10:48
When I think of a greenhouse, I think of plants quietly growing in the soil. But this one looks more like a factory. Plants are lining up to be scanned, 24/7, from all possible angles, and turned into data. The goal is to choose, or even consciously design, crops that are fit for the future. "I would call it a gym for plants." Because somehow, we need to feed 8 billion people in a hotter world, with less land, less water, and unpredictable weather. So, can this greenhouse help us figure out how?
We're in Wageningen in the Netherlands. It's all fields as far as the eye can see. The Netherlands is the world's second largest exporter for agricultural products - after the US, despite being about 230 times smaller. So they're doing something right. The country has been dubbed the 'Silicon Valley of agriculture' and is covered in greenhouses. We're on our way to one. It sits on the campus of Wageningen University. Everything is green here. So sour. But tasty. It's not built to harvest tons of tomatoes, but for research: to find out which plants perform best under stress. And we're here. That's the greenhouse over there.
It looks a bit like a space station growing food. Okay, it's very bright in here and not as humid as I expected. Some plants grow in open light. Others. They're locked up. Here, they can mimic almost any environmental condition on Earth and see how plants respond. "We can regulate the temperature. So we can give them a heat stress phase. We can freeze shortly, so we can simulate the night frost, which could be very disruptive for the young plants. We can introduce salinity by adding salt water in the thing."
It's like a time machine for global warming. The usual playbook of agriculture relies on predictable seasons - plant in spring, harvest in fall, for example. But now. "A lack of rain in recent years." "Repeated heatwaves." "A warming planet plus more extreme and unusual weather." All of that affects what farmers harvest. This is the curve of agricultural productivity in recent years. It's been plateauing lately. Farming is hitting limits: exhausted soils, less water, and less cultivable land. But the pressure is on: To feed 10 billion people by 2050,
we'll need much more food - 50% more, in fact. People are eating more meat, which takes more grain to produce; and demand for biofuels from maize is also growing. So, we need to know which crops will be productive and resilient. That's where these high-tech tools come in. Each of these different tree species is watered automatically, and sits on a highly sensitive scale that records its weight every three minutes - continuously, for weeks. That tells researchers which one uses the least water, and which turns it into biomass most efficiently. All of this is called phenotyping, the science of measuring plants and their features.
This is not new. We have observed and selected the best plants for the past ten thousand years. Back in the 19th century, botanists used this. A thin wire is tied to the plant. It moves a pen which draws a graph of the plant's vertical growth - but one plant at a time. They also took plants apart to measure them by hand - and still do it today. But it's a slow process. "All the lab measurements were restricting the size of the experiments. If you look around here in NPEC, we are talking about
thousands of plants in one experiment." That allows them to not just identify one good crop, but a whole range of plants with different strengths. "This is a new way of working, more data driven, more AI based, more, basically more complex data sets." I need to see all of this in practice. So, I'm going on a tour. First stop: Sicily - or its greenhouse simulation. Here, I'm meeting Gresheen, a researcher from Italy. She's investigating what's known as 'brassica' plants - that's broccoli and cauliflower.
"Before I started my PhD, I [didn't] usually eat broccoli, but now I really love [it]." Broccoli is a nutritional powerhouse packed with fiber, vitamins and antioxidants. Gresheen is comparing three brassica varieties: those sold in the supermarket, landraces - meaning traditional varieties, and their wild cousins. "A lot of these plants are originated in [the] Mediterranean environment. They are really resilient, I think, and adapted to the extreme conditions like high temperature and drought stress." That's currently being put to the test - they're only getting 30%
of their usual water. These guys don't really look like they're going through a dry spell. They're fine! Except maybe for this one. "In normal conditions, the leaves would be upwards and now it's more floppy and probably starts to get yellow, because it's a little stressed so the color changes." That's what the human eye can see. But these scanners record much more - twice a day. This machine generates 3D models of the broccoli by profiling it with 15 cameras. Within seconds, it identifies its shape, size and structure.
This individual is called F00308, and grew like that over the course of a month. Other scans can monitor how well the plant photosynthesizes. Special sensors detect tiny changes in light emitted by chlorophyll, which reveals how efficiently the plant is turning sunlight into energy. The pots rotate fully, so every angle is recorded. Apparently, the sensors are also hungry because this one tried to eat our phone.
Next door, we're in the greenhouse's version of Turkey. Here, Thijs Stegmann and his colleagues investigate a key defense mechanism of any plant: "A good root system can help you survive dry spells." All of this is wheat and barley. "It's a very different way of dealing with the outside world compared to what animals do and what we do, we just run away from our problems. They have to stay and make little, little changes." The containers are very thin and slightly tilted, so the root grows along the glass and reveals what's usually hidden underground.
The roots can grow larger than the shoot. "This root already does a little bit of a hide and seek in the middle where we're missing a part." The system automatically takes pictures of each container up to five times a day to record growth. And finally, we're entering tropical India. Today's forecast: 35°C and humid. The room is packed with cameras. The object of their desire: baby tomato saplings. They are just being planted into these starter cubes.
During India's summer and monsoon season, plants face either hot and dry, or hot and humid weather. Both can disrupt the tomato supply chain. So, Alan Pauls, who facilitates research while finishing his PhD, is showing me what they're measuring to stop that. "You have four sets of cameras. So, you have a fluorcam that can measure photosynthesis. What you see really is that just the plant which has chlorophyll is highlighted while everything else is not. So if you see these red dots, that is the highest." Photosynthesis indicates that the plant is healthy. "You have an RGB cam that can capture the colors." It allows them to track if the leaves move up and down in the right rhythm. If they do, it means the plant can handle heat and humidity well.
