Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorDavey, Kenneth Richard-
dc.contributor.advisorLewis, David Milton-
dc.contributor.authorHathurusingha Arachchige, Priyantha Indrajith-
dc.description.abstractFish farming with Recirculating Aquaculture Systems (RAS) is becoming widespread to fill the demand gap due to diminishing wild caught sea foods. Barramundi fish has a high demand as a premium Australian seafood, and is grown as an RAS farmed-fish. However, the accumulation of ‘earthy’ or ‘muddy’ off-flavours due to taint accumulation as geosmin (GSM) or 2-methylisoborneol (MIB) in the fish-flesh of is a major concern. Inconsistent quality of farmed barramundi has been identified as a major issue in buyer resistance. Established predictive models for chemical taint in fish-flesh have been based on steady-state assumptions. However, it was thought debatable as to whether a steady-state assumption could be upheld i.e. there was no evidence that the net chemicals exchange is zero across the fish body and RAS water phase. Against this background, an original, new and quantitative model that predicts the time dependent concentration of taste-taint chemicals as GSM and MIB in harvested fish-flesh was developed (Hathurusingha & Davey, 2013; Hathurusingha & Davey, 2014; Davey & Hathurusingha, 2014). This model is based on conservation of mass and energy, and thermodynamic processes established in (bio)chemical engineering with chemical uptake and elimination routes into and from the fish considered. The model was simulated for two RAS species, barramundi (Lates calcarifer) and rainbow trout (Onchorhynchus mykiss) with independent data (n ≥ 14) and showed good agreement with experimental observations. A major benefit of this new model is that simulations can be used to investigate a range of growth protocols in RAS farming to minimize taint in fish-flesh. An advantage is that it can readily be simulated in standard spread-sheeting tools by users with a range of sophistication. Extensive experimental testing of the new model was carried out in both pilot- and commercial-scale plants using low concentrations (≤ 10 mg L⁻¹) of hydrogen peroxide (H₂O₂) as a benign biocide to limit natural occurring taste-taint chemicals in the RAS growth water, and subsequently into the fish-flesh. A dedicated methodology and new dosing apparatus (ProMinent Fluid Control Pty Ltd, Germany) for controlled H₂O₂ dosing was developed. The analyses of taste-taint chemicals as GSM and MIB in water and fish-flesh was carried out with Solid-Phase Micro-Extraction (SPME) followed by Gas Chromatography Mass spectroscopy (GC-MS) (skills training was obtained at both the University of Laval and University of Waterloo, Canada). Preliminary investigations with a low concentration of H₂O₂ (5 mg L⁻¹) in pilot-scale (2,500 L) studies with barramundi fish demonstrated its potential to mitigate development of GSM and MIB in RAS water. It was found that controlled dosing of low concentrations of H₂O₂ did not impact the pH level in growth waters and was not detrimental to the health and well-being of the fish as fingerlings (0.01 kg) and until harvest at 240 days (0.8 kg). Additional benefits of H₂O₂ as benign biocide include a fish product of whiter colour, an increased dissolved oxygen concentration (Cₒₓ) in the growth water, a reduction in the number of gill flukes, and improved particles distribution with increased C:N ratio, and; improved availability of organic carbon in the growth water. Based on these preliminary investigations H₂O₂ was ‘optimised’ at a (low) concentration of 2.5 mg L⁻¹ as a benign biocide. This was investigated in commercial-scale studies (conducted at Barra Fresh Farm, South Australia) for a typical growth of 240 day for barramundi as the selected RAS fish. The emerging risk methodology of Davey and co-workers (e.g. Chandrakash et al., 2015) was applied for the first time to investigate quantitatively the impact of naturally occurring fluctuations in taste-taint chemicals in the RAS water and their accumulation in the fish-flesh. This predictive approach was justified because of the prohibitively expensive time and analytical costs that experimental studies would have necessitated. A Refined Monte Carlo (with Latin