Library › Part C · Article 22 in the series · Level 1 · 12 min read

Deep water culture: roots in a deep, aerated solution

Standfirst: After reading this article you will be able to explain how deep water culture keeps roots alive in a tank of solution, why its volume makes it forgiving, why the air pump can never stop, and how commercial raft ponds are laid out. You will also be able to size the solution volume, the oxygen store and the air flow for a square metre of lettuce with a calculator.

Before you read: Anatomy of a hydroponic system, Dissolved oxygen in the root zone, Managing a recirculating solution over time.

What deep water culture is

In deep water culture the roots hang in a deep volume of nutrient solution, and air is pushed into that solution so the roots do not run out of oxygen. Resh groups it under water culture, the family in which the roots are suspended in liquid and the crowns are held up by a polystyrene cover¹.

It comes in two forms that work by one mechanism. The home form is a bucket with a lid, one plant in a net pot through the lid and an aquarium air stone on the bottom; a University of Florida extension guide describes plastic buckets “with the plant contained in a net pot suspended from the center of the lid and the roots suspended in the nutrient solution”, oxygenated by an aquarium air stone¹⁵.

The commercial form is raft culture, also called raceway, floating or deep flow culture. Plants sit in holes in polystyrene foam boards that float on a shallow pond of aerated solution; the boards are pushed along the pond as the crop grows and pulled out at the harvest end^1,13. Merle Jensen of the University of Arizona built the prototype raceway in 1981 to 1982, with beds 15 to 20 cm deep holding a large, fairly static volume of solution¹. Lettuce, spinach, kale and herbs are its usual crops¹⁴.

What both forms share is depth: a nutrient film channel holds millimetres of solution per plant, a pond or bucket holds litres. That difference explains the method’s strength, its buffering, and its one absolute demand, continuous aeration.

The volume is the buffer

Everything that drifts in a nutrient solution drifts more slowly when there is more of it, because the plant removes the same mass of water, ions and heat per day whatever the tank size. Jones, citing Morgan (2002), lists among the advantages of the raft system that it “moderates the temperature conditions due to the larger volume of nutrient solution as well as its buffering capacity”². Managing a recirculating solution over time covers the drift itself; deep water culture is the system in which it is slowest.

The buffer slows drift; it does not stop it. Cornell’s deep flow programme mixes its ponds to an EC of 1,200 µS/cm, and Resh’s raft growers hold 1.2 to 1.3 mS/cm by automatic injection of stock solutions, with acid or potassium hydroxide dosed on the return line to hold pH^1,3. In a 30 L anthroponic deep water tub at PonicLabs in 2016, the pH rose above 7.5 and was brought down with phosphoric acid at the weekly check¹⁸.

Why oxygen must be added all the time

Roots respire, and in water the oxygen they use must arrive dissolved. Cornell is direct about what happens otherwise: “If no oxygen is added to the pond, DO levels will drop to nearly 0 ppm”, and the absence of oxygen “will stop the process of respiration and seriously damage and kill the plant”³. Dissolved oxygen in the root zone works through the store and the drain; the numbers that matter for a raft are these.

The store is small: water in contact with air holds 9.077 mg of oxygen per litre at 20 °C and 7.539 mg/L at 30 °C⁵. The drain is set by the roots. Chun and Takakura measured leaf lettuce roots 26 days after transplanting respiring 2.9 to 6.7 µmol of oxygen per hour per millilitre of roots at 23 °C, with a floor near 78 µmol/L, roughly 2.5 mg/L⁷. Goto and co-workers grew lettuce on floats for 24 days at 2.1, 4.2, 8.4 and 16.8 mg/L, found no difference in growth and no root damage, and put the critical level below 2.1 mg/L⁸. Cornell reports satisfactory growth at a level of at least 4 ppm and keeps its ponds at 8 ppm, within 7 to 10, with “no advantage to 20”³.

Still water also fails before it is empty: Chun and Takakura note that “the immersion of roots in unagitated water drastically reduces the diffusion of gases to the roots”⁷. Aeration therefore does two jobs, replacing the oxygen the roots take and stirring the tank so the solution at the root surface is the same as the solution in the middle.