And tiny color changes can indicate that it's picked up a disease. "When that happens, you need to have an accurate way of representing it, other than just saying that, hey, this is green-ish, you're going to get perfect pixel values. A thermal camera shows the temperature. Like us, plants have an ideal temperature at which their enzymes function best. So if it can cool down by sweating, it can handle heat well. "Four cameras, so it's like a paparazzi room." "Ah, yeah." The key advantage: they can measure the same plant again and again.
"If I want to know how much sugar is there in the plant, usually what people do is they have to take a leaf or a sample." But picking off leaves would reduce its productivity. Yield, or how much it produces per season, is a result of everything the plant experiences over its life. So, in order to draw conclusions about which of these is most productive, they need to stay fully intact. "What, say for example, a hyperspectral camera could do is that probably there is a particular wavelength in which sugar molecules, certain kinds of sugars, what you're interested in have a different reflectance. That can get captured
and then it's a nice proxy. So I'm not destroying the plant, but I can know how much sugar is there." But there's an obvious question. "How are you making sure that the plants 'pampered' in this highly controlled environment will behave the same way outside in a dynamic, messy field?" "That's the million-dollar question because there is no guarantee." And they can't recreate everything here: hail storms, high winds, or differing soil temperature, for example. That can create blind spots because plants react to everything.
That's why the researchers actually repeat many greenhouse experiments outside, some across the whole globe, to see if they get similar results. For that, we're moving into a fresh field of barley close by. Here, they can control way fewer parameters. These plants are one and a half months old. Tastes like grass. And a little nutty, I would say. But I'm not sure how barley usually tastes like so. Bas Jonkers is measuring them.
He runs this device over the field once or twice a week. Didn't think this was going to be so loud! The whole process is automated - of course - so the device basically runs itself. "So when you drive over these plot files, then you do the measurement for each separate plot." "And the computer knows via GPS tracking where you are?" "Yeah. It has a one-centimeter accuracy." Its imaging cabinet lowers over the plants and scans them with cameras similar to those in the greenhouse.
It can capture biomass, plant density and photosynthesis. For this field, with the TraitSeeker, you gather around 800 or 900 GB of data. Eventually, the scans can help improve the barley. "It's all about fitness. We simply score: how do these plants respond to the stress? Which ones can cope with it, which ones suffer from it. And then by looking at the response, we can also try to figure out how the DNA profiles can be linked to the tolerance against these stressors." Here, they are using top-performing varieties from breeders to compare which ones grow best.
Bas just told me there's more than 200 types of barley growing here, and they're all at different heights because they're different varieties. In other experiments, researchers go a step further: They cross two varieties, one with a high yield and one that is robust, for example, and produce a lot of offspring. Each one inherits a unique, scrambled mix of DNA. When the researchers see which offspring performed best overall, they look at their DNA, to find the genetic clues they share. The goal: fast forward evolution itself. "We enable researchers to screen large quantities of plants, but the actual breeding and the actual decision making is not up to us."
Much of the research is funded by the Dutch government and plant breeding companies. They can then use this data to eventually launch more robust varieties, faster and more targeted. I need some food to process this now. So I'm meeting Alan for a veggie burger with mushroom patty. I'm looking at this with totally different eyes now. Full disclosure - the taste took some getting used to. "The patty is interesting." "I mean that's also something that you can breed for."
"The patty?" "Yeah. I mean, so not the patty, but then the texture of the mushrooms that go into the patty." Turns out, my whole burger is optimized. "You actually have a variety of lettuce called burger leaf lettuce, where every leaf as you open it up is the size of a burger. So you don't need to cut it, you don't need to shape it." That's not random - it has been bred that way, along with its texture and thickness. This breeder advertises lettuce leaves that "fit perfectly in a bun, aided by its crisp bite and excellent flavor".
All the scans from earlier can technically feed into that process - but most of the research is intended to improve sustainable agriculture, not build the ideal burger. But Alan also experiences skepticism about his work. "I mean, especially from my family, actually. When you talk to the older generation, I definitely do feel like they don't trust it. The moment you talk about taking food into the lab or plants into the lab, immediately what you think about is GMO. 'Oh my God, you're working with fake food.
You're putting things that are not natural together.' " Genetically modified organisms, or GMOs, have had their genes changed artificially. For example, by inserting a gene from a different species, like bacteria, into a plant to create new traits. In Wageningen, they work with GMOs too, but only in 5% of the experiments. It's strictly regulated. With phenotyping, they can avoid genetic modification, and speed up natural processes - just with more precision. Still, many people seem to believe too much machine interference makes food artificial. Like these people who commented under a video about this exact lab.
"The point is, what do we think is artificial, right? Are we thinking that this is artificial because a machine is involved, or are we thinking it is artificial because it's happening so fast, or it's happening in a lab? I can still harvest rice with the sickle. But is that gonna feed a billion people in time? No." Most of our food already comes from large-scale farming systems that rely on machinery and automation. In fact, some of what's grown here might even end up at an organic market. "We grow organic lettuce here and stress them, especially the organic domain needs to rely more heavily on more robust plants because they are not allowed to spray with all kinds of chemicals
against like diseases and insects." Looking at our food system though, improving crops is only one piece of the pie. Research shows the world already produces enough food, even for future generations, but it's not accessible to all. Recent data suggests that up to 40% of all food is lost or wasted along the value chain and that's one of the main causes of hunger. And adapting crops to a heating world does not tackle the root problem. We still need to protect soils and cut emissions for these to have a future. Would you be happy eating these high-tech veggies like this one?
Tell me in the comments, don't forget to like and subscribe if you want more videos like this, and check out dw.com for more stories.
