Soil, Coco or Hydro? Wrong Question. Here’s the Right One.

R07 Synthesis for Everyone — the draft public article

From the Cannabis Cultivation Knowledge Engine. Derived from synthesis v1.0 (signed off 2026-09-11). Every factual statement traces to a verified claim in the R07 ledger (IDs like C013); sources in the plain-English ledger. This article is not evidence; see the ledger.

Ask ten growers whether soil, coco or hydro grows the best cannabis and you will get ten confident answers and a fight. Hydro grows faster and yields more. Soil tastes better. Coco is the best of both worlds. Hydro is unforgiving; soil forgives mistakes. Hydro uses 90% less water. Rockwool goes to landfill. You can’t grow organic in hydro. Sterile media mean fewer root diseases.

We went looking for the science behind all of it. What we found is that the science never asked the question that way, and once you see why, the fight stops making sense.

The labels dissolve on contact

Not one study we located compared “soil,” “coco” and “hydro” as categories [C001]. What the studies measure is a system: what the root zone is made of and what grade it is, how big the pot is, how it drains, whether the water drips, floods or recirculates, how often it is fed, whether the nutrients come dissolved or have to be broken down by microbes, how warm and oxygenated the roots are, what is in the water, and what is in the feed. Outcomes belong to that whole bundle.

A practitioner-record passage is withheld from the public copy.

Then the second casualty: “soil” is usually soilless too. Almost every “soil” arm in the research is a bagged peat mix, which is the third class, not the fourth. No indoor trial of a native-soil or amended bed exists [C034]. Danny’s licensed San Jose greenhouse, growing in bagged peat “soil,” was the same class [C035].

The one experiment, and why it can’t say what people want it to say

Only one controlled trial has grown cannabis in three root-zone systems side by side. At a licensed facility in Ontario, an 11-litre peat pot fed a synthetic solution, labelled “hydroponic” by the authors, grew 42–116% more flower than the other two setups: an 11-litre peat pot fed fish-farm effluent, and a 3-litre peat pot with its lower roots in an aquaponic tank [C002].

Read that carefully. The “hydroponic” winner was a peat pot. The three arms differed at the same time in nutrient source, pot size and plumbing. The gap is real and well verified; nobody can say which of those differences caused it. It is the textbook example of a bundle comparison, and it cannot support “hydro beats soil” or “peat beats aquaponics” or anything else with a single label in it [C002].

The three things that actually change outcomes

Once the labels are gone, three variables keep showing up in the measurements.

Which variety you are growing. In a German trial across peat, green fibre and coir, one variety lost more than half its flower in coir (3.8 g per plant against 8.6 g in peat) while another held 8–9 g in all three [C007]. On rockwool, salt-stress “crop steering” cut flower yield about 20% in a pattern that depended on the cultivar, with no potency gain to show for it [C014]. A substrate ranking from one variety does not carry to the next.

How much water the root zone makes available. Across five coir/peat/perlite mixes on the same feed, the mix’s available water was the main thing that predicted growth [C009]. Within coir, the drier of two formulations gave 29% more dry flower in the flowering stage [C010]. But the growth-stage study favoured more available water and the flowering study favoured less, so “drier is better” is not a rule; the direction changes with stage. When three media were compared under their own watering schedules, the authors attributed the differences to water-holding and irrigation management, not the material, and the CBD percentage barely moved [C008].

How the water is delivered. With the same clay pebbles and the same feed, recirculating the solution instead of draining it to waste raised the total THC harvested, mainly after week nine, but did not raise flower weight [C013]. That result comes from one clone, in clay pebbles, and the two setups ran one after the other rather than at the same time, so it is a lead, not a law. Outdoors, burying the drip line cut irrigation water 18.6%, with slightly more flower and CBD [C017]. Plumbing is its own dial.

Coco is not one thing

The property behind “coco is the best of both worlds” is measurable, and measuring it dissolves the slogan. Across 13 coir sources, salinity ran from 39 to 597 mS/m and the fibre’s capacity to hold nutrient ions from 31.7 to 95.4 cmol(c)/kg, and both varied within a source [C003]. A “coco” number without the source, grade, compression history and pot size means nothing.

