGreat Sandy Straights, well to the south. ( currently14/11/2011 there has been recorded tag data fromAwoonga barramundi captured in the Burnett River,Bundaberg, approximately 174 km south) so the trend forwide spread movement is frimly documented.Figure G Nice BarraAt this stage (late April) during the time line it was lookingfine, barramundi were being caught from the freshwaterreaches of the Boyne, the harbour and now in numerousreaches of estuary systems over big areas. Somesensational lure fishing was still being had in the lowerBoyne River whilst commercial netters still worked outsidethe mouth. By mid April it was obvious that barramundiwere scattered over large distances of coast line creatingimpressive fisheries for both rec and commercial anglers.I joined Dr Ben Diggles and Kurt Hutchbyfrom GAWB on a science trip below the damwall to catch by line as many barra aspossible and check each one for health anddisease.Figure H. There were numerous issuesnoted by Dr Ben Diggles143

May/June- FinsMy initial observations were of anal fins andthe tips of barra tails eroding away or beingeaten away during the month of May.Photographs of this occurrence wereimmediately sent to both Bill Sawynok andDr Ben Diggles for comment.Also a few barra with caligid copepods aswell as a form of mucus or a monogeneanparasites were captured- photos also sent toDr B. Diggles for identification.Figure I. Tail Fins of Barramundi and barramundi caligid copepods144

Figure J. A definite 'thinning out' of barra body condition became more common place by theend of May and by June it was typical to catch escapee barra that looked nothing like theirprime when in Lake AwoongaFigure K. This image shows a mint conditioned Awoonga barramundi.145

Thes e escapee were now also in aworse condition to that of typicalwild stock saltwater barramundi thathave always been known as a 'slim,fit fish'.The tide was definitely turning withthe barramundi beginning to looktatty and no longer carrying that'prime brilliant look' that barra areknow for.By late June most anglers hadreduced their fishing efforts unableto continue to catch barra, day inand day out, mostly governed bywater temperature- even at theGladstone Power Station warmwater outlet it was easy to see lessanglers and less fish being caught,however they were still in strongnumbers but required moreknowledge to catch. Wesuccessfully captured barra allwinter, in great numbers. Guidesexcel while the average anglerstruggles. In this June period it wasnoted that in the 'best paddocks' inthe harbour where forage food wasabundant, the escapee lake fishwere doing the best by retaining themost condition while in other areaswith far less food it was clear to seethe barra were worse off in healthand condition.Figure L. This image shows a typically fit, slim,saltwater wild stock barramundiOne phenomenon I witnessedduring late May was an intensedensity of barramundi north ofGladstone toward Ramsay'scrossing. Over a distance of 2.4nautical miles, barramundicontinually showed on the sounderscreen, rarely a moment passingwith not a fish on the screen. I jigfished barramundi in this deeperarea- no mistake, they werebarramundi. To date, 14/11/2011 theschool has departed; the nomadicways of the barramundi, along withchanging seasons have moved thefish away.146

Figure M. Prime conditioned barramundi from AwoongaJuly/ August- Getting UglyIt was only in late July when I saw the first signs of eye problems, ( see A 6), and in thosefish that had eye problems, it was also clearly visible that these fish had busted throughcommercial nets with tight monofilament strangle marks across the forhead and/or snout.147

Figure N. Barramundi showing net marks and cloudy eyesThis may have just been coincidence as we also captured escapee Awoonga barra, for manymonths, some with net marks on the snout above the eyes showing no eye issues at all.The eye problems seemed to be highlighted by a milky colouration – this obviously didn'thinder their ability strike a lure.Figure O. Another barra caught on lure with cloudy eyes148

Barra have been caught on lines for decades, across the country with occassional eyedamage or in worst case- no eyes). Their ability to hunt by using vibration and sound tolocate food source is a key to their survival without any eyes....... in those individual cases,unrelated to this current eye problem in some escapee Awoonga barramundi.Eye damage wasn't common in our line caught specimens, but may have been morecommon than our catch rates dictated. Ill or suffering fish may not have been eating asmuch, if at all. What was again obvious and common was skinny fish, tattered fins and theloss of that glossy gleaming iconic barra glint. The fish were more dull and the body slimeseemed more whitish rather than opaque.Figure P. Barramundi with blemishes on skin and fins described below.Blemishes on the skin and fins were more frequent and many captures of escapee lake fishhad some kind of sign of deteriorating condition/health. At this point in time one of my closefishing companions and I discussed if these fish would ever regain their condition again orcontinue downhill until death. For every story of 'doom and gloom escapee barra caught' wealso had stories of escapee barra from other areas within the harbour that were in way bettercondition. During July and early August I did manage to dive numerous times and film in thelower Boyne River in winter waters ( cleaner) to observe hundreds of barramundi with manyof these fish having hollow stomachs.( concave bellies) At no stage during diving orcharter/personal barra fishing in the lower Boyne ( mouth to 1st bridge) did I notice thepoorer quality fish found from areas like the mid Calliope River, however in the upperreaches of the Boyne River's tidal extremities there were multiple cases of dead barra, sickbarra, and dying barra in August.149

