Unearthing Nature’s Pesticides
The four boxes sit on a wooden pallet in the middle of four acres of organic strawberries, on a farm just north of Toronto. They’re each roughly the size of a shoebox, with air vents perforating their lids. And they’re buzzing—a light drone that tickles the eardrums.
I’d planned to see these boxes for months, selecting a week in late spring when the Canadian strawberry fields would be in bloom and the weather, I hoped, would be reasonably dry and warm. Instead, bruised clouds fill the sky and a frigid wind cuts through my borrowed jacket.
The bumblebees inside the boxes don’t seem to like it, either. My host from Bee Vectoring Technology, a Toronto startup, tells me the insects prefer calmer days and warmer temperatures. In better weather, I might have seen the pollinators buzz out of the nickel-size holes at the ends of the boxes at a regular clip, dipping from flower to flower in the surrounding field, each carrying an unusual delivery: a white dust formulated to protect the strawberries from a type of rot known as Botrytis cinerea, or gray mold. The dust contains a benign fungus, Clonostachys rosea. It colonizes the inside of plants, blocking the growth of the nastier mold—a biologically based alternative to a cocktail of synthetic fungicides, which are getting more difficult to use.
Todd Mason, BVT’s lead scientist, strides into the strawberry field, ruddy-faced and in short-sleeves despite the weather. He raps on a hive. The buzzing crescendos, but no bees come out to investigate the source of the disruption.
Mason shrugs and then surveys the field, rubbing his hands together. “I’m going to take some samples,” he says, grabbing a handful of Ziploc bags. His goal: to gather strawberry blooms so he can measure how much of the white dust the bees left on more pleasant days. This field is one of several demo trials in North America and abroad. BVT is already convinced—based, in part, on decades of research from scientists at the University of Guelph in Ontario—that the white dust can fend off the gray mold that afflicts strawberries and numerous other crops. The purpose of the trials is to prove to farmers that this unconventional pesticide, with its unconventional delivery method, works in real fields, where the weather—and the bees—don’t always cooperate.
BVT is among a swell of companies, from startups to the world’s agrochemical giants, trying to bring pesticides derived from natural materials, or biopesticides, to the agricultural mainstream. Biopesticides are growing faster than synthetic pesticides, which make up the bulk of crop protection. Newer materials are making their way through the regulatory pipeline. There are several forces driving this trend. For example, pests and pathogens have grown resistant to many pesticides, while the EPA has phased out older chemistries, due to environmental and health concerns. These concerns have not only increased government regulation and driven up the cost of developing new chemical pesticides, but have also increased demand from consumers for farmers to grow more organic produce.
“Twenty years ago, if there was a problem and there was a pest, you’d kill it,” says Michael Collinson, BVT’s chairman and former CEO. “When a new chemical came out and it worked, you’d pour it on. Today we have to change our thinking if we’re going to be sustainable.”
On the surface, leveraging nature’s vast biodiversity to protect crops against some of agriculture’s thorniest pests looks elegant and appealingly Earth-friendly. The possibilities are nearly endless: Untold millions of species of microbes, insects, and other overlooked organisms live in our soils, fields, and streams, waiting to be tapped. These critters already infect and eat one another and manufacture chemicals that they use to protect themselves. Why not enlist them to duke it out on our farms? We’ve already done so successfully for decades with products derived from Bacillus bacteria, which can be found in practically any farm or garden supply store. After all that time, Bacillus is still the primary ingredient in about three-quarters of all biopesticides.
But finding the right organism to counter a specific pest or blight is no easy task. And biopesticides are often specific to one type of pest or disease. That’s a plus from an environmental perspective, as it means they won’t wipe out beneficial organisms along with pathogens, but it means most biopesticides have to be used in concert with a host of other tools—a more difficult proposition than the easy shoot-and-spray of conventional pesticides.
Convincing farmers to adopt a radically different approach to crop protection is also tricky. Many farmers, and some pesticide experts, believe biopesticides are subpar compared to synthetic chemicals, partly because of the products’ lower toxicity—they just don’t kill as effectively—and partly because they often require that we harness living things to serve as delivery mechanisms, which can be difficult to control in the field. Skeptics who prefer the ease of conventional sprays that knock pests dead dismiss biopesticides as “bugs in jugs.”