Hypercube) simulation of GSM and MIB in the growth water (Cᴡ), water temperature (T) and growth time (t) was used to simulate typical RAS farmed barramundi. It was found in RAS farming of barramundi it would be expected some 10.10 % of all 240 day harvests, averaged over the long term, would result in fish with taste-taint as GSM above the desired consumer rejection threshold concentration (0.74 μg kg⁻¹) due to natural fluctuations in an uncontrolled RAS environment. For MIB this predicted failure rate was 10.56 % (Hathurusingha & Davey, 2016). The vulnerability to taste-taint failure as GSM and MIB was shown to be principally controlled by the time to fish harvest, and to a lesser extent by concentration and fluctuation of these taint chemicals in the RAS water. This work was of practical benefit because growth time can be readily controlled by farmers. The methodology appears generalizable and therefore is applicable to a range of RAS farmed fish (and possible crustaceans e.g. prawns- Macrobrachium sp.). In extensive commercial-scale RAS studies with barramundi and controlled H2O2 dosing, fish grown from fingerlings to harvest at 240 day was investigated. This was to observe an entire production cycle. Results from a H₂O₂ ‘treated’ growth tank (30,000 L) were compared directly with those obtained from an identical ‘control’ tank (30,000 L). Increased organic matter (three (3) to four (4) times pilot-scale findings) reduced H₂O₂ efficacy through inhibiting generation of reactive oxygen species (ROSs). This is thought to be a consequence of the need to scale (48 times volume) the pilot-scale studies for in-tank mixing. Analyses of fish-flesh (n ≥ 167) showed (moderate) predicted exponential correlation between taste-taint concentrations in the fish-flesh and the growth-mass of the fish for both GSM and MIB as predicted. In addition, the research findings highlighted that accumulation of taste-taint compounds was mainly governed by the combined effect of mass of the fish (mᵳ) and taste-taint concentrations in the growth water (Cᴡ). Comparisons between the model predictions and experimental observations showed good agreement over the range of low taste-taint concentration (0 to 2, μg kg⁻¹), especially below the consumer rejection threshold (~ 0.7 μg kg⁻¹). However, a minor anomaly was an over-prediction for greater concentrations (2 to 11, μg kg⁻¹). Current predictions are therefore conservative or ‘safe’ by about 20 %. Possible reasons for over prediction might be attributed to rapid fluctuation of taste-taint concentration in growth water with growth time and different (exponential) growth constants shown by larger and smaller fish, and; errors in obtaining representative samples from fish-flesh. Model predictions and experiments further highlighted that the new model could be meaningfully applied to RAS systems with lower variations and/or lower taste-taint concentrations in RAS growth water. These theoretical and experimental results are the first for RAS farmed fish covering an entire production period to harvest. Approval for this research was gained from both The University of Adelaide Animal Ethics Committee Science and, Australian Pesticides and Veterinary Medicines Authority (see Appendices F and G). Research findings will be of immediate benefit to RAS farmers, fish processors and risk analysts in foods processing.en
dc.subject2-Methylisoborneol (MIB)en
dc.subjectGeosmin (GSM)en
dc.subjectBarramundi (Lates calcarifer)en
dc.subjectRecirculating Aquaculture System (RAS)en
dc.subjectexperimental studiesen
dc.subjectmodel validationen
dc.titlePredictive modelling and experimental studies on taste-taint as geosmin (GSM) and 2-methylisoborneol (MIB) in farmed barramundi (Lates calcarifer)en
dc.contributor.schoolSchool of Chemical Engineeringen
dc.provenanceThis electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at:
dc.description.dissertationThesis (Ph.D.) -- University of Adelaide, School of Chemical Engineering, 2016.en
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
01front.pdf413.19 kBAdobe PDFView/Open
02whole.pdf5.92 MBAdobe PDFView/Open
PermissionsLibrary staff access only202.28 kBAdobe PDFView/Open
RestrictedLibrary staff access only5.78 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.