White tips are an expensive way to find out your lighting strategy needs work. On fresh flowers, they can look almost deliberate: pretty, sometimes nearly white tissue perched on otherwise normal flowers. After drying, that novelty can become a brownish discoloration that does very little for bag appeal.
Even when the chemistry remains commercially useful, appearance can push flower toward a lower price or a different product stream. A lighting strategy that saves OPEX but erodes the value of your best buds might not be best for your bottom line.
For several years now, part of my team’s research has been focused on studying the causes of photobleaching in cannabis, otherwise known as white-tipping. The research connected it to red photon exposure, which suggested a biochemical explanation and, conveniently, a possible solution: add far-red light.
I liked this hypothesis. If true, it would have enabled us to create even more efficacious lighting solutions for cannabis cultivation while maintaining premium flower quality. It would be nice feather in my cap, validating the proposed mechanisms underlying cannabis photobleaching that I’d been presenting on for a couple of years now. And a nice bonus: I could finally stop being the bad guy, always saying that I’ve not yet seen compelling, data-supported evidence for any benefits to using far-red in cannabis cultivation.
Unfortunately, the plants were not particularly impressed.
Why We Want More Red
Red light is attractive because red LEDs can convert electricity into photons very efficiently. Increasing their contribution to a fixture’s spectrum can reduce the electricity needed to deliver a given photon flux. Across a commercial facility, that is worth pursuing.
But “more efficient light” and “more profitable flower” are not interchangeable. We need enough red to capture the economic benefit without crossing into a plant response that erodes crop value.
An earlier study we conducted approached that problem with a two-dimensional lighting gradient. Red percentage increased in one direction across the growing area, while photosynthetic photon flux density (PPFD) increased in the other. PPFD counts photons between 400 and 700 nanometers arriving per square meter each second. We measured the light quality and intensity at each individual plant in the study so we could correlate photobleaching incidence, depth, and visual severity with different aspects of the lighting.
Across six crop cycles, the analysis included 892 plants representing seven cultivars. We fitted curves describing how bleaching changed with both the percentage of red light and the absolute red photon flux. The stronger predictor was absolute red photon flux. Percentage alone did not tell the whole story. For example, 50% red at a PPFD of 800 delivers 400 µmol·m⁻²·s⁻¹ of red photons. The same percentage at 1,200 delivers 600. Those are meaningfully different exposures, even though the spectrum has the same name on the box.
Our work places the approximate onset of concern around 400–600 µmol·m⁻²·s⁻¹ of red photons, depending on cultivar. Even recognizing genetics as a confounding factor, our data model still only partially explains the response; there’s more to learn, but for now, our model gets us a pretty good idea of when bleaching happens.
A Proposed Biological Mode of Action for Photobleaching
Here is what (I thought) was going on. Flowers contain photosynthetic tissue, and chlorophyll can absorb more excitation energy than the tissue can productively use. Under sufficiently red-rich light, we hypothesized that this imbalance could contribute to chlorophyll destruction and the appearance of white tips.
To understand why far-red might help, we need to visit the light reactions of photosynthesis. There are two cooperating photosystems, photosystem II and photosystem I. Photosystem II comes first in the main electron-flow sequence. Pigments in the antenna complexes absorb photons and transfer excitation energy toward reaction centers. At photosystem II, light drives charge separation, and electrons are replaced by extracting them from water. Those electrons move through an electron transport chain toward photosystem I, where another input of light energy enables their onward transfer, ultimately supporting adenosine triphosphate (ATP) and nicotinamide adenine dinucleotide phosphate (NADPH) production. These go on to help make all the biochemical stuff plants need to live.
We suspected that photosystem I might be insufficiently excited relative to photosystem II. If that slowed electron throughput, absorbed energy could exceed the capacity for useful photochemistry, increasing the burden on protective mechanisms around photosystem II.
Plants have ways to dissipate excess excitation and limit damage. Those protections are not unlimited. Excited chlorophyll can enter a triplet state and transfer energy to oxygen, forming singlet oxygen, a reactive oxygen species. If protection is overwhelmed, oxidative damage can affect pigments, proteins, and membranes. Chlorophyll loss could then leave visibly pale tissue.
