Feed Loading and Stocking Density in RAS

Why daily feed is the variable that sizes every RAS component, what stocking densities studies report per species and stage, where welfare limits appear and how feed conversion is tracked.

Feed as the master design variable

SRAC Publication 453 describes fish density as the primary controller of system stability and notes that density ultimately sets the feed application rate. In a typical design exercise the holding capacity of the system is defined first, the peak daily feed ration is calculated from it, and every component is then sized to handle that same peak feed load, often with a uniform safety factor such as 1.5. Feed matters because almost every waste product scales with it: SRAC Publication 451 explains that uneaten feed and metabolic by-products generate carbon dioxide and ammonia and consume oxygen.

Loads expressed per kilogram of feed

Extension and FAO sources express many design and operating figures per unit of daily feed. The values below illustrate the principle; the dedicated RAS guides on biofilters, solids and oxygenation treat each process in detail.

Ammonia: in a worked example in SRAC 451, about 3% of a 32% protein feed becomes ammonia-nitrogen.
Oxygen: SRAC 451 states that systems that remove solids quickly may consume as little as 0.3 kg of oxygen per kg of feed, while systems that hold solids between filter backwashes may consume up to 0.75 kg per kg of feed.
Solids: SRAC 453 states that about half of the feed consumed is excreted as solids, and SRAC 452 cites an estimate that more than 60% of feed placed in a system ends up as particulates.
Water flow: SRAC 453 gives typical recirculation rates of 42 to 84 litres per minute per kg of daily feed ration.
Biofilter media: SRAC 453 sizes moving bed reactors for about 8 to 16 kg of daily feed per cubic metre of media, and coarse sand fluidised beds for about 16 to 32 kg per cubic metre.
New water: the FAO guide states that more than 300 litres of new water per kg of feed is usually enough to dilute nitrate; below that, denitrification becomes worth considering.

Feeding rates

SRAC 451 reports that high-protein pelleted diets are fed at 1.5 to 15% of body weight per day, about 15% for juveniles and 1.5% for market-size fish. SRAC 452 states that fingerlings eat 3 to 4% of body weight daily until they reach about 1/4 to 1/2 lb (about 113 to 227 g, converted here), then 2 to 3% until harvest. It recommends splitting stocks into new tanks when feed demand exceeds what the biofilter can handle, feeding several times a day to spread the waste load and avoid sudden oxygen drops, and treating a poor feeding response as an immediate alarm to check water quality and fish health. Overfeeding wastes feed, degrades water quality and can overload the biofilter.

Design density bands and species examples

SRAC 453 gives design bands rather than species norms: very low densities below 15 kg/m3 for broodstock and display stock, moderate loading below 30 kg/m3 for fingerlings, baitfish and ornamental fish, and about 60 kg/m3 (0.5 lb per gallon) as a widely accepted growout design figure. About 120 kg/m3 can be reached, but such systems often have unstable water quality and more disease and growth problems. Published trials show how species and life stage change the picture.

Atlantic salmon (Salmo salar) post-smolts, about 115 g, in flow-through seawater at 9.3 °C for 8 weeks (Calabrese et al., 2017): growth slowed at 50 kg/m3 and above, feed conversion ratio rose linearly from 25 to 125 kg/m3, fin damage increased at 100 kg/m3 or more, and stress markers appeared at 125 kg/m3. The authors considered densities up to 75 kg/m3 feasible without compromising performance and welfare at that temperature and size.
Atlantic salmon in RAS for 66 days (Liu et al., 2017): fish held at about 29 to 54 kg/m3 had lower final weight and higher cortisol than fish at about 10 to 18 kg/m3, which also had the lowest feed conversion ratio.
Rainbow trout (Oncorhynchus mykiss), juveniles, 30 days at 10, 40 and 80 kg/m3 (Yarahmadi et al., 2016): stress indicators rose and immune parameters fell at 40 and 80 kg/m3. In contrast, a UK thesis (North, 2004) found that around 80 kg/m3 did not produce consistent negative effects when good water quality was maintained, and a RAS study on adult trout of about 1.5 kg (Raymo et al., 2024) found stress markers at a low density of 12 kg/m3 compared with 43 kg/m3.
Turbot (Scophthalmus maximus), juveniles in RAS for 120 days (Liu et al., 2019): the group ending at about 52 kg per square metre of tank bottom grew more slowly, converted feed less efficiently and showed higher stress markers. Flatfish densities are reported per area, not per volume.
Olive flounder (Paralichthys olivaceus), about 60 to 180 g in RAS (Seo and Park, 2022): the authors suggest a final density of about 20 kg/m2.
Tilapia in tanks (SRAC 452, Table 6): recommended stocking falls from 225 fish per cubic foot for 0.02 g fry to 3 fish per cubic foot for fish growing from 250 to 450 g; 1 cubic foot is about 28.3 litres, a conversion added here.

Feed conversion ratio

The FAO guide defines the feed conversion rate (FCR) as kilograms of feed used per kilogram of fish produced and states that feed is by far the largest production cost in RAS, so improving FCR has a significant effect on efficiency. Wikipedia gives FCRs of around 1 for farmed Atlantic salmon and catfish and about 1.5 for tilapia, and notes that fish can reach values below 1 because dry feed is more energy-dense than wet fish flesh. SRAC 452 uses an FCR of 1.5:1 for tilapia in its feeding schedule and shows how to track it: weigh a sample of fish, estimate tank biomass, divide feed given by weight gained (in its example, 100 lb of feed for 70 lb of gain gives 1.43), then reset the daily ration. SRAC 451 notes that tank systems generally achieve better feed conversion than ponds. Several density trials above report FCR worsening as density rises.