Crop steering is the most confidently quoted set of numbers in commercial cannabis cultivation. A “generative” substrate EC of six to nine. A ten percent overnight dryback. Shot sizes, target bands for volumetric water content, substrate targets for each phase. We went looking for the trials behind those numbers. Peer-reviewed cannabis studies on drought stress and root-zone EC do exist. None of the ones we found validates those specific thresholds. Every specific threshold we could trace leads back to education content from equipment vendors. The only published factorial trial we found that included dryback does not support the assumption that a generative dryback raises yield, at least in the two cultivars it tested. That trial was commissioned by a substrate manufacturer, has no journal, DOI or peer review, and was described in thin detail. The physiology is sound. In our view, the specific numbers are vendor guidance, not established findings.
1. What crop steering claims to be
The concept is to manage substrate electrical conductivity (EC) and volumetric water content (VWC) on purpose through the crop cycle, to push the plant toward vegetative growth or toward generative (reproductive) development. Higher substrate EC and larger drybacks push generative. Lower EC and smaller drybacks push vegetative. The tools are sensors in the substrate, irrigation delivered in small “shots”, and a phase plan across the cycle.
The doctrine is usually taught as a sequence, not a single direction. Growers steer generatively around the flip to limit stretch. They steer vegetatively, with smaller drybacks, lower substrate EC and more frequent shots, while the plant builds structure and bulks. They return to generative steering late in flower to “finish” the crop. The trade-off is usually framed as yield against quality: vegetative steering for mass, generative steering for density, resin and cannabinoid content, sometimes at some cost in weight. Any test of the doctrine has to be read against that sequence. It is not simply a rule that bigger drybacks are always better.
The underlying plant physiology is not in dispute. Across horticultural crops it is well established that mild water stress and higher root-zone salinity shift resources from vegetative growth toward reproduction. Sensor-based irrigation management is a real and mature discipline in greenhouse production generally.
This article is about the second layer: the specific numeric thresholds that get quoted as though they were established for cannabis.
2. What we searched for, and what we found
Peer-reviewed cannabis studies on drought and root-zone stress do exist. Caplan, Dixon and Zheng (2019, HortScience) applied controlled drought stress to container-grown medical cannabis in flower. They withheld fertigation until midday leaf water potential fell to about −1.5 MPa. Compared with well-watered controls, the stressed plants had THCA and CBDA concentrations about 12% and 13% higher, and THCA and CBDA yield per unit growing area about 43% and 47% higher. The authors note that results may differ for other cultivars and chemotypes. In plain terms, that result points in the direction the doctrine expects: water stress late in the cycle favoured cannabinoid output.
On root-zone EC, Allred, Fatzinger and Bugbee (2025, Journal of Cannabis Research) tested osmotic stress as a crop steering tool in two medical cultivars. The treatments were a 4 mS/cm control, a constant 8 mS/cm, and 8 mS/cm for the first four short-day weeks followed by 4 mS/cm. The higher-EC treatments reduced plant height by about 15%. They also cut flower yield by about 20% in some cultivar–treatment combinations and did not significantly change THC or CBD. The authors concluded that it is difficult to reduce plant size without reducing yield. The Bernstein group in Israel has published controlled studies on how mineral nutrition shapes cannabinoid and terpenoid profiles in medical cannabis, for example nitrogen supply (Saloner and Bernstein 2021, Industrial Crops and Products).
None of these studies validates the specific vendor thresholds: substrate EC 6–9, a 10% overnight dryback, or particular shot sizes. Caplan defined stress by plant water potential during a single drought episode, not by a nightly dryback percentage. Allred and colleagues tested EC in the feed and found a cost to yield.
Every specific recommendation we could trace leads back to vendor education content (AROYA/Addium, Grodan, TSRgrow, Front Row Ag, Growlink) or to calculator websites built on it.
These companies sell substrate sensors, substrates, irrigation controllers and fertigation systems. That does not make their guidance wrong. They may well hold more hands-on cannabis irrigation data than most universities, and much of their material is thoughtful. They also have commercial reasons to want it to work. What it means is that the numbers have not been independently replicated or peer reviewed, yet they are repeated as established fact.
How to spot this problem is simple: ask where a number came from. “Generative EC 6–9” is quoted constantly. Ask which trial, on which cultivar, in which substrate, measuring which outcome. In our searching, the trail ended at an education page every time.
3. The only published factorial trial we found that included dryback
Grodan ran this trial with Royal Queen Seeds at CRIC Labs in Québec, Canada. It used a 2×2×2 factorial design. The three factors were variety (Quantum Terp F1 and Gaia F1), maintenance (Regular or None) and overnight dryback (Small or Large). There were six plants per cell, grown in stone wool and harvested on 13 August 2025.
- Maintenance means scheduled canopy work: defoliation (removing fan leaves), pruning of lower growth that will not make usable flower, and training (bending or tying stems). “No Maintenance” plants got none of this.