The biopesticide industry’s future depends on finding new materials that work—and convincing more farmers to try them, quirks and all. A few months before my Toronto visit, I asked Sara Olson, a research analyst at Boston-based Lux Research who specializes in emerging agriculture technology, about the outlook for biopesticides. She was cautiously optimistic about the industry as a whole and called BVT’s technology “really, really promising.” She then added the caveat that it was still “way too soon to tell” and worried about “giving the impression that there should be unbridled enthusiasm for all biopesticides at all times. The reality is some of them just don’t work as well as synthetics.”
On the strawberry farm, I huddle closer to a layer of fragrant straw strewn between the rows of strawberry plants and watch the hives, waiting for a bee to appear.
Biopesticides may be trendy, but the concept of harnessing natural materials to protect crops isn’t new. Toxic plant extracts such as nicotine and pyrethrum, made from crushed chrysanthemum petals, have been used as insecticides for centuries.
The most commercially successful modern biopesticide is arguably Bacillus thuringiensis (Bt), a soil bacterium discovered by a Japanese bacteriologist in 1901. The French started selling Bt in the 1930s; Americans followed in the 1950s. Different strains kill different insect species, such as moths and mosquitoes, but they all work the same way. The bacteria produce crystal-like proteins that, when ingested, bind to insects’ guts, poking holes in the lining and eventually killing them. (The proteins are harmless to most other animals, including people.) Bt remains popular in organic farming and scientists have also inserted genes that code for the bacterial toxins into some of the world’s most common genetically engineered crops, so the plants produce their own insecticide.
BVT’s approach to biopesticides dates back to the 1980s. John Sutton, then a plant pathologist at the University of Guelph, was looking for biological alternatives to fungicides to control the gray mold Botrytis, which hides on dead plant debris in practically every farm and greenhouse. Its spores carry on wind, rain, or irrigation systems, landing on damaged or dying plant matter, then grow into the tissues and rot them. Botrytis afflicts 240 plants worldwide. Sutton was particularly interested in finding a way to control its spread in strawberries, a high-value crop, without dousing the plants—and the people who handled them in the fields all day—with so many harsh chemicals.
In the years leading up to his project, Sutton says, concern over “residues of fungicides and other pesticides in foods and the environment” was on the rise. There were also worries about the health of the farm workers, he adds, because the directions for use “were pretty sloppy then, compared to now.”
To find the right biopesticide, Sutton and his team analyzed 1,400 species of microbes—including bacteria, fungi, and yeast—scraped from strawberry plants to find one that naturally blocks gray mold. In what Sutton calls a “microbial Olympics,” they winnowed the list by isolating the candidates, spraying them on strawberry leaves, flower petals, and stamens in petri dishes—then introducing gray mold spores.
Microbes that muscled out gray mold in lab dishes moved to the next rounds: tests in growth rooms, greenhouses, and eventually, for some, open fields. The fungus Alternaria alternata blocked gray mold, but its inconsistent performance and the fact that some strains are pathogens ruled it out. Various species of Penicillium beat Botrytis but sometimes rotted strawberries themselves. The fungus that “won the gold,” Sutton says, was Clonostachys rosea, which blocked gray mold in bout after bout. Better yet: the scientists found that the fungus was as effective as captan, a common fungicide.
Clonostachys is an endophyte—it infects a plant without harming it—and it snakes between the cells of the plant, occupying virtually every available cranny. Inside the plant, Sutton says, “occupation is nine-tenths of the law. Once the tissue is occupied by the fungus, other fungi cannot invade.”
Sutton now had a potent weapon against Botrytis—he just needed a way to reliably deliver it to strawberry plants. He tried spraying the spores on the strawberries, which worked. But sprays drift where they aren’t wanted into the surrounding environment—the material only needed to hit the flower. “We said, why do we have to spray the entire plant?” he recalls. Then, in 1988, Sutton heard about the work of Peter Kevan, an entomologist and botanist who was also at Guelph. Kevan and colleagues were trying to suppress milkweed by infecting it with yeast that would interfere with seed production. The researchers hoped to use honeybees to carry yeast to the weed. But based on the small studies they were able to do, the yeast didn’t appear to work. They ran out of funding and abandoned the project.