Why Far-Red Seemed Promising
Far-red light can preferentially excite photosystem I (figure 1). If electron flow was getting held up in photosystem I, maybe supplying more energy to that downstream photosystem via far-red would improve electron flow and reduce the pressure upstream. Research in lettuce has shown that supplemental far-red can improve photosystem II photochemical efficiency and net photosynthesis under the conditions tested. And while cannabis and lettuce plants are very different, a lot of the core biochemical machinery tends to be conserved between plant species.
Figure 1: Light reactions of photosynthesis. Electron flow is induced at photosystem II. We'd theorized that photosystem I was acting as an energetic bottleneck that could have been released via preferential excitation with far-red light.Courtesy of David Hawley
This reasoning is related to the Emerson enhancement effect. In the classic experiments, combining shorter-wavelength light with long-wavelength red or far-red light produced more photosynthesis than the sum of the responses to those lights separately. The result helped establish that two cooperating photochemical systems were involved.
Our hypothesis borrowed that principle of balanced excitation and asked a more specific question: Could balancing photosystem excitation via far-red prevent cannabis flower bleaching under a red-rich spectrum?
Far-Red Has a Following and a Variability Problem
Growers have explored far-red for managing crop architecture, accelerating flowering, and shortening production cycles. Much of that interest involves phytochrome, a family of photoreceptors that helps plants interpret their light environment. A shift toward far-red can signal competition from neighboring vegetation and encourage elongation.
More stretch could help with an exceptionally compact cultivar. However, in a room already short on headroom, it could also be an excellent way to create additional work. Whether the response occurs, how large it is, and whether it helps the business are separate questions.
Published cannabis studies illustrate that variability. Perotti and colleagues studied the CBD-dominant cultivar Mango under continuous or end-of-day far-red. They observed changes in plant architecture, including longer petioles under continuous far-red, but no change in flowering time.
This, and other similar studies have been cited to me as proof of far-red’s utility in cannabis, but the part that is always missing is the context. Their artificial PAR lighting supplied roughly 300–320 µmol·m⁻²·s⁻¹, initially supplementing greenhouse sunlight and later serving as the sole source indoors. The indoor flowering phase lasted just 12 days. That design addressed early development and flower initiation; it did not test a full commercial flowering cycle or measure mature flower yield. The indoor PPFD was also substantially below PPFD’s commercially relevant in today’s market. These are useful physiological observations, but their transfer to intensive commercial flower production remains a question.
In a separate study, Peterswald and colleagues studied Cannatonic, Hindu Kush, and Northern Lights under different photoperiods and far-red schedules. Their flowering treatments supplied 854 PPFD for the first eight hours, then 404 for the remaining light hours, with far-red applied around lights-off. Northern Lights showed increased total THC yield under one schedule, while the other cultivars did not show corresponding increases in total cannabinoid yield. Another schedule, extending far-red into darkness after a 12-hour day, strongly inhibited flowering.
Those experiments used small plants and 40 days of flowering treatments. They provide useful evidence of cultivar and timing dependence, but not a universal recipe. A result from one cultivar under one combination of intensity, spectrum, and photoperiod is exciting directional information, but needs much broader evaluation before it becomes a production recommendation.
Our experience reinforces that caution. We have conducted multiple far-red studies, including far-red sunsetting, far-red deployed throughout flower, and far-red as under-canopy lighting. While we’ve seen a couple interesting and positive responses (and a couple negative responses, along the way), we have not yet observed a reliable far-red response across that work. The challenge is finding a response consistent enough to build a lighting solution and cultivation strategy around.
Testing the Proposed Rescue
Our photobleaching experiment at Texas Original Compassionate Cultivation used one cultivar: TK’s Dream, a Blue Dream × ON Haze cross. We chose it because its moderate sensitivity to photobleaching made it useful for this question. It typically begins bleaching around a red photon flux of 600 µmol·m⁻²·s⁻¹.
