This fall, for the first time since 1970, farmers throughout the United States were set to harvest federally legal hemp. From New York to Kansas to Oregon, they had applied for licenses to plant nearly half-a-million acres of the non-psychoactive cannabis. Farmers had purchased seeds and tended fields throughout the summer. But toward the end of the growing season, many of them realized that their plants weren’t turning out as expected — and at worst, that their entire crop would have to be destroyed.
While the corn and soybeans had, as usual, sprouted all at once and grown to a uniform size, new hemp plants tended to grow on different schedules, to different heights. Sometimes there weren’t that many hemp plants at all, because a relatively small percentage of the seeds had sprouted. The most dramatic surprise, however, was to farmers who had intended to plant hemp with a high concentration of the chemical compound known as CBD — a potentially lucrative prospect given the ingredient’s addition to everything from lattes to pet treats — and instead ended up growing what the federal government classifies as marijuana.
All cannabis plants, including hemp, have some amount of the psychoactive chemical compound THC. Cannabis sold legally as marijuana may have as much as 20% THC on a dry-weight basis, while farmers growing hemp for textiles, food, or CBD are required by law to meet a much lower limit of 0.3%. Neither a 0.3% THC cannabis plant nor a 0.4% THC cannabis plant will get you any higher than smoking grass from your lawn, but in the eyes of the law, that tiny tenth of a percent makes all the difference.
To comply with federal rules, states must ensure the “disposal” of a hemp crop that tests over the 0.3% limit, a process that usually involves plowing it down or burning it on-site. A draft of new hemp-growing guidelines proposed by the USDA, the first such guidelines from the federal government, establishes a margin of error for meeting the THC limit, but is strict about the consequences of failing the test. If the rules go into effect as written, farmers with what the industry calls “hot” crops that exceed the 0.3% limit will be responsible for paying an estimated $200 per acre to destroy their own crops, on top of losing the commercial value of the plants. This is true even if the grower is operating in a state that has legalized recreational marijuana, which requires a different type of license.
The problem is that while the race to plant hemp has taken off at a sprint, the race to study hemp is a marathon that is just getting started.
Almost 90 hemp sites in Colorado have tested “hot” this year, the state department of agriculture told OneZero. The Oregon Department of Agriculture said eight hemp sites have tipped the limit, while Hawaii reportedly destroyed more than half of its hemp crops because of high THC content. Meanwhile, the advocacy organization Vote Hemp estimates that only 50% to 60% of the hemp planted this year will actually be harvested, due to what it calls “crop failure, non-compliant crops, and other factors.”
The problem is that while the race to plant hemp has taken off at a sprint, with 17,000 farmers applying for licenses this year alone, the race to study hemp, which has also been essentially banned wherever the crop was illegal, is a marathon that is just getting started. Published research, seed certification processes, and understanding of the cannabis plant’s genome are all lagging behind the plant’s rapid commercial adoption.
At a time when the genetics of other crops are so well developed that corn farmers can choose between certified seed that will be ready to harvest in 79 days, 84 days, or 89 days, hemp farmers have no way to know for sure how much THC any variety of cannabis plant will produce under a variety of conditions — not to mention whether they’re buying high-quality seed at all. As Michael Bowman, the co-founder of a public benefit corporation that partners with small hemp growers, puts it, “it’s been a crap shoot.”
Hemp has been cultivated for use in textiles, paper, and food since ancient times, and it was once a staple crop in North America. George Washington famously grew a variety of industrial hemp on his farm in Mount Vernon, and predicted that the crop would eventually become more profitable than tobacco. (He was wrong — so far, at least.) While the last revival of commercial hemp farming petered out in the United States shortly after a call for farmers to plant “hemp for victory” during World War II, and hemp became a controlled substance in 1970, plenty of remnants of the plant’s agricultural heyday are still growing in ditches and behind farm fields.
George Weiblen, a plant biologist at the University of Minnesota who studies hemp, knew exactly where these feral hemp populations were growing when he began applying for permission to study the plant in the early 2000s. But he wasn’t permitted to cut so much as a single leaf, even after he became one of the first biologists granted permission in 2002 by the Drug Enforcement Administration to conduct cannabis research.
Weiblen was initially focused on the possibility of creating a cannabis plant that was completely THC-free. It was slow work. “We are still interested in that,” he says, “it just turned out to be a lot more complicated. We had to first understand the genetic basis for the drug content, and that took 15 years.” Even samples used to analyze the plant’s DNA had to be kept under lock and key — and still must be if the plants contain more than 0.3% THC, because that means they were derived from a federally illegal drug.