But Sutton was drawn to the approach, since pollinators like bees were going exactly where he wanted to deliver his new biopesticide: straight to the flowers. He asked Kevan to collaborate on a method of delivering Clonostachys to strawberries to fight off gray mold.
Sutton and Kevan started with honeybees, but Sutton went on to test bumblebees—the species prefer different flowering plants, and the latter are more likely to pollinate strawberries. The bee-delivery approach worked. Not only that, but the bees targeted crops more precisely than spraying the biopesticide across a field.
Over the next two decades, the Guelph team honed their process, finding the best way to raise the right strain of the fungus in the lab and testing it in the field again and again. (The details of their method, which involves growing the Clonostachys in large bags of grain, are proprietary.) But it wasn’t until after Sutton retired at the end of 2004 and attitudes toward biological controls began to soften that commercializing the technology seemed possible. They passed the intellectual property to BVT, both staying on the project as scientific advisors. The company raised money to bring the approach to market, and worked to reformulate the dust and dispenser to make it easier for the bees to deliver the product.
In 2012, BVT began pitching its biopesticide to farmers. The white dust comes in small foil-covered trays that slot into commercial hives. (Competitors in Europe sell similar products: BeeTreat in Finland and Flying Doctors in Belgium.) To exit her hive en route to the field, a bumblebee has to walk through the dust. It clings to her furry body, long evolved to pick up microscopic particles of pollen. When the bee alights on a bloom, she shimmies her wing muscles in rapid-fire contractions, knocking out pollen—a move, called “buzz pollination,” made by bumblebees, but not honeybees. The shimmy drops some of the white dust onto the bloom, too, inoculating it against Botrytis. By BVT’s calculations, each tray holds, at a minimum, more than 2 billion Clonostachys spores, and individual bees can carry around 300,000. To protect a plant from gray mold, it only takes a few.
At BVT’s headquarters, tucked into an office park in Mississauga, west of Toronto, I’m sitting at a small conference table with Collinson and Mason as they show me a bumblebee hive with a custom-made lid that fits one of their trays of Clonostachys biopesticide. Mason hands me a tray, its foil peeled back to reveal the white dust inside.
“Can I touch it?” I ask.
“You could eat it,” Collinson says.
That, I decline. But I run my fingers through the dust. It’s fine, like talcum powder, with an occasional piece of grit. This, Mason tells me, is a silica gel that helps keep the dust from clumping in damp weather.
As I play with the dust, Collinson and Mason begin the hard sell, ticking off reasons why BVT’s product is superior to synthetic pesticides. In addition to reducing fungicide use, they tell me, their approach could improve the overall quality of a crop. Farmers don’t always use commercial pollinators for strawberries; wind pollination is more common, at least in North America. But bees may do a better job spreading pollen to each of a flower’s pistils—the female reproductive organs, which must be fertilized with pollen from the male stamens.
If pollen only spreads to part of the bloom the resulting fruit is crinkled—farmers call it “cat-faced.” The market judges on appearance: Perfect strawberries sell fresh at a premium price, while disfigured fruit ends up in cheaper processed food. The argument has merit: a 2013 study by German scientists in the Proceedings of the Royal Society B found that bees improved strawberry quality, quantity, and sales price compared to wind pollination.
But the most compelling argument for the biopesticide is its edge with pesticide resistance. With synthetic pesticides, resistance is inevitable if they aren’t managed well. Although a chemical may initially kill most of its targets, some individual germs and bugs are naturally resistant to the poison and survive the treatment to pass their hardiness to the next generation. Using the same pesticide again and again makes resistance more likely. On strawberry farms in the U.S., farmers spray fungicide from the time the plants bloom until they are harvested, often every week. Pesticide resistance is spreading fast, with some farms down to just a few products that still work.
Resistance is especially likely for most of the fungicides used to treat gray mold, which target very specific characteristics in the fungus. Take benzimidazoles, a class of pesticide that’s been on the market since the 1960s. Benzimidazoles disable a key protein necessary for Botrytis cells to divide. To get around it, gray mold needs only a slight evolutionary nudge—a change in just one letter in its genetic code—to continue making this essential protein even when doused with the fungicide.