We wanted to challenge that cultivar across a range of red exposures, then ask whether far-red could reduce the resulting bleaching. Our design was such that we’d be able to see exactly how effective far-red could be at reducing photobleaching and would inform us of how much far-red to theoretically add to one of our lights in the future.
We held our primary PAR PPFD at 1,350 µmol·m⁻²·s⁻¹ and deployed two perpendicular spectral gradients across the room. From front to back, the red fraction increased from 40% to 85%. That produced a red photon flux ranging from approximately 540 to 1,147 µmol·m⁻²·s⁻¹, spanning below and well above TK’s Dream’s expected bleaching threshold.
Across the other axis, far-red decreased from 300 µmol·m⁻²·s⁻¹ on the left to 30 on the right. All of that far-red was added on top of the background PAR. Heat maps of the lighting layout are illustrated in figure 2.
Figure 2. Left: Front-to-back gradient of red percent. Right: Left-to-right gradient of far-red light percent.Courtesy of David Hawley
The result was a room containing many combinations of red and far-red, with spectral and plant data collected at 126 positions. This let us relate photobleaching to the light actually reaching plants across the two gradients and investigate whether increasing far-red moderated the response to increasing red.
The experiment included two replicates, with both gradient directions reversed between them to account for any plant-position-in-the-room effects that might have been influencing development.
We also considered photon fractions within extended photosynthetic photon flux density, or ePPFD, commonly defined over 400–750 nm. This includes far-red immediately beyond conventional PAR. Using the stated background and added far-red fluxes, the combined total ranged from approximately 1,380 to 1,650 µmol·m⁻²·s⁻¹. Far-red therefore approximately 2%–18% of that total.
Measuring both absolute photon flux and spectral fraction matters. A percentage describes a mixture; the photon flux tells us how much light the plant receives. Our earlier red-light work had already shown why relying on percentage alone could be misleading.
Because far-red was added on top, total photon flux also increased along its gradient. That design tested the practical proposition of adding far-red to alleviate bleaching. It leaves additional questions about separating the effects of extra photons from those of their wavelength.
TL;DR: It Didn’t Work
Far-red did not appear to reduce photobleaching. At least, no more so than what we can achieve by just adding any other non-red PAR light. I had mentioned earlier that the absolute dose of red light is the main factor that predicts whether photobleaching will occur. What I didn’t mention is that the percent of red – particularly at higher total PPFD – also matters. It’s not nearly as important as that absolute dose, but the data indicate that, as you increase photon flux density (PFD), there is an opportunity to “dilute” red in a way that makes it a little less damaging to plants. The nuances of this require some refining on our part and really aren’t worth worrying about in most cultivation scenarios. The relevant takeaway here is that, at those higher PFDs, far-red is behaving similarly to non-red-PAR in diluting problems induced by red light.
The observation challenges the idea that insufficient photosystem I excitation is the main problem and that supplying far-red will resolve it. Perhaps another step limits electron transport, or flower tissue handles excitation differently than we assumed. I guess the good news is, I’m not going to run out of things to study in the foreseeable future.
For growers, the strongest practical lead remains understanding how much red a particular cultivar tolerates. Our earlier research places that concern in roughly the 400–600 µmol·m⁻²·s⁻¹ range.
Being Wrong Productively
One of the hazards of a rich biochemical education is that you can convince yourself of a very satisfying explanation for something a plant has absolutely no intention of doing. Still, progress is progress; we’ve now learned one more way to not solve photobleaching!
I still like hypotheses that connect “torque on the screw” biology to cultivation economics. And I can’t say too much just yet, but we’ve got a new working hypothesis for exactly how red light is inducing photobleaching, and what we might be able to do about it. Stay tuned, and thanks, as always, for supporting cannabis science!
David Hawley heads research, development, and horticultural services at Fluence, where he directs the company's global cannabis lighting research program. He is the inventor of Fluence's Red Sandwich method, which cuts cannabis lighting operating costs by 26%, and a co-inventor of the T48 method, which cuts them by roughly 20%. He also led Foursquare, the multi-year spectrum study behind Fluence's R-series lineup. Hawley holds a Ph.D. in Environmental Science and publishes peer-reviewed cannabis lighting research.