September- Deaths and SicknessFigure Q. Dead barramundiIn September in the upper tidal reaches of the Boyne there were dead barra, dying barra,dead bream, dying bream, dead mullet, dying mullet, dead catfish, dying catfish as well as adying King Threadfin salmon noticed all on the same day. For a succession of days most ofthese species were again seen struggling, or dead. ( King salmon spotted only oncecoveredin a fungus or mucus or something) My observations labelled that section of wateras the warmest in the region, which was warmer than the hot water outlet in the CalliopeRiver. From memory and notes it cloked 80 degrees F which was up to 8 degrees F warmerthan other areas. It was common to see discoloured barramundi from this region, their headsand bodies covered in a white growth as well as green moss/alge growing on their bodies inplaces. I was still guiding clients in the area until the Harbour Closure mid September. I wasglad to stop fishing in there as touching the fish was a great concern to me and for thewelfare of my clients. We never captured a fish in that area that I considered 100% healthynor in good condition. Some were rakey, skinny, tattered, rough looking barramundi.150

Figure R. 135 cm barra described below.Included was a massive 135cm barramundi with white eyes, torn fins and off coloured milkybody slime that also carried an odour. Surprisingly, this was the heaviest and fattest fish I'dseen all year, even though very unhealthy.In the mid Calliope River, fish were getting skinnier and even large fish were found floating oron the river bank, dead.Figure S. 116 cm and 119 cm barra caught in the Calliope RiverOn one charter on 10 September, 2011 my clients caught a 116cm escapee Awoongabarramundi and a 119cm wild saltwater barra from the mid Calliope River.This leads to my most valid point for the whole entire time line- not once throughout thewhole ordeal have we caught a wild saltwater barramundi that has shown any sign of illness,infection or health issue. We catch escapee barra right beside wild salt water barra in certainlocations and even with deteriorating escapee barra by the droves, we still have perfectlymoulded, perfectly coloured, vivid, shiney, strong, healthy wild barramundi. They look likethey are in mint condition which leads me to my confidence point- this issue of health isn'taffecting the true wild barramundi stocks, only the escapee barra from Awoonga.151

There is numerous distinguishing features between the escapee Awoonga barramundi andtrue wild barra from local waters. All of these differences can be detected by the eye.Examples of this are pectoral fin colour, body colour, head shape, head features, bodyshape, right down to body density- the list goes on but these fish can be picked by a trainedexpert eye. I have spent over a decade looking at them closely- I can pick the differencesvery easily.Figure T. The above image shows 5 Gladstone caught barra - 2 true wild barramundi and 3escapee Awoonga barramundi, one showing distinct scale loss from the free fall down thewall- trained eyes can pick the difference.OctoberCommercial effort increased in the local area once again prior to the seasonal closure onNovember 1st, 2011.The escapee fish were still in bad shape and obviously the commercial sector netted more illfish than could be captured on a line due to the fact that very ill fish with bad health or closerto death don't feed, therefore not striking lures. The ugliest barra with the worst eyeinjuries/problems were either found dead, dying or captured in nets, not by line anglers.Many commercial catches were rejected by buyers due to unhealthy fish with red lesions andeye problems; fish I wouldn't dare eat.152

NovemberBarramundi season closes in the salt water on Nov 1st. Commercial effort stops and recangling slows so information about fish health is reduced although reports from the uppertidal section of the Boyne highlights fatter fish and no signs of illness; a contrast to that ofSeptember.( data from a commercial fisher in Bundaberg highlights barra of ill health bearingtags from Awoonga). During 4 personal, local fishing expeditions in late October and veryearly November we managed to catch 31 barramundi averaging 97cm. The catch wasdominated by saltwater barramundi, 24 wild saltys to 7 escapee lake barra. Of the 24 wildsalty barramundi none had any signs of ill health ( as expected) whilst 4 of the 7 escapeelake barramundi had tiny red lesions but were mostly of a good strong condition.153

Department of Agriculture, Fisheries and ForestryCall: 13 25 23 or +61 7 3404 6999Visit: www.daff.qld.gov.au154