In contrast, Clonostachys doesn’t kill gray mold by knocking out a key genetic building block—it jumps into a plant and closes the door behind it. Gray mold would have to make a more dramatic transformation, with multiple genetic mutations falling in just the right way, to sneak in that door first, or knock it down by sheer force.
BVT’s biopesticide won’t replace fungicides on strawberry fields entirely. For one thing, the bees only deliver the biopesticide to the flower, which won’t protect the strawberries against pathogenic fungi that enter through the soil. Still, Collinson hopes the product will decrease dependence on synthetic fungicides. One way it might do so, he suggests, is by allowing farmers to use the chemical less frequently—thus slowing the development of resistance.
The company plans to extend the same principle to other crops. Collinson smooths a spreadsheet across the conference table and points at columns of numbers. “There are actually 87 crops that require pollination,” he says. BVT is focusing on the 20 crops with the most acreage, which include apple, canola seed, squash, strawberry, sunflower, tomato, watermelon, and zucchini.
Alone, none of these are major crops on the scale of commodities such as corn or wheat. But together, across North America and Mexico, these so-called minor crops add up. And “if you start getting into Europe—Germany and France and Turkey—you can see there are massive acreages,” Collinson says. “Twenty crops in twenty countries, you end up with a huge, huge potential.”
The company hopes to eventually expand even more, by offering its technology for honeybee hives, which are more common in the commercial pollinating business and are specific to a different range of crops. An individual honeybee won’t likely carry as much dust as a bumblebee: they’re smaller and smoother, and they don’t shimmy when they pollinate. But honeybee hives are far bigger—the bumblebee hive sitting on the conference table maxes out at 300, while a commercial honeybee operation may have 30,000. BVT is building new trays to accommodate the honeybees’ smaller size and larger hives.
The company also plans to experiment with adding other biopesticides to the trays. That way, bumblebees and honeybees can carry multiple materials, helping protect plants from several pests and diseases at once. In addition to the fungal spores for gray mold, for example, a tray might hold Bt bacteria, to kill certain pest insects and leave the bees unharmed. This combo dust, Mason says, is akin to “throwing a bunch of stuff in a FedEx box.”
When BVT’s technology was in its infancy, “few people gave much credibility to biological control,” says Sutton—there were “lots of naysayers.” Since the advent of DDT in 1939, human-made chemicals had been the workhorses for controlling pest insects, diseases, and weeds. While chemists who make synthetic pesticides often draw inspiration from nature, they engineer the chemicals—plucking or adding atoms to make a form that is more potent and longer lasting. For example, pyrethroids, a common modern class of insecticide, are synthetic takes on pyrethrum—the natural pesticide made from crushed chrysanthemum. But pyrethreum breaks down in sunlight, while pyrethroids are chemically tweaked to withstand it.
Synthetics used to be relatively cheap to produce. But when the EPA formed in the 1970s, health and environmental regulations gradually grew more robust, and the price tag for pesticide development rose.
To find new pesticides, scientists comb through massive chemical libraries, testing substances known to have characteristics that may help them kill or control a pest or disease. If a chemical works in preliminary tests, it goes through yet more studies to determine its toxicity to humans or the environment and to show whether it works in the field. According to the most recent estimates from CropLife, a trade association, major chemical companies search an average of 159,574 different compounds to find just one they can sell. The discovery of that single chemical takes around 11 years and costs an average of $286 million. Forty years ago, it cost $23.1 million.
Biopesticides are cheap by comparison, particularly in the U.S. Twenty years ago, the fledgling biopesticide industry lobbied the EPA to make it easier to bring their products to market, arguing that because these biologically based products are fundamentally different from synthetic pesticides—and are, in general, gentler on the environment—they should have a different regulatory path. The EPA agreed. “We were the first country in the world to create a separate unit for the licensing of biopesticides,” says Jim Jones, assistant administrator for the agency’s Office of Chemical Safety and Pollution Prevention. This way, he adds, biopesticides “don’t get crowded out by the synthetics chemistries.”
Biopesticides still must meet safety requirements, and those that are especially toxic are regulated like synthetics. Still, on average each active ingredient in a biopesticide costs less than $10 million to develop and takes around four years to test—just one to two percent of the cost to develop a synthetic pesticide, in around a third of the time.