The calcium story is half true. Fresh, unbuffered coir arrives with its exchange sites loaded with sodium and potassium and can pull calcium and magnesium out of your first feeds; rinsing removes loose salts but does not buffer [C004]. But nobody has measured how much cannabis loses this way, and no cannabis trial has compared buffered with unbuffered coir. “Coco always locks out calcium” is rejected as a rule. If you run reverse-osmosis water, as Danny mostly did, this is the situation where it matters most, and his sites never had the water tested [C004, C036].

What the evidence says about the ten claims

Hydro grows faster and yields more. No cannabis study reports days to harvest or cycles per year by root-zone system with variety and environment fixed. “Faster” has no measurement and no measured mechanism [C018]. The one yield comparison is the bundle above [C002].

Soil tastes better. No blind taste test or trained panel has ever compared soil, soilless and water-culture cannabis [C022]. Chemistry does differ by system; nobody has shown it changes what people taste [C023].

Coco is the best of both worlds. Coir grade, water availability and container capacity produce measured gains and penalties depending on variety and stage [C003, C007, C009, C010]. Not evidence-based as “best.”

Hydro is unforgiving; soil buffers mistakes. A plausible idea from general substrate physics. No cannabis dataset measures how fast problems appear or how fast they can be fixed in one system versus another, and those are two different things [C020, C021].

Hydroponics uses 90% less water. That number comes from lettuce greenhouses versus fields. No cannabis study meters water by root-zone system. The real cannabis savings, deficit watering and buried drip, come from how you irrigate, not what you grow in [C029, C016, C017].

Rockwool goes to landfill. Plausibly, but nobody has weighed spent media per unit of flower or tracked where it goes [C030].

You can’t grow organic in hydro or coco. False as an absolute. Organic liquid feed works in coir with a measured dose curve, and aquaponic cannabis has been grown to harvest [C028, C011]. The catch is real: organic nutrients need microbes and time to become available, and water culture or inert media only supply that through a biological compartment. Whether it competes is the bundle trial again. And there is no USDA-organic cannabis; California’s OCal is the “comparable-to-organic” scheme [C027].

Sterile media mean fewer root diseases. Rejected by two independent labs. Disease-causing Pythium and Fusarium were recovered from coir and rockwool under sick plants in Connecticut, Pythium root rot was confirmed in British Columbia greenhouses, and Fusarium was found in a facility’s recirculating solution [C024, C025]. But “soil has fewer diseases” is equally unsupported: nobody has ranked disease by medium on a fair basis.

Living soil has a lower footprint. Coir-versus-peat life-cycle rankings flip depending on geography and what you measure, and they are per cubic metre of medium, not per kilogram of flower; rockwool was not even in them [C031].

A practitioner-record passage is withheld from the public copy.

Practitioner-record section withheld from the public copy.

What nobody has measured

No matched trial of days-to-harvest or cycles per year by system [C018]. No blind taste test [C022]. No pot-size trial with everything else fixed [C006]. No buffered-versus-unbuffered coir trial in cannabis [C004]. No disease ranking by medium [C024]. No multi-cycle living-soil bed versus soilless comparison [C034]. No root-zone temperature or dissolved oxygen logged by system in any located trial [C025]. No spent-media inventory, no nutrient-discharge measurement, no cost-and-labor comparison [C030, C029, C032]. Six of the key papers have only been read as abstracts. Every trial left the irrigation water chemistry unreported.

So what should a grower do?

Choose a root-zone system by the properties you need to control, not the label, and expect the literature to have measured few of them [C001, C021]. Treat coco as a material with a source and a grade, and neither assume calcium lockout nor assume its absence [C003, C004]. Try your variety in your system before you scale, because rankings do not transfer [C007, C014]. Manage water availability and pot size as yield variables, knowing the preferred direction changes between growth and flowering [C009, C010]. Take water savings from how you irrigate, which has evidence, not from the medium, which does not [C016, C017, C029]. Assume no medium is pathogen-free and none is proven cleaner; test the crop, not the category [C024]. Step your pot sizes up rather than drowning a small plant in a big one, knowing that is convention rather than a measured effect, and expect overwatering to be the commonest failure [C006, C036]. And treat every “faster,” “cleaner” or “more forgiving” you hear as a question about a property nobody has yet measured in cannabis.

Cannabis root zones are well described as systems and unmeasured as categories. Grow accordingly.