The selection of the starting material is always crucial. While many factors contribute to cultivation, especially when dealing with complex systems such as CEA infrastructures, the starting material is the very first step and dictates most of the grow performance. Usually, the choice is between seeds and propagated material. Many growers prefer using clones rather than starting from seeds, as genetic variability in a batch of seeds is too high to ensure uniform plant development. This is a well-established fact in the cannabis industry and is widely accepted. Most growers are fully aware of this when making decisions about starting material. This was the premise of the speech given by Dominique van Gruisen from Innexo at the Cannabis Summit in The Hague. According to him, this is about to change.
Unpredictable and unacceptable
Innexo is a research organization that conducts trials ranging from developing substrates to pressure testing genetics. He began his talk by pointing out that the unpredictability of seeds is essentially unacceptable in an industry moving toward pharmaceutical precision. "We want to be able to scan the seeds, to bring a new level of control to precision cannabis," he said.
He recounted how, months ago, he was speaking with an acquaintance who explained that it is possible to sort tomato seeds based on the vitamin C content they produce. Dominique then realized that technology already existed in other agricultural sectors for scanning seeds -thanks, in part, to advancements in AI. All he had to do was apply that technology to cannabis. Easier said than done, of course, but that didn't stop him from setting up a proof-of-concept trial. In his own words, he wanted to "show such a thing wouldn't work on cannabis." A healthy dose of skepticism may have been necessary, as the results he obtained ended up proving his initial assumption wrong.
To analyze what's inside a seed, they had access to a custom data collector that, with the help of cameras and AI, gathered around 500 data points for each seed. These seeds were then labeled, and different characteristics were assessed, such as germination potential, flower color, flower development, and whether the seed would produce a male or female plant.
Relying on such information before putting the seed into the soil - or any other medium - could be a game-changer for the industry. First of all, seeds are less costly than clones, which would reduce operating expenses. On top of that, growers wouldn't necessarily need a nursery on-site for propagation, allowing them to focus on maximizing canopy space.
Returning to the data points this technology collects, it goes without saying that knowing each seed's characteristics in advance represents a huge advantage. First, growers could select only those seeds that are 100% guaranteed to germinate and do so uniformly. When it comes to flower color, growers could choose seeds that will produce flowers with a unique hue to stand out in the market. On the production side, flower development is one of the most critical aspects. Given the industry's high level of control and push for efficiency, it's essential for all plants in a grow to develop their flowers in unison to keep processes streamlined.
Regarding the ability to determine male from female plants, Dominique recounted the discovery of the feminization process. While feminized seeds have been crucial for growers, they also come with issues - such as the risk of developing hermaphrodites due to stress factors. Scanning seeds the way Dominique demonstrated would allow growers to cultivate plants from seeds that are naturally female.
Reliable seeds
"For each cultivar, we can create sorting algorithms to generate data for production," he explained, describing how this "seed scanning" process could be implemented in the current industry. In other words, seed companies could differentiate themselves by incorporating seed scanning into their production line, offering seeds that are 100% guaranteed to possess specific traits. This technology wouldn't only benefit growers but also seed companies, which could specialize in specific genetics with desirable characteristics. "We could even improve F1 hybrids and empower seed companies and producers to avoid fraudulent seeds in their batches," he added.
One might think it will take years for such technology to become viable for cannabis. However, that's not the case. "In six to twelve months, we'll have a high-throughput commercial sorting," Dominique said. "This is happening right now. This technology is already being used in other crops - very successfully, too." This explains why it won't take long before seeds with fully validated cultivation characteristics become available. As Dominique pointed out in his final remarks, "Every seed from now on can be the best seed."