Two ways to add it

Resh names the two routes: forced aeration, bubbling air “through a perforated pipe or airstone placed at the bottom of the bed and/or in the nutrient tank”, or circulating the solution through the beds and back to a reservoir, with baffles at the bed ends to aerate the returning water¹. For a raft bed about 30 m (100 ft) long holding 10 to 15 cm of solution he gives “about 1–2 complete changes per hour”¹. On the next page he describes Jensen’s 3,600 L raceway beds circulating at only 2 to 3 L/min, about one exchange per 24 hours¹. The two figures differ by a factor of about 30; our reading is that the first aerates the bed by flow alone and the second only turns over a pond that is aerated and chilled in the tank.

Bubbling works mostly through the surface, not the bubbles. For fine-bubble membrane diffusers in deep tanks the measured oxygen transfer is 5.7 to 6.0% per metre of depth¹¹, so at 15 cm under 1% of the oxygen in the air would dissolve, if that figure held at such shallow depths, which is this article’s extrapolation. The rest of the air leaves at the surface, having lifted and stirred the water, and that stirring is what keeps oxygen diffusing in through the surface⁵. Cornell adds pure oxygen to its ponds, but says an air pump and aquarium air stone will do for sufficiently small systems³.

Raft culture: the numbers growers use

Depth. Jensen’s beds were 15 to 20 cm deep¹. Resh’s beds for small greenhouses have “a minimum depth of at least 8 in. (20 cm)”, and the two ponds he built in Anguilla were 30 cm deep¹. Jones gives 7.6 to 25 cm, reports research in which depths above 7.6 cm “did not improve growth or yield”, and adds that 12.7 cm “tends to result in more consistent growth”²; a Florida backyard float garden keeps its board on at least 12.7 cm (5 in.) of solution¹⁶. Depth is chosen for buffer volume and for time between top-ups, not for yield.

Density. Cornell grows 150 g leaf lettuce at 97 plants per square metre from day 11 to day 21, then re-spaces to 38 plants per square metre until harvest on day 35³. Resh’s raceway greenhouse plans 32 heads per square metre of growing area, 64 per 1.2 m by 1.2 m board at 15 cm spacing¹. Hydronov, which Resh calls a world leader in large raft greenhouses, sets 18 plants in each 2 ft by 4 ft board for the final stage¹. At 38 plants per square metre and 15 cm of depth each plant has 3.9 L of solution, our arithmetic from^1,3.

Temperature. Resh calls a chiller in the nutrient tank “standard practice” for raft lettuce, holding the solution at 18 to 23 °C to delay bolting in hot regions¹. In Anguilla, solution at 18 °C under 32 °C air delayed bolting by 3 to 4 days and “reduces fungal infection of the roots by Pythium”¹. Cornell’s set-point is “No higher than 25C, cool at 26C, heat at 24C”³, and the experiment behind it found that a pond held at 24 °C “maintained market quality lettuce head production in 31 deg C air”⁴. Root-zone temperature covers the methods.

Layout and boards. The continuous-harvest layout is a long pond with transplanting at one end and harvest at the other; boards are pushed in at the transplant end and pulled to the harvest end by a winch on a string under the rafts¹. Resh uses 2.5 cm high-density polystyrene insulation board, the blue “roofmate” type, and Cornell drills its floats from 1 in. (2.5 cm) insulation^1,3. Boards return to the pond only after cleaning: Resh hoses them and dips them in 10% bleach and cleans each bed after every crop; Cornell washes floats “with a weak bleach (2%) solution” between cycles^1,3. The boards also insulate the solution and keep light off it¹.

The home bucket

The Florida guide builds its bucket from a 5 US gallon (18.9 L) bucket with a lid, a 15 cm (6 in.) pot in a hole in the lid, an aquarium air pump, tubing and one or two air stones, and recommends it for “long-term” crops such as tomatoes, peppers, cucumbers and beans¹⁵. A bucket gives one plant 18.9 L against 3.9 L per lettuce in a commercial pond, which is why it can carry a tomato^1,3,15.