Many of the companies in the biopesticide market are scrappy startups, but big agriculture is elbowing in, too. In 2012, Bayer CropScience bought a biopesticide company called AgraQuest for a reported $425 million. Over the past several years, Monsanto, Syngenta, DuPont, and others have also invested heavily in biopesticides. Even Bee Vectoring Technology has ties to Big Ag: A former Bayer executive recently accepted the CEO position, and a former Syngenta scientist serves as both a scientific advisor and board member.
Some of these companies, like BVT, seek fungi that can block disease. Others are using bacteria, yeasts, and viruses that are lethal to insects. There are natural growth hormones, too, such as a neem oil extract called azadirachtin, which stops insect pests from maturing and reproducing. Then there are pheromones—chemicals that insects use to communicate. Some pheromones attract insects, which could be useful in traps, while others act as an alarm system to warn insects away from danger, which might repel a pest from a field.
All this growth means more options for farmers. “A lot more successful biopesticide products have come on the market,” says Michael Braverman, manager of the biopesticide and organic support program at Rutgers University’s IR-4 Project, which helps register pest control products for minor crops such as artichokes and strawberries. “And the increase in the quality of these products has given them better standing than they’ve had in the past, making them more attractive.”
For example, chemicals made from fermenting the microbe Chromobacterium subtsugae, Braverman adds, work well as an insecticide, while the fungus Aureobasidium pullulans is effective at treating orchards for fire blight in the Pacific Northwest.
Alan Schreiber, an agricultural consultant in Washington State, agrees. He says growers and chemical companies often hire him to test both conventional and bio-based products, including experimental materials that are five or more years from market.
Some of his clients are skeptical of anything green or hippie-dippie, he says. “These guys are right of right of center. They don’t want anything organic, they don’t want any biopesticides.” Their attitude is “give me a goddamn hardcore pesticide.” Yet, left with few alternatives, Schreiber says, “even people who have no interest in biopesticides are getting into situations where they feel they need to evaluate them, because they are just running out of alternatives.”
Despite the promise of biopesticides, the naysayers may have a point. Synthetic pesticides are fundamentally different from biopesticides not only in how they’re made, but in how they are used in the field. Synthetic chemicals are predictable. Point and spray, and the offending insects die, or the disease stops withering a crop’s leaves, or the weeds recede. Biopesticides, however, require exploiting the biology or behavior of an organism, and that’s tricky.
On one hand, you can count on a bumblebee’s natural instinct to gather pollen or on the fact that a fungus will sprout a certain way inside a plant it infects. On the other hand, a fungus can only infect a plant once it’s been deposited there—and bees prefer to fly when the weather is nice.
Bees may also wander where they aren’t wanted. According to a 2014 study by Dave Goulson, a biologist and bumblebee expert at the University of Sussex, 73 percent of the pollen collected from commercial bumblebees on three Scottish farms came from wildflowers, rather than the intended fruit. Bees will eventually hit the right plants—if they didn’t the commercial bee industry wouldn’t exist. But a bee coated in a biopesticide that’s attracted to a wildflower may carry the white dust to weeds or other plants that don’t warrant a farmer’s protection. That would not only be a waste of money—it might also allow the material to spread to the surrounding environment, including onto weeds, strengthening their resistance to disease.
Sprinkling weeds and wildflowers with the dust “might have some ecological impact, but I’d guess not too much, Goulson says. “A bigger concern would be if the fungus is used in regions where it does not naturally occur, as it could easily spread into the wild with unknown consequences.”
Mason of BVT points out that, although the biopesticide will inevitably end up on non-target plants, the fungus will die unless it germinates inside a plant within six to eight hours. According to the company’s internal research, the bees can’t carry the dust beyond about 1,150 feet. And because the biopesticide is a physical block, rather than a chemical killer, the chance of weeds getting hardier and more resistant to the white dust isn’t likely. “The chances of this creating super weeds is almost impossible,” Mason says.
It’s not just the bees that can be fickle. BVT’s flagship fungus degrades at high temperatures, so farmers need to keep the trays refrigerated up until they pop them into the hives in the field, swapping out fresh ones every week.