- Dryback was defined as the fall in substrate water content between the last irrigation of one day and the first irrigation of the next. The published report does not give numeric values for the Small and Large arms. Readers therefore cannot tell how far apart the two treatments were, or how they compare with the vendor targets.
- The trial did not test substrate EC. It tested maintenance × dryback in two varieties.
The result: in both varieties the best yield came from No Maintenance plus Small Dryback, 392 g and 396 g of dry trimmed flower per plant. Regular Maintenance gave lower yields in every comparison. The picture was not uniform, though. In Gaia F1 under Regular Maintenance, the Large Dryback arm out-yielded Small Dryback (299 g vs 249 g).
How should that be read against the doctrine? A small dryback giving more weight is what vegetative steering is supposed to do, so on its own it does not overturn the phase plan. What it does is fail to support the generative-dryback assumption for yield in these two cultivars. Larger drybacks did not produce more flower in three of the four comparisons. THC was reported only as relative rankings, and those rankings went in opposite directions in the two varieties. So the trial says little about any quality benefit of generative steering.
Reasons not to build anything on it:
- No journal, no DOI, no peer review. It was published as a blog post by one of the commissioning parties, and we could find no independent source for its details.
- Commissioned by a substrate manufacturer and a seed company.
- THC was reported “in relative terms” only, with no absolute figures.
- Terpene data came from a single sample, and the authors themselves say it should not be generalised.
- Six plants per cell across eight cells is a pilot.
- No numeric dryback levels were reported, so the result cannot be translated into a target.
So the state of the evidence is this. There is a large body of confidently quoted numbers with no traceable trial behind them. There are a few peer-reviewed studies that support the general direction of stress-based steering but not the specific thresholds, and one of them shows a yield cost from high EC. And there is one small, unreviewed company trial that does not support larger drybacks for yield. That is not a foundation for a protocol.
4. What we are not saying
We want to be precise here, because this article could easily be misread as “crop steering does not work.”
We are not saying substrate sensors are useless. Knowing your substrate water content and EC in real time is clearly better than not knowing. Measurement is not the disputed part.
We are not saying the physiology is wrong. Water and salinity stress shifting allocation toward reproduction is well established in horticulture, and Caplan et al. found a cannabinoid benefit from controlled drought in cannabis.
We are not saying vendors are lying. They may well be right. Commercial growers using these protocols report good results, and a company with sensors in thousands of rooms has real data.
What we are saying is narrower and, we think, more useful: the specific numbers have not been validated by anyone independent, yet they are taught and repeated as though they were established. A grower who adopts a number because it is widely quoted and gets a poor result has no way to tell whether the number was wrong, wrong for their cultivar or wrong for their substrate, because there is no published basis to reason from.
There is also a specific reason for caution in this region. The studies above already show that responses depend on genotype. In Allred et al., the two cultivars responded differently to the same EC treatment. In the Grodan/RQS trial, the two varieties ranked the dryback arms differently under regular maintenance. We infer that crop steering thresholds are exactly the kind of parameter that will depend on cultivar and substrate. Numbers drawn mostly from North American commercial production in stone wool are the least likely thing to transfer unchanged to Thai and Southeast Asian growing conditions.
5. The tropical complication: low VPD in unconditioned rooms
This section applies to unconditioned rooms, greenhouses and net houses in the wet season. It does not apply to well-sealed indoor rooms.
Dryback is the net loss of water from the substrate between irrigations. Most of it is the plant’s transpiration, but drainage after the last shot and evaporation from the substrate surface also contribute. Transpiration, and to a lesser degree evaporation, depend strongly on vapour pressure deficit (VPD).
In the Thai wet season, a room or greenhouse without active dehumidification can sit at a VPD far below the targets usually recommended for flowering cannabis. At very low VPD the plant transpires slowly, and the substrate dries back slowly. In our view, a large overnight dryback target may then be hard or impossible to reach. The reason would not be a wrong irrigation strategy. The plant is simply moving little water.
A sealed indoor room with air conditioning and an adequately sized dehumidifier is a different case. It can hold VPD close to target all year, and there dryback should behave much as the vendor material assumes.
We infer two consequences for unconditioned spaces. First, a controller set to chase a dryback percentage it cannot reach may withhold irrigation for longer than intended, depending on how its logic is written. Check what yours does when a target is not met. Second, dryback stops being an independent variable you steer with and becomes partly an output of your humidity control.
Some vendor material discusses VPD alongside irrigation. What we did not find was guidance written for growers who cannot hold VPD near temperate targets for part of the year. Most examples assume a greenhouse or sealed room with a controlled climate.
6. What to do instead
These are our recommendations, stated as such.
Measure, and treat the numbers as your own. Install sensors if you can afford them. Log substrate EC, volumetric water content and dryback. Then build your own targets from your own crop rather than importing someone else’s.