The starting water level follows the logic of a Kratky container. Kratky sets the net pots “such that their lower portion is initially immersed in nutrient solution”, and the level “may remain the same or be lowered, but it should not be raised because submerging the oxygen roots will cause the plant to ‘drown’"¹⁷. So start with the solution touching the base of the net pot, let it drop a little as roots emerge, and never refill over roots that have grown in air. This is our inference from Kratky’s method, not a measured result for buckets.

Light must be kept out. Resh explains that light promotes algae, which compete for nutrients and, at night, for oxygen from the solution, and prescribes opaque containers or covers¹. A translucent bucket needs painting or wrapping.

Two PonicLabs trials in Malmö used this format with organic inputs: a 13.65 L box with 12 L of water, an air stone and six basil plants fed with dried vermicompost powder gave visibly healthy, full-size plants at 50 days with EC relatively stable throughout¹⁹; the 2016 anthroponic tub needed weekly pH correction¹⁸. Both are observations on bioponic variants (see the PonicLabs sustainable-hydroponics pages), not general rules.

Worked example: one square metre of lettuce raft

Take a 1 m² raft pond at Cornell’s final density of 38 plants per square metre³, 15 cm deep, the lower end of Jensen’s beds¹. Every number below is this article’s arithmetic from the cited inputs; the script is in the notes.

1. Volume. 1 m² × 0.15 m = 150 L, 3.9 L per plant^1,3.
2. Oxygen store. The Benson and Krause equation used by the USGS gives saturation of 9.09 mg/L at 20 °C and 7.83 mg/L at 28 °C⁶, within 0.02 mg/L of the table value at 20 °C⁵. The pond holds 1,364 mg at 20 °C and 1,174 mg at 28 °C^5,6. Above Cornell’s 4 mg/L floor³, 764 mg and 574 mg are usable.
3. Root demand, measured. Chun and Takakura’s regression gives 0.0698 × 1.1230^26 = 1.42 mL of roots per plant at 26 days after transplanting; at their upper rate of 6.7 µmol per hour per millilitre, 38 plants use 362 µmol/h, which at 32 mg per millimole is 11.6 mg/h; at the lower rate, 5.0 mg/h⁷. Between 0 and 30 °C respiration roughly doubles per 10 °C⁹, so from 23 to 28 °C the upper rate rises by 2^0.5 = 1.41 to 16.4 mg/h^7,9.
4. Root demand, upper bound. Marschner, citing Cannell, gives up to 17 L of oxygen per square metre per day for a dense crop stand in soil¹⁰; at 1.42905 mg/mL⁶ that is 24.3 g/day, about 1,010 mg/h. No measurement of a mature raft at that rate was found; treat it as a ceiling.
5. Time if the air stops. At 20 °C the measured demand empties the usable store in 764 / 11.6 = 66 hours; at 28 °C in 574 / 16.4 = 35 hours. At the ceiling the same store lasts 45 minutes at 20 °C and 34 minutes at 28 °C^3,5,6,7,9,10.
6. Air flow. One litre of air carries 20.946% oxygen⁵ at 1.42905 mg/mL⁶, about 299 mg. If the fine-pore transfer of 5.7 to 6.0% per metre¹¹ held at 0.15 m, about 0.9% would dissolve. The measured demand at 28 °C then needs 16.4 / 0.009 / 299 = 6.1 L of air per hour, 0.1 L/min; the ceiling needs 376 L/h, about 6.3 L/min^{5,6,7,9,10,11}.

The spread between steps 3 and 4 is the honest answer: a young crop of small roots barely dents the store, a dense mature one can empty it in under an hour, and only a meter in your own pond tells you where you are. A small air pump is enough for a square metre of lettuce; its failure matters more in summer than in winter.

