Aquaponics combines fish production with soilless crop production, so its performance is measured across water, energy, feed, harvests and nutrient management. The figures below span the 2023 U.S. Census of Aquaculture, published case studies, controlled trials and a 2023 design model; the periods and study settings are kept explicit because results are not interchangeable.
Contents
- U.S. aquaponics operations
- Water use and energy inputs
- Crop yields and growing cycles
- Fish stocking, growth and harvests
- Nutrient recovery and water quality
- Operating and economic indicators
U.S. aquaponics operations
The 2023 U.S. Census of Aquaculture counted 108 farms using aquaponics systems. Those farms reported 940 aquaponics tanks with a combined capacity of 1,297,097 gallons. The reported average tank volume was 1,380 gallons. These are census figures for U.S. farms in 2023, rather than a global estimate or a projection. The source is Table 12, Methods Used for Aquaculture Production: 2023.
The census numbers describe the scale of operations that reported aquaponics production, while the studies below describe individual systems. A farm count cannot by itself indicate output: the 940 tanks vary in size, configuration and purpose, and the reported average is not a production average.
Water use and energy inputs
Water replacement and energy demand differ substantially by system design. In a two-year study of a small-scale raft aquaponics system in Baltimore, published in 2015, the system lost roughly 1% of its water volume per day. It used an average of 35,950 liters of replenishment water per year and consumed an average of 19,526 kWh of propane and electricity annually. The average annual energy cost was $2,055. These measurements come from Energy and water use of a small-scale raft aquaponics system in Baltimore, Maryland, United States.
On a product basis, the Baltimore system required 104 liters of water and 56 kWh of energy to produce 1 kg of crops. A 1 kg increase in tilapia required 292 liters of water and 159 kWh of energy. The figures apply to that small-scale raft system and its study conditions; they should not be treated as a universal aquaponics conversion factor.
The USDA Hydroponic Task Force case study reported a different water-loss pattern for deep-water-culture systems operated for six years. Each system held approximately 18,000 U.S. gallons, and transpiration plus evaporation used approximately 0.5% of system volume per day. Each system had two fish-rearing tanks and four tanks for solids removal and mineralization of fish effluents. The report is the Hydroponic Task Force Report, reported in 2016.
The contrast between roughly 1% daily loss in the Baltimore raft study and approximately 0.5% daily use for transpiration and evaporation in the USDA case study illustrates why water statistics need a system boundary and a defined loss category. The first includes the operating conditions of a specific small-scale system; the second describes a case-study DWC configuration and its reported water use.
Crop yields and growing cycles
Crop results vary with crop type, coupling method and growing medium. In a tomato trial running from 7 April to 7 September 2015, or 154 days, the decoupled tomato unit harvested 123.5 kg of tomatoes. The coupled unit harvested 90.9 kg, and the decoupled yield was 36% higher than the coupled unit. The same study reported 63.7 kg of leaves, 5.8 kg of roots and 5.8 kg of stem biomass from the coupled unit during the trial. These results are from Decoupled systems on trial: Eliminating bottlenecks to improve aquaponic processes.
The decoupled tomato unit used 1.4 liters of water per plant per day at the beginning and end of the 154-day trial. Its peak water consumption ranged from 2.0 to 2.4 liters per plant per day between 7 May and 6 August 2015. The change over the growing cycle matters: a single daily rate would hide the difference between the early and late trial values and the mid-season peak.
Another experiment, reported in 2024, tested baby-leaf vegetables with rainbow-trout wastewater in a high-tech vertical decoupled aquaponic system. Mix–Peat lettuce yield was 2,497 g per square meter, approximately 6% higher than Hoagland–Peat. Mix–Peat used 38% less water and 10% less fertilizer than Hoagland–Peat. For rocket, Hoagland–Peat yield was 18% higher than Mix–Peat and 30% higher than Fish–Peat. The findings are reported in Greenhouse production of baby leaf vegetables using rainbow trout wastewater in a high-tech vertical decoupled aquaponic system.
| Crop comparison | Reported result | Period or setting |
|---|---|---|
| Decoupled tomatoes vs. coupled tomatoes | 36% higher yield | 154-day 2015 trial |
| Mix–Peat lettuce yield | 2,497 g/m² | Vertical study reported 2024 |
| Mix–Peat water use vs. Hoagland–Peat | 38% lower | Vertical study reported 2024 |
| Hoagland–Peat rocket vs. Mix–Peat | 18% higher yield | Vertical study reported 2024 |
These comparisons show that a favorable result for one crop does not automatically transfer to another. Lettuce, rocket and tomatoes respond to different production conditions, and the vertical-study percentages compare specific growing treatments rather than all aquaponic systems.
