Farmers of the future may be able to use plants as sensors that encode information about their environmental conditions in their own DNA. Then, they could use that information to order seeds custom-tailored to the conditions on their farms.
If it sounds like science fiction, that's because it is, a little. Yet it's grounded in existing technologies. This design provocation is the product of a collaboration between design firm IDEO and the Lim Lab at the University of Southern California, whom we last saw growing consumer goods with bacteria.
Agriculture has a problem. Its one-size-fits-all model requires outsize amounts of pesticide, fertilizer, and water to create a homogenous environment for a homogenized product. Monocropping means covering as much area as possible with a single, often genetically identical crop, and commercial genetic modification techniques only exacerbate the problem by self-destructing after a single generation, preventing seed-saving and the development of new varieties by farmers themselves.
Nature long ago solve this problem through the process of mass customization we know as natural selection. Now the challenge for agricultural scientists is to reintroduce biodiversity in the one place it's been most thoroughly eradicated--the industrialized farm.
"This is challenging our current understanding of genetically modified foods and monoculture," says Will Carey, a designer at IDEO who worked on the project. Both are usually associated with the race to cover the planet with plants that are genetically indistinguishable, which is great for spreading desirable traits fast, but requires constant updating as pests adapt to the new strains of staple crops like wheat, corn, soybeans, and rice.
The first step to transforming agriculture from mass production into mass customization--a trend that long ago took hold among consumer goods--is to measure the conditions in which a farmer hopes to grow crops. This could be accomplished with conventional information technology, of course, but there could be a cheaper way: Use the sensors plants already have built into them to record everything from soil acidity and mineral content to average rainfall.
"Plants sense their environment and exhibit sophisticated responses--the idea is to engineer that," says Reid Williams, a PhD candidate at UCSF who also worked on the project.
We already know that plants can sense gravity, touch, and probably dozens of other environmental factors. Regardless of their innate responses, all living things also encode information about their environment through a process called DNA methylation. When DNA is methylated, it changes its expression during the life of an organism, and there are ways to determine precisely which genes have been altered in this way.
Add some "big data" to this equation--massive, automated studies that statistically match DNA changes to the conditions in which a plant is grown--and you could begin to build a system that can examine the genetic code of a particular plant and spit out a record of the conditions in which it developed. Something along these lines has already been developed, namely a plant that turns red in the presence of land mines.
The next step would be creating plants that are custom-tailored to the farms, even individuals micro-climates, in which the sensor plants grew. Genetic engineering is significantly easier in plants than in animals, and it's been practiced commercially for decades.
Even conventional plant-breeding programs are beginning to use whole-genome sequencing to improve the plant varieties that scientists can produce. One way having a complete genetic sequence of all your seed stock is helpful is that it allows researchers to figure out which genes are associated with desirable traits, so that they can make sure to include those genes in any given engineered crop.
Another way genetic sequencing accelerates the development of new strains of crops is that under normal circumstances, you have to wait until a crop reaches a certain age to determine whether or not it has the traits you want. Genetic sequencing short-circuits that process, allowing you to know whether individual seeds have the traits you want, before they ever grow up.
The ultimate result would be "personalized seeds" (think of it in terms of personalized medicine) that are tailored to the environment in which they'll grow. This application of synthetic biology turns the usual sequence of events in agriculture on its head. Usually, knowing the preferences of his or her crop, a farmer would try to add things to the soil to adjust it to the plant's preferences. This scheme instead adjusts the plants to what's available, potentially making marginal land more usable and allowing a level of customization, even within a single plot of land, that's currently beyond the reach of even developing world agriculture.
Mass-customization of seeds and crops feels like the end-point of biomimetic thinking, because rather than just making something that is like life, it uses life itself as its starting materials. Hence the name of this field, synthetic biology, which implies that humans are at the stage that we are in a position to directly influence the substance of evolution itself, and not just in the accidental ways we have in the past.