What goes wrong, and why

Pythium root rot. Pythium is a water mould whose swimming spores, zoospores, “are dispersed by transport in nutrient solution”¹², so in a pond one infected plant seeds the crop; Jones warns that Pythium “can quickly destroy a crop and require that the water and the entire growing system be sterilized”². A warm, crowded pond gives it both of the conditions it favours. Pythium aphanidermatum causes severe root rot “when root zone temperature is moderate or high (≥23-27°C)”; Sutton and colleagues report from their own trials that colonised roots stay almost symptomless at 16 to 18 °C “yet develop severe symptoms within minutes or hours when the temperature is raised to 24-28°C”¹². And “In general, root rot increases when oxygen levels are low”, while the oxygen demand of roots and root-zone microbes rises with temperature¹². Disinfesting the recirculating solution usually has a minor effect on epidemics compared with treatments in the root zone¹², and Resh notes that UV sterilisation does not combat Pythium¹. Temperature, oxygen and hygiene between crops are the controls that work.

Warm nights and pump failure. At night respiration continues while photosynthesis stops, and the water, which lags the air⁵, is still warm from the day, so demand is high while the store is smallest^5,9. If the air stops, the worked example gives the scale: small roots have days, a dense mature crop has well under an hour, and warm water shortens both^3,5,7,9,10. A pond and a bucket fail the same way; the difference is how many plants fail together.

Electricity next to water. Keep pumps and connections above the water line so a stopped pump cannot siphon solution back, and put the circuit on a residual-current device, which is general electrical practice rather than a hydroponic rule.

Key numbers

Common mistakes

1. Treating the big tank as a reason to check less often. The volume slows drift; it does not stop it, and a pond at 1.2 mS/cm still needs its EC and pH held by dosing^1,3.
2. Running a pond without aeration because the water is deep. Without added oxygen the level falls to nearly zero and the roots stop respiring³; depth buffers salts and heat, not oxygen.
3. Letting the solution warm with the greenhouse. Above about 25 °C the store is smaller, respiration faster and Pythium turns from silent to severe within hours^3,9,12.
4. Refilling a bucket to the top after roots have grown into the air gap. Submerging roots that formed in moist air drowns them¹⁷.

Summary

Deep water culture puts roots in litres of solution rather than millimetres, as a bucket with an air stone at home or as a pond of floating boards in a greenhouse. The volume slows every drift of temperature, EC and pH, which is why beginners and commercial lettuce growers both use it. The one thing it cannot buffer is oxygen: the solution holds under 10 mg/L at pond temperatures⁵, roots draw on it day and night, and still water loses it from the roots outward, so air must be bubbled or circulated without pause. Commercial ponds are 15 to 30 cm deep, carry about 30 to 40 lettuce per square metre and are held at 18 to 25 °C^1,3. A small air pump covers the measured demand of young lettuce many times over, but a dense mature crop in warm water can exhaust the store in under an hour if the pump stops. Warm, oxygen-poor water is also what Pythium needs, so solution temperature, aeration and hygiene between crops are the three habits of a deep water grower.

Next in the series: Nutrient film technique.