Similar challenges extend to practically every biopesticide. For example, some products are made with nematodes, small organisms that live in the soil and eat up other tiny creatures. Nematodes need to be kept cool, and if they aren’t deployed in damp soil or leaves, they’ll die. Chemicals extracted from plants or microbes may break down in sunlight more rapidly than synthetics, and require more applications, which translates to higher labor costs. Other biopesticides, such as the insecticidal fungus Beauveria bassiana, only germinate in high humidity. These are virtually useless in drier croplands, says David Haviland, a farm advisor and researcher with the University of California Division of Agriculture and Natural Resources.
And no matter what a pesticide is made from, there is always the looming threat of resistance. Evolution doesn’t care about a pesticide’s origin. While some biopesticides, such the Clonostachys in BVT’s white dust, aren’t as likely to have this problem, products made with specific toxins that kill in a precise, straightforward way are vulnerable. Although it took decades of active use, and often overuse, the first signs of Bt resistance emerged in Hawaii in the early 1990s in diamondback moths, a pest common to brassicas such as broccoli and cabbage. Resistance has since spread to other insect species, helped along by the popularity of GMO crops that incorporate Bt toxins.
Because biopesticides usually work in radically different ways than point-and-shoot chemicals, it’s hard to directly compare their effectiveness. Haviland says he’s tested numerous biopesticides and isn’t impressed. “Usually we test them a year or two and they prove to be worthless, and we move on,” he says. “That is the reality in most cases. There are some definite exceptions, but generally speaking, they don’t work that well.”
Other researchers see the tests used on biopesticides as the problem. A typical protocol for synthetic insecticides, for example, involves spraying and then counting how many insects died, and how quickly. A biopesticide that works in a different way—say, as a repellent, or by forcing the insects to stop laying eggs—requires a different, tailored test. “One of the challenges, frankly, as a scientist is trying to design experiments for these materials,” says Hannah Burrack, an entomologist at North Carolina State University. Using a standard protocol for a biopesticide, she adds, is often “not really a fair test of that product’s potential, because they aren’t intended to work in the same way.”
If the scientists have trouble working with biopesticides, the same is also true of farmers. Using any pesticide correctly takes training. Biopesticides can be even “less forgiving,” says Haviland. And farmers who are interested in biopesticides don’t always have access to education to help coax the products to work.
“The microbes—that new frontier is fantastic, but it needs more applied research,” says Amy Hepworth, an organic farmer in New York. If a farmer tries a biopesticide but doesn’t get it right, and then their field gets infested, they may have to use the harsher materials they were trying to avoid to begin with in order to save their crops.
Without research and educational support, Hepworth says, “you can’t just keep burdening the farmer with the stricter regulations, the heavier fines, keep telling us more, more, more.”
Back at the Canadian strawberry field, the wind is still whipping the white blooms into frenetic sways. A few more bees have appeared, but they’re not carrying any white dust. Some have snuck through the entrance of one of the hives, a maneuver that’s not supposed to be possible. On the inside of the entrance, there’s a small plastic flap that hangs like a little curtain. A bee entering the hive from the outside can push the curtain out of the way; from inside the hive, the curtain forms a seal. But gusts of wind are pushing the curtain open, allowing bees to slip through the gap. I see other bees, too, but they seem to be coming from a rogue hive that my hosts just discovered. One of the bees escaped her hive and started a new operation in the wooden pallets below.
But one disappointing day might not matter, says Ian Collinson, a project manager at BVT and Michael Collinson’s son. Since the bees are left out in the field for weeks, all they need are a few breaks in the weather to eventually pollinate the field.
The morning wears on, and Mason, BVT’s head scientist, comes back from collecting his strawberry blooms, to help build the case that—notwithstanding this cold, windy day—the bees are indeed delivering the white dust.
The sun breaks through the clouds in bursts, warming my back. Finally, a bumblebee, fat and cute, pushes through the exit hole in one of the boxes. She’s coated in white, as though dusted in ghostly powdered makeup. She pauses, preening an antenna with her foot.
The bee’s wings buzz, lifting her round body into the air. Then she’s gone, in search of pollen.
Interactive by James Davidson