In 2014, changes to the farm bill allowed states to create their own programs for legal hemp cultivation and to grant scientists licenses to run pilot research projects. Weiblen was finally able to collect wild hemp as part of a separate project licensed through the state of Minnesota. Still, he and other scientists had to follow strict rules about where they collected samples, how many plants they grew and who had access to them. “You apply, you submit protocol, investigators come out and look at where you’re gonna do it — there’s a lot of record keeping” says Weiblen. “It has not been feasible to do much in cannabis because of all the regulatory requirements.”
Whether hemp could cross state lines, for instance, was still a legal gray area, which made collecting diverse samples of hemp plants challenging. Larry Smart, a professor at Cornell University, kicked off a preliminary research program for hemp in 2016, a year after the state passed its hemp bill. If he had been studying, say, apples, he would have called up the library of seed and plant tissue that the USDA maintains for exactly this purpose, which is conveniently located right on Cornell’s Geneva campus. It contains genetic material from thousands of varieties of apple categorized by traits like “fire blight resistance.” He would put in a request, and if it were approved, the library would collect budwood for him and send it over for free.
The USDA used to have a similar library for hemp — the database used to track samples still returns 150 entries — but it was presumably destroyed when growing hemp became illegal. All of the entries are now listed as “unavailable.” So instead, when Smart goes looking for hemp to plant in his trial plots, he says, “I have to call quote-unquote breeders in Colorado and actually buy seed from them. At exorbitant prices.” While an entire bushel’s worth of corn seed will run farmers less than a dollar, some hemp varieties can cost $1 or $2 per seed.
That’s if Smart is lucky. Before the 2018 Farm Bill went into effect, hemp breeders could not patent or protect what was still considered an illegal product by the federal government. Even though patents and equivalent plant protections are becoming available for the plant, the culture of secrecy that is characteristic of a long-underground industry remains. The creator of one hemp variety that is widespread in New York, for instance, refuses to sell it to Smart for fear that the researcher would somehow steal or reveal its secrets. “I would like to be able to provide data, comparative data, side by side of those cultivars in the same field trials with other cultivars, so that I can present to growers unbiased results in terms of yield and performance,” he says. “If they’re not willing to let me include those in my trials, I can’t provide a full analysis of what is commercially available.”
In August, U.S. Senator Charles Schumer announced $500,000 of federal funding to build a hemp seed bank co-located at Cornell’s Geneva campus. But it will take years to develop the sorts of resources for hemp that are already in place for other crops like apple.
While states have been permitted to run hemp programs since 2014, the enormous demand for CBD — which grew an estimated 706% this year — and the legalization of hemp on a federal level have accelerated adoption of the crop. In 2018, the growing season before the new Farm Bill made it legal to plant hemp anywhere in the United States, hemp production in the country tripled. When the 2018 Farm Bill went into effect this season, the growth continued. In Colorado, around 1,200 farmers were granted hemp licenses in 2018. This year, the state department of agriculture told OneZero it had registered about 2,600. Oregon also jumped from around 1,100 farmers to 3,000. In states that had not previously opened their own hemp licensing programs, like Illinois and Kansas, farmers planted for the first time.
“All the sudden in Illinois, it was like, ‘ta da, you’re legal to grow,’” says Winthrop Phippen, a professor of plant breeding and genetics at Western Illinois University. “The permitting process didn’t open until May, so there were last-minute situations.” He only knew of one producer in Illinois selling hemp seed. Instead, “everyone had to outsource from somewhere.”
Some aspiring hemp farmers went on road trips to buy seeds, or bought them online. Many went with the cheapest option, but with the risk that they might be getting duds. “When we talk about corn and soybeans, we’ve got corporations with 50 or 60 breeders working on different traits,” Phippen says. “Most of the hemp stuff has been backdoor, home gardeners working on traits.”
Even a decent hemp seed is only the first step to a viable crop. How much THC a hemp plant produces is affected by environmental factors as well — a seed may perform differently in Kentucky, for instance, than it does in Colorado. Why? Scientists haven’t performed enough research to say. “We’re in the middle of a dry spell here in Kentucky, and people say ‘that’s going to raise the THC level,’” says Bob Pearce, a professor at the University of Kentucky who leads the school’s hemp research program. “We don’t really know that for sure… We can infer from our experience with other crops, but we don’t know that hemp will work in the same way.”
For crops like soybeans and corn, farmers can buy seed that produces a plant with known characteristics that is certified by a third-party who ensures it is grown according to set standards. But high-CBD varieties come almost exclusively from the United States, which has only had federally legal hemp for a year — not enough time to determine if a variety is eligible to be certified by the Association of Official Seed Certifying Agencies (AOSCA).