References

1. Resh, H. M. (2013). Hydroponic Food Production: A Definitive Guidebook for the Advanced Home Gardener and the Commercial Hydroponic Grower, 7th ed. CRC Press, Boca Raton. Chapter 5, Water culture.
2. Jones, J. B. Jr. (2005). Hydroponics: A Practical Guide for the Soilless Grower, 2nd ed. CRC Press, Boca Raton.
3. Brechner, M., Both, A. J. and CEA Staff (2013). Hydroponic Lettuce Handbook. Cornell University Controlled Environment Agriculture Program, Ithaca, NY. https://cpb-us-e1.wpmucdn.com/blogs.cornell.edu/dist/8/8824/files/2019/06/Cornell-CEA-Lettuce-Handbook-.pdf
4. Thompson, H. C., Langhans, R. W., Both, A. J. and Albright, L. D. (1998). Shoot and root temperature effects on lettuce growth in a floating hydroponic system. Journal of the American Society for Horticultural Science, 123(3), 361–364. https://doi.org/10.21273/JASHS.123.3.361 (abstract read in reference 3, p.44)
5. Lucas, J. S. and Southgate, P. C. (eds) (2012). Aquaculture: Farming Aquatic Animals and Plants, 2nd ed. Chapter 3, Water quality. Wiley-Blackwell, Oxford. Table 3.2 reproduced from Colt, J. (1984), American Fisheries Society Special Publication 14.
6. U.S. Geological Survey, Office of Water Quality (2011). Change to solubility equations for oxygen in water. Technical Memorandum 2011.03, attachment pages A2 and A4. https://water.usgs.gov/admin/memo/QW/qw11.03.pdf Underlying paper: Benson, B. B. and Krause, D. Jr. (1984). Limnology and Oceanography, 29(3), 620–632. https://doi.org/10.4319/lo.1984.29.3.0620
7. Chun, C. and Takakura, T. (1994). Rate of root respiration of lettuce under various dissolved oxygen concentrations in hydroponics. Environment Control in Biology, 32(2), 125–135. https://doi.org/10.2525/ecb1963.32.125
8. Goto, E., Both, A. J., Albright, L. D., Langhans, R. W. and Leed, A. R. (1996). Effect of dissolved oxygen concentration on lettuce growth in floating hydroponics. Acta Horticulturae, 440, article 36. https://doi.org/10.17660/ActaHortic.1996.440.36
9. Taiz, L. and Zeiger, E. (2002). Plant Physiology, 3rd ed. Sinauer Associates, Sunderland, MA. Chapter 11.
10. Marschner, P. (ed.) (2012). Marschner's Mineral Nutrition of Higher Plants, 3rd ed., Chapter 13, section 13.4.3 Aeration. Academic Press, London.
11. Mueller, J. A., Boyle, W. C. and Pöpel, H. J. (2002). Aeration: Principles and Practice. Water Quality Management Library, Volume 11. CRC Press, Boca Raton.
12. Sutton, J. C., Sopher, C. R., Owen-Going, T. N., Liu, W., Grodzinski, B., Hall, J. C. and Benchimol, R. L. (2006). Etiology and epidemiology of Pythium root rot in hydroponic crops: current knowledge and perspectives. Summa Phytopathologica, 32(4), 307–321. https://doi.org/10.1590/S0100-54052006000400001 (open-access PDF: https://www.alice.cnptia.embrapa.br/alice/bitstream/doc/576855/1/a01v32n4.pdf
13. Mason, J. Commercial Hydroponics, 3rd ed. (e-book). ACS Distance Education.
14. Nelson, M., Langellotto, G. and Nackley, L. (2025). Hydro hints: Deep water culture. Oregon State University Extension Service, EM 9455. https://extension.oregonstate.edu/catalog/pub/em9455 (PDF: https://extension.oregonstate.edu/sites/extd8/files/catalog/auto/EM9455.pdf
15. Sewards, J. Making Your Own, Inexpensive Hydroponic Bucket (Deep Water Culture). UF/IFAS Extension, Volusia County, slide set. https://www.volusia.org/core/fileparse.php/4461/urlt/making-your-own-hydroponic-bucket.pdf
16. Sweat, M., Tyson, R. and Hochmuth, R. Building a Floating Hydroponic Garden. University of Florida IFAS Extension, publication HS184. https://ask.ifas.ufl.edu/publication/HS184
17. Kratky, B. A. (2009). Three non-circulating hydroponic methods for growing lettuce. Acta Horticulturae, 843, 65–72.
18. Sanchez, H. (2016). Ocimum basilicum and Coriandrum sativum cultivation in a decoupled anthroponics system. PonicLabs technical report, Malmö. (summarised in library/poniclabs_compressed_context/reports_compressed.md, lines 213–254)
19. Sanchez, H. (2019). Cultivating Ocimum basilicum in a hydroponics DWC system using dried vermicompost powder. PonicLabs technical report, Malmö. (summarised in library/poniclabs_compressed_context/reports_compressed.md, lines 288–322)

This article was drafted with the help of AI tools and reviewed and edited by PonicLabs. Sources are listed above; tell us if you find an error.