Effects of feed loading on nitrogen balances and fish performance in replicated recirculating aquaculture systems

Publication: Research - peer-reviewJournal article – Annual report year: 2012

View graph of relations

This study investigated the effects of applying four fixed feed loadings to three replicated recirculating aquaculture systems (RAS) on water quality changes, nitrogenous balances and growth performance of rainbow trout (Oncorhynchus mykiss).Feed loadings ranged from 1.6 to 6.3kgfeed/m3 make-up water, with a constant make-up water renewal of 4.7% of total water volume per day in all twelve RAS. Fish densities ranged from 14 to 92kg/m3 during the prolonged trial of 10weeks. Selected water quality parameters were measured during two intensive sampling campaigns, evaluating biofilter nitrification performance and diurnal patterns of total ammonia nitrogen (TAN) and nitrite concentrations. No fish mortality occurred during the study. Feed conversion ratios varied between 0.91±0.04 and 0.95±0.02, and were unaffected by feeding load. Mean nitrate-nitrogen levels ranged from 54±7 to 196±10mg/L at steady state, and the concentration of nitrogenous compounds and organic matter were all positively correlated to feed loading.The TAN loading to the RAS from the specific feed type was assessed in a separate mass-balance study and used as input in a descriptive mathematical model (AQUASIM® software) developed to simulate processes affecting N mass-balances in the RAS. Nitrification kinetic rate constants were applied to the biofilter, and fractions of nitrifiers in suspended solids in the water phase were estimated based on existing information from waste water treatment processes. Two model scenarios successfully simulated the measured TAN concentration in the experimental RAS. The first model scenario applied a first-order area-based nitrification rate (k1a) constant of 0.2m/d, estimating a fraction of active nitrifiers (fN) in the water phase of 4% of the total suspended solids. The second model scenario used a k1a of 0.1, estimating a fN of 8% with similar predictability as in the first scenario. Overall, this study provided new information on fish performance and resulting water quality during steady state RAS operation. Furthermore, the study demonstrated that kinetic modeling can be applied to simulate measured TAN concentrations in experimental RAS.
Original languageEnglish
Pages (from-to)237-245
StatePublished - 2012
CitationsWeb of Science® Times Cited: 26
Download as:
Download as PDF
Select render style:
Download as HTML
Select render style:
Download as Word
Select render style:

ID: 7607511