To keep up with ever increasing seafood demand, aquaculture in the US and abroad must expand beyond traditional coastal culture methods and dramatically increase production. A major bottleneck experienced when bringing aquaculture onshore is the accumulation of the off-flavor compounds geosmin (trans-1, 10-dimethyl-trans-9-decalol) and 2-methylisoborneol (MIB)—particularly in recirculating aquaculture systems (RAS). These terpenoids give a “muddy” flavor to farmed fish, especially high value and popular fish like Atlantic salmon (Salmo salar). Knowledge regarding the rates at which these compounds are taken up by fish muscle, eliminated, and biologically transformed remains unestablished. Various technologies have been developed for commercial scale removal of off-flavor compounds and other applications. No comprehensive models exist to predict the bioaccumulation or elimination of off-flavor compounds in RAS with these technologies in mind, or to assess the efficacy of the technologies themselves. This research attempts to fill that gap with a model which not only quantifies the mass transport of off-flavor in RAS, but allows for the objective comparison of different remediation and removal approaches.
IMET’s research has shown that UV/TiO2 photocatalysis is moderately more effective (80% reduction in salmon GSM concentration in ng/kg over 10 days), than direct application of H2O2 (67% reduction), or the control (57% reduction), in freshwater RAS. Indirect ammonia electro-oxidation showed promising removal in saltwater RAS with 99% removal of GSM over 100 hours compared to 28% from a control system. Lipid extraction (LE) also demonstrated effective removal with >95% reduction in ng/L of GSM on a system scale water only test when compared to 26% reduction from the control.
To further understand and get a direct comparison, we developed a mathematical model, building upon previous work from Arnot and Gobas 2004, and Schram et al., 2017. Using a three-compartment approach, different technologies/strategies can be added to the model to determine a single parameter, R, which represents a removal rate of any off-flavor compound from a RAS system.
The three compartments are the change in the concentration in the fish, the water, and the change caused by a ‘reactor’ represented by equations one through three below.