Fish stocking, growth and harvests
The USDA case-study systems stocked 1,250 tilapia in each 3,000-gallon fish tank. Fry grew from 1 g to 50 g in about four months before entering the aquaponics tanks, then grew from 50 g to approximately 900 g in another six months. Tilapia survival was greater than 99%. The case study was reported in 2016 and describes systems operated for six years; its source is the Hydroponic Task Force Report.
In the 7 April–7 September 2015 decoupled-systems trial, average fish feed-conversion ratios ranged from 1.2 to 1.3. Fish specific growth rate averaged 1.0 before declining toward 0.5–0.6 late in the trial. Fish-unit weight gain differed by less than 2% among the coupled, decoupled and control systems, while fish mortality was below 1% in all systems. Those figures come from the decoupled-systems trial.
A separate scalable coupled-aquaponics design model, published in 2023, uses a 35-week fry-to-harvest schedule ending at an average fish weight of 624 g. Its five growth phases specify feed-conversion ratios of 1.1, 1.2, 1.4, 1.6 and 1.8, with phase feed rates from 0.54 to 6.94 g per fish per day. The model averages 3.96 g of feed per fish per day across the five phases. These are design-model specifications, not measurements from a completed commercial farm. The source is Scalable coupled aquaponics design: Lettuce and tilapia production using a parallel unit process approach.
At a modeled feeding rate of 6.63 kg per day, the design would require 1,674 fish distributed across all growth phases, or 335 fish per phase in a five-phase grow-out. Each modeled harvest would yield 209 kg of fish, with modeled annual production of 1,553 kg. The same model allocates 75% of a 267.56-square-meter greenhouse to hydroponics and 15%, or 40.13 square meters, to fish rearing and water treatment.
Nutrient recovery and water quality
The PAFF Box study used a daily feed input of 42 g per square meter of deep-water-culture beds. Its deep-water-culture beds produced 3–10 times higher plant yields than ebb-and-flow beds. Potassium and phosphorus concentrations remained below 10 mg per liter, while iron, copper, zinc, manganese and molybdenum remained below 11 micrograms per liter. These measurements are from the 2017 study Plant and fish production performance, nutrient mass balances, energy and water use of the PAFF Box.
The same PAFF Box analysis required a 3.6% daily water-exchange rate to manage nutrient accumulation. That exchange rate was associated with more than 50% nutrient loss to the environment in the analysis. The result is important when evaluating “closed-loop” claims: recirculation can reduce replacement water, but nutrient management may still require discharge or exchange under particular operating conditions.
The 2024 vertical trout-wastewater experiment also reported a land-use difference. Mix–Peat used approximately 7% less land than Hoagland–Peat. That result belongs to the tested vertical system and treatments, so it is best read alongside the reported lettuce, rocket, water and fertilizer comparisons rather than as a general land-use benchmark.
The USDA case-study systems reported only city water, fish feed and supplemental iron as added inputs during six years of operation. This describes the listed added inputs for those systems; it does not establish that every aquaponics design can operate with the same input profile.
Operating and economic indicators
The Baltimore study linked 0.5 kg of feed with 1 kg of crops and 1.3 kg of feed with a 1 kg increase in tilapia. It assigned about $6 in energy cost to each kilogram of crops produced and about $12 in energy cost to each kilogram increase in tilapia. These are study-specific allocations from the two-year Baltimore system, not current prices or a universal cost model. See Energy and water use of a small-scale raft aquaponics system in Baltimore, Maryland, United States.
Infrastructure and operating schedules also shape the economics. In the USDA DWC case study, one water pump ran continuously, 24 hours per day and 7 days per week. Each system had two deep-water hydroponic troughs measuring 8 by 75 by 1 feet. Each system also included the two fish-rearing tanks and four solids-removal and mineralization tanks described above. These dimensions and operating details come from the Hydroponic Task Force Report.
The scalable 2023 design model provides a planning view rather than an observed cost outcome. Its 267.56-square-meter greenhouse assigns 75% to hydroponics and 15%, or 40.13 square meters, to fish rearing and water treatment. Together with the modeled 35-week fish cycle, 624 g average harvest weight, 209 kg per modeled harvest and 1,553 kg annual production, these figures define a proposed production arrangement—not a forecast of revenue, profit or payback.
For growers and market planners, the most useful takeaway is to keep the metrics paired with their boundaries. Farm counts describe reported U.S. operations in 2023; water and energy rates describe named systems; crop and fish results belong to particular trials; and model outputs are estimates based on design assumptions. That distinction makes aquaponics statistics more useful for comparing systems without turning one study’s result into a promise for every garden or market operation.