In order for a plant variety to be eligible for certification, plant breeders must take carefully documented steps to prove that it is distinct, that it has uniform characteristics, and that it is stable from one generation to the next, meaning that over time, each generation looks and performs as the previous one. It involves growing multiple generations of plants and running trials to understand what climate, growing season length, and other factors they require. “It has not been uncommon for us to get calls from a seed vendor saying, ‘I have hemp seed, it’s this variety, it’s ready for sale, I want you to certify,’” says Chet Boruff, AOSCA’s CEO. “We can’t do that, because we don’t know where it came from, we really don’t know what variety it was, nobody there was to watch over it and make sure it maintained varietal purity.”
“Plant breeders are working overtime right now to develop consistent, low-THC varieties that will work in the marketplace,” Boruff says, but in the meantime, “there are a lot of people who are trying to sell seed and take advantage of this excited marketplace and farmers end up buying seed that is low quality and, in the worst case scenario, has a [illegally] high THC level.” Commercially available seeds tested by Smart regularly sprout at well below the rate promised by their sellers. One hemp variety from Oregon only germinated at a 39% rate, according to a presentation Smart gave at Cornell that was later posted online.
Meanwhile, since both CBD and THC in a plant increase over time as it grows, most farmers who buy even otherwise high-quality seed are still caught playing a game of chicken: They want to leave their plants in the field long enough to maximize CBD. But if they wait too long to harvest, they’ll end up with too much THC — and a crop that has to be destroyed.
Though it’s been known for almost a decade that THC and CBD production in the cannabis plant are controlled by more than one gene, it wasn’t known exactly how cannabis plants inherited one trait or the other. But late last year, researchers published a genome map that showed where the two genes resided on the chromosomes of a marijuana plant, a hemp plant, and a cross of the two plants. The marijuana plant had a working version of the gene key to THC production, while the hemp plant only had a working version of the CBD gene (that was a little leaky for THC). It still isn’t fully understood why certain plants produce so much CBD or THC.
“In one day we can test a plant and tell you if it’s going to produce THC or CBD. We can use real science, which means it’s reproducible. It’s not just something you heard from your buddy down the street.”
George Weiblen, the University of Minnesota professor, has submitted research for peer review that further decodes the cannabis genome. The paper builds on work he did in collaboration with Sunrise Genetics, a company that applies genetic tools to cannabis and which has invested $500,000 to create a more detailed map of the cannabis genome. “Before we didn’t know where any of these genes were relative to each other,” Sunrise Genetics CEO CJ Schwartz says. “So if we wanted to connect two of them, it was essentially like trying to go from Kansas City to San Diego without a map. Now we know exactly where San Diego and Kansas City are located relative to each other, and the quickest way to get between them.”
A clearer picture of how cannabis’s genes impact its traits will allow breeders to better select plants for their programs based on their DNA while they’re still seedlings, without waiting for them to grow. That makes the breeding process more precise and efficient — better products, and fewer farmers forced to destroy their crops. “In one day we can test a plant and tell you if it’s going to produce THC or CBD,” Schwartz says. “We can use real science, which means it’s reproducible. It’s not just something you heard from your buddy down the street.”
Understanding the cannabis genome also opens cannabis up to genetic engineering through tools like CRISPR, which some scientists think could be used to remove THC from the CBD-producing plant altogether. Though Sunrise Genetics has not yet experimented with gene editing, the company is in the process of testing five high-CBD cannabis varieties it has developed through traditional breeding. The varieties will be grown at trial plots located in 12 different states, so the company can learn and document how they are affected by different environments, and which traits work best in different environments. Disease resistance, drought resistance, time to flower, and other traits are all on the table. “Anything that anyone has looked at in a modern crop is a potential target for us,” says Schwartz.
The 0.3% limit for THC in cannabis plants is by all accounts arbitrary. It has no scientific basis. If the limit were higher, farmers who have no intention of growing drugs would have more room for error, without any additional risk of psychoactive cannabis flooding the market. But the government is unlikely to change the limit. Eric Steenstra, the president of Vote Hemp, an industry advocacy group, told me before the drafted USDA rules were released that he hoped the regulation would be more lenient, either raising the allowable level of THC or testing material from the whole plant instead of just the top flower, where the highest level of THC is found. Neither of these changes made it into the draft released last month.
But as scientists develop a better understanding of hemp, science could help farmers sidestep politics. Breeders will develop varieties with specific traits, certification process will help farmers predict results, and eventually the results from low-THC certified varieties will be so well-known that states won’t even need to test plants themselves. “I think that’s possible,” Steenstra says. “But it will take some time. “