Deep water culture buys growth rate by removing the substrate. Roots sit in the nutrient solution, uptake is not mediated by anything, and the crop responds. The same design decision removes the buffer: there is no soil air, no pore space, and no reserve. Everything the root gets, the water carries.
That is why root rot in a recirculating system is rarely a slow decline. It is an oxygen budget that goes negative in the warmest week of the cycle, followed by an opportunist that was present the whole time and only now has a host that cannot resist it.
Why is deep water culture so vulnerable to root rot?
Deep water culture suspends the whole root mass in water, so every root depends on what one reservoir carries. Warm water holds less oxygen, a heavy root mass consumes more of it, and a shared solution circulates any pathogen to every plant on the system. The buffer that soil provides is not there.
Start with the ceiling. Under air at one atmosphere, fresh water holds about 9.1 mg/L of dissolved oxygen at 20 °C and about 8.3 mg/L at 25 °C. That is a solubility limit set by Henry's law, not a target, and it falls as the room warms. A reservoir that drifts up a few degrees in late flower loses part of its ceiling at exactly the point the crop is asking for the most.
Demand moves the other way on the same thermometer. Respiration rate rises with temperature, and root mass is at its maximum late in the cycle, so the two curves converge. Add the microbial demand in the biofilm on every surface in the loop, and a system that was comfortable in week three can be running a deficit in week eight with nothing on the schedule having changed.
Then there is the shared-water problem, which is a design consequence rather than a hygiene failure. One infected site seeds a common solution, and the same pump that delivers nutrients delivers zoospores. This is why a DWC crash presents as a room-wide event rather than a plant-by-plant one.
Finally, the reading and the reality diverge at the root surface. A probe in a well-mixed reservoir reports the bulk concentration, while every root sits behind its own boundary layer and a mat of roots restricts flow through its interior. Dissolved oxygen is not the same as oxygen availability works through why a healthy reading and a hypoxic root are entirely compatible.
How does dissolved oxygen change Pythium pressure?
Oxygen does not kill Pythium. It changes the contest. Pythium tolerates low oxygen better than a root does, so hypoxia delivers both a weakened host and less competition from aerobic organisms. Restoring oxygen removes that advantage: root tissue stays functional, and the aerobic community around it stays in the game.
The pathogen's advantage is specific. Pythium species produce motile zoospores that swim in free water, and free water is exactly what a DWC system is. They colonize stressed and damaged tissue far more readily than healthy tissue, so a root already compromised by low oxygen is a substantially easier target than the same root a week earlier.
The competition argument is the second half, and it is weaker than the first. Raising dissolved oxygen shifts which organisms dominate the root zone, and aerobes that occupy the same space and the same exudates are competition a pathogen has to displace. How much protection that provides is not quantified, and the Kairospace CFU counts across three dissolved oxygen levels are a caution as much as a result: the highest level tested returned fewer colonies than the middle one. The community reorganized rather than simply growing, and a dose-response curve drawn through those points would be an invention.
What none of this supports is a threshold. There is no published dissolved oxygen level above which Pythium stops being a problem, and any grower offered one should ask which trial it came from. Oxygen shifts the odds in a contest that is still being run.
Read it as pressure management. The pathogen's opportunity is a function of host condition, competition, temperature, inoculum load, and time. Oxygen acts on the first two of those five. That is worth having, and it is not a control program.
What did we measure in cannabis?
In the GB Sciences deployment we measured two distinct outcomes: biomass up 30% and final yield up 15%. Those are different metrics and both are real. More plant does not convert one for one into more saleable flower. Read the pair as Kairospace field data from one site rather than as a rate to expect on yours.
Two different metrics, both real. One commercial site under its own conditions.
Keeping the two apart matters more than either number. Biomass is everything the plant built. Yield is what came off the trim table and made weight after drying. The ratio between them is set by cultivar, environment, and post-harvest handling, and a system that adds vegetative mass late does not necessarily add flower in proportion.
It also matters for how the figures travel. Biomass and yield are easy to blur once a number leaves the page it was measured on, and the split is stated here so the larger figure cannot be read as the saleable one. If a claim does not say which metric it is, treat it as the smaller one.
The provenance is the other half. This is field data from a commercial deployment, not a replicated multi-season trial with randomized blocks, and the honest read is a demonstrated result at one site under that site's conditions. It supports the conclusion that the oxygen constraint was real and binding there. It does not establish an expected value for a facility with different water, different cultivars, and a different room.
What it does license is the question to ask. If a deficit that size existed in a professionally run facility, the odds that one exists in a warmer room with a heavier canopy are not low, and the cost of finding out is one instrumented cycle.
Does this replace sterilization and IPM?
No. Oxygen management changes the pressure a pathogen can apply; it does not sanitize a system. Reservoir cleaning between crops, filtration, incoming water control, UV or ozone where the design calls for it, and the rest of an integrated pest management program all stay in place. Dropping them because a probe reads high removes controls and keeps the exposure.
The controls do different jobs and fail in different ways, which is the whole argument for layering them.
- Sanitation. Reduces inoculum.
- Filtration and disinfection. Intercept what is circulating.
- Oxygen. Strengthens the host and the competing community.
A system with high dissolved oxygen and a contaminated feed line still has a contaminated feed line.
This is also where the technology gets described dishonestly, and the correction is worth stating plainly. Hydrodynamic cavitation changes the physical state of the water — what it carries, how finely it disperses what is dissolved in it, and how much oxygen reaches the root. It improves how well the chemistry you are already running can do its job. It does not displace that chemistry, and anyone telling a grower that a bubble system retires their sanitation program is selling something.
Sequencing matters too. If the room runs hot, fix the temperature before anything else, because every term in this post moves with it. If the reservoir is not cleaned between crops, that is a cheaper problem to solve than a hardware purchase. Oxygen applied on top of an unaddressed hygiene failure will read as a disappointing result when what happened was a misdiagnosis.
The narrow version: oxygen is a layer that most DWC systems do not currently have, added to layers they should already have.
What DO level should I target?
There is no published set point for deep water culture, and treating one number as a target is the wrong frame. Dissolved oxygen above 30 mg/L is achievable with ultrafine bubble delivery, across published research and our own deployments, but that is a demonstrated concentration rather than a level any system holds against demand. Measure stability instead.
The reason the headline figure is interesting is that it sits in a different regime from the saturation ceiling. Water under air tops out near 8 mg/L in a warm room; supersaturated water is not a bigger version of the same thing, because the gradient into every root and biofilm is steeper. It is also not permanent — wherever treated water meets the atmosphere, it trends back toward the air-saturation ceiling at a rate set by surface area, turbulence, temperature, and demand.
So the useful measurements are where and when rather than how high. Read at the root mass, not at the injector or the return manifold, because the difference between those points is the thing you actually want to know. Read at the worst hour of the worst week: late flower, lights on, warmest reservoir temperature, heaviest canopy.
Then watch the shape of the curve rather than the peak. A system that reaches a high number after treatment and gives it all back within an hour is telling you the demand is larger than the delivery. A system that holds an elevated concentration through the photoperiod is telling you the deficit is closed.
The peak is a specification; the hold is the answer.
Our controlled environment page sets out where these systems are typically sized, the cannabis ROI calculator runs the economics on your own canopy and price, and the classroom section on hydroponics carries the underlying literature, including the supersaturation work the achievable figure comes from.