Off-flavor compounds, geosmin (GSM) and 2-methylisoborneol (MIB), bioaccumulate in Atlantic salmon Salmo salar and other species produced in recirculating aquaculture systems (RAS) causing earthy/musty taints that are unacceptable to consumers and an increased likelihood of economic loss due to rejected products. Improved knowledge of environmental dynamics that inhibit off-flavor in land-based systems would help to ensure that edible products are off-flavor-free and could lead to solutions to this critical industry challenge.
Observations from two Atlantic salmon studies recently improved our understanding of variables that could influence off-flavor accumulation in RAS. During these trials, increased GSM and/or MIB production was attempted via discontinued system cleaning, first in eight identical partial reuse systems (PRS; Study 1) and later in six replicate RAS with fluidized sand biofilters (Study 2). Partial reuse systems were pre-disinfected by recycling 200 mg/L hydrogen peroxide, while microbially mature RAS that had been operated for >1 year without shutdown were utilized during the second study. Time-series water samples were collected, at minimum, every other week and shipped overnight to a contract laboratory for analysis. Further, salmon fillets (Study 1) containing 211-231, 305-319, and 432-467 ng/kg GSM were tested against control fillets (43-103 ng/kg GSM) by an untrained taste panel to determine if differences could be detected.
Average waterborne GSM increased from 4.6 ± 0.5 ng/L to 15.5 ± 1.3 ng/L in replicate PRS and reached a maximum level of 20.5 ng/L after one month without cleaning, but MIB was rarely detected. Conversely, GSM and MIB levels measured in replicate RAS did not increase beyond 3 ng/L for two months despite the lack of cleaning. Additional attempts to increase off-flavor levels were largely unsuccessful over the next two months, including discontinued use of an ultraviolet light treating makeup water, increased water flushing, sand transfer from a RAS with higher off-flavor levels, and on/off cycling of pumps to disrupt microbial biofilms in the fluidized sand biofilter. Moreover, sensory analysis showed that participants detected a difference in salmon fillets with 432-476 ng/kg GSM vs. control fillets; however, qualitative remarks did not separate earthy/muddy off-flavor as the direct cause.
These studies indicate that GSM and MIB production may be influenced by microbial maturity, water flushing rate, and/or the presence of a fluidized sand biofilter. This work also shows that PRS are not exempt from off-flavor accumulation. Additional research is needed to understand the microbial dynamics related to these conditions. Although new information was gleaned relative to GSM taste threshold, a trained taste panel is still required to determine if consumers with sensitive palates can detect low-level GSM and MIB in RAS-produced Atlantic salmon.