Log VPD alongside dryback, always. In this climate the two are linked, and a dryback figure without a VPD figure is hard to interpret.
Change one variable at a time, across cycles, and keep a control: one group of plants on your current practice and one with the change. This is tedious, and it is the only way to generate real knowledge about your own crop.
There is no independently validated starting range to recommend. The practical starting point is your current irrigation routine, logged for one full cycle as a baseline. Make any change in small steps from there.
Treat “generative” and “vegetative” steering as hypotheses to test on your genetics, not as settings.
Box: measuring dryback and VPD
With a substrate sensor (VWC):
- Dryback = peak VWC after the last shot of the day − VWC just before the first shot the next day.
- Example: peak 60% VWC, morning reading 51% VWC → a dryback of 9 percentage points. Some controllers instead report dryback relative to the peak (9 ÷ 60 = 15%). Check which convention your software uses before comparing your figures with anyone else’s.
- Read the peak once drainage has stopped, and read it at the same point every day. Otherwise drainage will inflate your number.
- Put the sensor in a representative block or pot, not at the edge of the bench, and keep it in the same position all cycle.
Log VPD next to it:
- Record air temperature and relative humidity at canopy height at the same interval as VWC, at least hourly overnight.
- VPD (kPa) ≈ 0.6108 × e^(17.27 × T ÷ (T + 237.3)) × (1 − RH ÷ 100), with T in °C. This uses air temperature. Leaf temperature is more accurate if you have an infrared thermometer.
- Each day, write down the night’s average VPD beside that night’s dryback.
No sensors (most home and small growers):
- Pot weighing: weigh one marked pot (on a cheap hanging or platform scale) right after the last irrigation, once drainage has stopped, and again before the first irrigation the next morning. Overnight loss % = (evening weight − morning weight) ÷ (evening weight − weight of the pot with dry substrate and no water) × 100. The plant adds weight as it grows, so track the trend rather than any single figure.
- Runoff EC: measure the EC of what you feed and of the runoff each day. Runoff EC rising steadily above input EC shows salts building up in the root zone. Runoff close to input EC shows they are not.
- Keep a notebook: date, feed EC, runoff EC, evening and morning weights, and min/max temperature and humidity. One logged cycle is worth more than any number quoted online.
Be sceptical of any number offered without a trial behind it, including the numbers in this article. The figures above are sourced in the References. Where we are reasoning rather than reporting, we say so.
7. What would settle it
The gap is specific, and someone will eventually fill it: a peer-reviewed, cultivar-replicated, factorial trial combining substrate EC and dryback in cannabis, with numeric dryback levels, absolute cannabinoid and terpene figures, and adequate sample sizes. We did not find one. Until it exists, crop steering is a practice with a sound rationale and unvalidated parameters.
For this region the more tractable question is narrower: in rooms and greenhouses without dehumidification, does dryback-based steering work at all at monsoon VPD? Or does the strategy simply become unavailable for part of the year?
If anyone here is running substrate sensors in Thailand, what dryback are you actually achieving in the wet season? Can you hit a 10% overnight dryback at all when VPD is under half a kilopascal? Please say whether your room is sealed with AC and a dehumidifier, or open. That question has a factual answer, nobody seems to have published it, and it decides whether this body of practice even applies here for half the year. A single logged wet-season cycle, with VWC and VPD side by side, would answer it.
References
- Caplan D, Dixon M, Zheng Y (2019). Increasing inflorescence dry weight and cannabinoid content in medical cannabis using controlled drought stress. HortScience 54(5): 964–969. hort
- Allred J, Fatzinger B, Bugbee B (2025). Crop steering through osmotic stress can reduce height but reduced yield in medical Cannabis. Journal of Cannabis Research 7: 92. Client Challenge
- Saloner A, Bernstein N (2021). Nitrogen supply affects cannabinoid and terpenoid profile in medical cannabis (Cannabis sativa L.). Industrial Crops and Products 167: 113516.
- RQS Editorial Team (Royal Queen Seeds) (2026, 10 July). Plant Maintenance and Overnight Dryback: Effects on Cannabis Yield, THC and Terpenes (company blog; trial run by Grodan at CRIC Labs, Québec, with Royal Queen Seeds; not peer reviewed). Plant Maintenance and Dryback: Cannabis Yield and THC - RQS Blog
- Examples of vendor education content: AROYA, Mastering dry backs: The key to optimizing yields and plant quality, Mastering dry backs to optimize yield and quality | AROYA; TSRgrow, Crop Steering with Irrigation: Dry Backs, Shots, EC, and more, Crop Steering with Irrigation: Dry Backs, Shots, EC, and more
This article is technical horticultural information for readers operating legally in their jurisdiction. Cannabis cultivation is regulated differently in every country. In Thailand, cannabis flower is a controlled herb under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis) B.E. 2568. It may be dispensed only against a prescription, it may not be sold or advertised online, and cultivation and sale require licensing. Nothing here is legal advice.