The Cannabis Cultivation Strategies of the Future
Cannabis has been cultivated by humans for thousands of years across a huge number of different cultures. But despite this long history, the golden age of cannabis cultivation innovation may just be beginning.
As new regulations have come and paved the way for the development of a medicinal cannabis market, a new focus has arisen on developing cultivation techniques that allow for consistent and easily replicable growth. At the recent Annual Symposium on Medical Cannabis, hosted by the Danish Technological Institute and the national investment promotion agency Invest in Denmark, attendees heard the latest cultivation research breakthroughs from top cannabis scientists and explored what the future of large-scale cannabis might look like.
Propagating the cannabis plant
In recent years, tissue culture techniques and micropropagation methods have significantly altered how large-scale cannabis cultivations operate. These methods can save cultivators a significant amount of floor space while increasing peak multiplication rates and making it easier than ever to produce certifiable virus-free plants. But within these methods, there is still room for further optimization.
“Stage zero [in micropropagation] is the selection and maintenance of stock plants, this is your mother plant that you're going to start with,” Dr Max Jones, an associate professor at the University of Guelph, told audiences during his presentation “Plant Sciences - Medical Cannabis.”
“[Then] we take a small piece of tissue, usually a node in the case of cannabis, and we sterilize it. Then we go to Stage two, and this is really where the magic happens. In stage two, you take one plant and grow it into ten. Take those ten plants, grow them into a hundred, then a thousand, then ten thousand – and it’s the exponential growth that’s really important to make micropropagation work.”
But when following published research papers on cannabis propagation, Jones and his colleagues were not seeing this exponential growth pattern. They found an initial flush of growth, but then, after a few generations of micropropagation, the plants would begin to develop health problems and die off.
“The question here that we were asking ourselves was, can we use flowering plants to increase multiplication rates?” asked Jones. “And the answer is yes.”
Instead of propagating using nodes taken from a mother plant, Jones and his colleagues began to experiment with using light cycles to induce flowering in cannabis plants, and then propagating tissue samples from these flowers to maximize observed multiplication rates.
“We have our plant, we put it under short days to grow flower, we take the flowers, chop them up, plate them out, and then we get more plants. And so this is a micropropagation system using flowering plants,” Jones explained.
After running calculations, the group determined that, for the same amount of labor, this method could achieve high multiplication rates that would be economically feasible for cultivation operations to adopt.
“But what’s important is that micropropagation through flowers should also be genetically pretty stable,” Jones added.
Applying machine learning to cannabis cultivation
While cultivators are already beginning to apply these propagation techniques to their operations, researchers are still looking to further improve the quantity and quality of cannabis crops that can be produced commercially. To this end, the University of Guelph researchers and their collaborators are now applying machine learning algorithms to different aspects of the cannabis cultivation process.
“We’ve come into the 21st century trying to modernize research, and we’re trying to use machine learning,” Jones explained.
“Humans are fundamentally not very good at seeing patterns and big sets of data. It takes many years of experience, and that takes a lot of trial and error. This is where machine learning comes in. Computers are really good at this kind of stuff; if you feed them enough data, they can sort through the patterns and they can run simulations.”
The first of these machine learning studies from the University of Guelph group looked at modeling and optimizing seed germination, which is a fairly simple process with relatively straightforward datasets, according to Jones. Now the group has also looked at optimizing a disinfection protocol and scarification method for seed germination, as well as optimizing the light and carbohydrate sources needed for in vitro shoot growth and development.
“You can change the amount of light, you can change the spectrum of light, it’s actually a very complex thing. There’s no way we could actually test the thousands and thousands of combinations that are possible,” said Jones.
Instead, the researchers developed different treatments using unique combinations of light and sugars, then measured the outcomes when these treatments were used.
“So you have your input dataset, you put it into the machine, you model it using a few different machine learning algorithms, and then use an optimization algorithm, which tells you the theoretical optimum combination,” explained Jones. “And what I find really cool about machine learning is that the prediction that comes out is often things that you never even combined in the first place.”
Simulating stress for ideal plant growth
But improving upon cultivation strategies requires more than just maximizing the quantity and basic qualities of cannabis plants. With the growth of the medical cannabis sector, it is also important that cannabis plants are highly standardized and contain predictable cannabinoid profiles across the entire crop.
“When we’re dealing with a medical plant which needs to support a medical industry, the need to develop very, very high chemical standardization is crucial,” Dr Nirit Bernstein told conference attendees during her “Cultivation Strategies of the Future” presentation. Bernstein is the principal researcher at the Institute of Soil Water and Environmental Sciences at the Volcani Research Center, Israel.
“We are looking at standardization throughout the plant, to make sure that an inflorescence on the top [of the plant] is similar to an inflorescence on the bottom in terms of chemical composition. Otherwise, what the patient will receive will be very different between one batch he received to another.”
Even very minute changes in cultivation conditions can have a significant impact on cannabis’ secondary metabolism, Bernstein explained. As a result, researchers are largely concerned with trying to understand what is responsible for this lack of uniformity observed between inflorescences, so that the effect can be reverse-engineered to create effective cultivation conditions that promote high levels of standardization.
“One of the things that we checked several years ago was trying to introduce humic acid,” said Bernstein. “Why humic acid? Because there was no information at that time about the cultivation condition of plants, but many of the growers used to apply humic acid with the belief that it was going to increase production of the secondary metabolites.”
They found that humic acid actually reduced the natural variability of the cannabinoids across the plant, meaning that the inflorescences at the top and bottom of the plant resembled each other more closely in terms of cannabinoid composition. However, this improved uniformity came at the expense of moderate reductions in the levels of CBD and THC being produced by the plants.
“What we’re working on now in the lab and the protocol that we are putting together is that, by applying different exogenous treatments, we are able to develop plants which function at their optimum,” said Bernstein. “We trick them to think that they are under stress conditions and therefore, secondary metabolite production is highest.”
The plants can be tricked in this way through many different means; some very low-tech, such as manually shaking the plants, and others that are a little more high-tech, such as altering the nutrient levels that are fed into hydroponic growth media. When the plants “believe” that they are stressed they will produce higher levels of cannabinoids. But since this stress is being artificially applied, the plants are still able to grow in a normal, healthy manner.
“In terms of the breakthroughs and innovations which are needed, we need development of arable technologies for the improvement of cultivation,” Bernstein concluded. “So some of this will include more sensors and controllers because cannabis is all about precision agriculture.”
“It is understanding the plant response and dealing with precision agriculture directed to achieve the physiological state of the plant that we are interested in. Development of advanced technologies for directing the chemical qualities, and so interfering with the physiological responses of the plant for achieving specific biochemical responses. And, of course, controlling the chemical quality and the profile for specific medical indications.”
This article originally appeared in Analytical Cannabis' Advances in Cannabis Cultivation eBook in December 2021.
