Skip to content
Lead: Equipment marketing promises spectrum miracles, UV boosters and
perfect photoperiods. The research of the past five years paints a much
clearer, almost unromantic picture: one factor beats nearly all the
others, and it is not the one people talk about loudest.
What this is about
If you grow cannabis indoors, lighting eats the largest share of your
running energy budget. So the temptation is strong to squeeze out a few
more percent through spectrum tweaks, UV boosters or new photoperiods.
We went through nine open-access papers published between 2019 and 2024,
almost all of them from the big platforms Frontiers, PLOS and
HortScience, and asked: what is solid evidence, what is marketing, and
which levers are worth pulling for a grow in Germany?
What works: PPFD is king
If you boil the nine papers down to one statement, it is this one:
more light (within sensible limits) means more flower. The
relationship is surprisingly linear, and in the studies reviewed here no
saturation was reported, not even at the highest levels tested.
Rodriguez-Morrison et al. (2021) show this clearly in a controlled
indoor study: flower mass, apical flower density and harvest index all
rise linearly with PPFD. Worth noting: cannabinoid concentration did
not change in this trial. So if you are chasing THC or CBD
percentages, higher PPFD
will not get you there. If you are chasing total cannabinoids per unit
of area, it will, because more flower multiplies through.
Llewellyn et al. (2022) go one step further: in a comparison of high
versus low PPFD, total biomass came out 1.3 to 1.5x higher, and flower
even 1.6x.
Under strong light the plant does not just produce more material, it
also shifts more of it into the flower (harvest index +7 %).
Huber et al. (2021) remind us that light rarely works on its own:
raising DLI combined with elevated CO2 produced 24 to 33 % more
growth at comparable energy input. So if you are watching the
electricity bill, think of light and CO2 together.
Take-away: if you want to maximize one lever as a grower, PPFD or DLI
is the safest bet, provided your climate management (temperature, VPD,
CO2) keeps up.
What matters less: the spectrum
This is where it gets interesting, and uncomfortable for part of the LED
industry.
Westmoreland et al. (2021) held PPFD constant across three consecutive
studies and varied only the blue fraction, from 4 % (HPS) to 20 %
(various LED mixes). The result: more blue lowers yield linearly, by
12 % in total. In these trials, spectrum had no effect on THC or CBD
concentration.
Even more relevant in practice: the white+red LED setup produced 4.6 %
less per square meter than HPS, but 27 % more yield per euro of
electricity spent, because the LED efficacy (µmol per joule) was
considerably better. In other words: the win was not the spectrum, it
was light output per watt.
Bilodeau et al. (2019) and Stamford et al. (2023) are both reviews
with a grower focus, and they put this in context: spectrum strategies
have their place, but mostly for plant shape and habit, not for a
jump in yield or potency. Stamford explicitly models the electricity
cost of different LED “recipes” and offers rules of thumb for when a
given approach pays off.
Take-away: spectrum optimization is a fine-tuning tool, not a yield
lever. If you are spending money, look at fixture efficacy first.
What is probably overrated: UV-B
UV-B lamps have been sold as cannabinoid boosters for years.
Llewellyn et al. (2022) tested this in the same study as the PPFD
question and came back with a sobering answer.
UV-A plus UV-B during the last 20 days of flowering did raise THC
concentration in the sugar leaves (the small leaves around the flower)
by roughly 30 %. But total THC per tissue stayed the same, and in the
flower itself the study found no statistically significant increase in
cannabinoids. The
authors put it plainly in their conclusion: no commercially relevant
benefit.
That does not mean UV-B has no physiological effect. It means the effort
does not clearly pay for itself economically, at least not in this
configuration.
What most people get wrong: 12:12 as gospel
The standard practice of “flowering starts with 12 hours of light” comes
from grow convention rather than from data, and in many cases it is
suboptimal.
Peterswald et al. (2023) ran nine photoperiod treatments across three
cultivars: one CBD line (Cannatonic) and two THC lines (Northern Lights,
Hindu Kush). The result:
- In all three cultivars, 14L:10D produced the highest flower yield.
- In the THC lines, however, THC concentration dropped under a static
14L:10D. - In the CBD line everything lined up: higher yield AND higher CBD
concentration, adding up to 50 to 100 % more total CBD.
Ahrens et al. (2023) support that picture: in most cultivars flowering
can also be triggered at 13 or 14 hours, and some manage 15. The yield
response is often quadratic, with a definable sweet spot that sits
between 12 and 13 hours depending on the cultivar.
Take-away: photoperiod is cultivar-specific. Treat twelve hours as a
universal law and, depending on the cultivar, you leave 30 to 50 % of
yield or cannabinoid output on the table.
Bonus: greenhouse growing
For those of you who do not work purely indoors: Collado et al. (2024)
looked at supplemental lighting in a greenhouse. The result: a clear
increase in photosynthesis, water use efficiency and growth, with no
saturation reached. Same story here: more light, more plant.
Practical take-aways
- Lever 1 (safe): push PPFD or DLI up as far as your climate (CO2,
temperature, VPD) can follow. It scaled linearly in most trials. - Lever 2 (cultivar-dependent): test the photoperiod. 13 or 14 hours
can deliver noticeably more flower or cannabinoids in some cultivars.
Try it on a small batch first. - Lever 3 (on the investment side): when buying fixtures, look at
µmol per joule first, not at spectrum recipes. - Lever 4 (stay skeptical): UV-B as a cannabinoid booster is not
clearly supported by the evidence so far. Run your own tests before
buying hardware.
Sources
- Rodriguez-Morrison et al. (2021), Front. Plant Sci.,
doi:10.3389/fpls.2021.646020 - Westmoreland et al. (2021), PLOS ONE,
doi:10.1371/journal.pone.0248988 - Llewellyn et al. (2022), Front. Plant Sci.,
doi:10.3389/fpls.2022.974018 - Peterswald et al. (2023), Plants, doi:10.3390/plants12051061
- Ahrens et al. (2023), Plants, doi:10.3390/plants12142605
- Bilodeau et al. (2019), Front. Plant Sci., doi:10.3389/fpls.2019.00296
- Huber et al. (2021), Front. Plant Sci., doi:10.3389/fpls.2021.615853
- Stamford et al. (2023), HortScience, doi:10.21273/hortsci16823-22
- Collado et al. (2024), Front. Plant Sci.,
doi:10.3389/fpls.2024.1371702
