Transforming Agriculture, Perennially

Issue No. 113
Fall 2015
View full issue PDF


Are there other crop candidates?

Only a handful of perennials enjoy attention, but they number in the thousands

BY Scott Bontz

British immigrant Henry Shaw built his fortune in frontier St. Louis, starting with hardware and cutlery, expanding to agricultural commodities, mines, furs, and real estate. He retired in 1840, age 39. With about 1,000 acres to his name, for his adopted home at the confluence of the Missouri and the Mississippi he wanted to build a public garden. Shaw’s friend George Engelmann was an immigrant from Germany, and not a businessman, but a physician. He also was one of the young country’s leading botanists. Engelmann wanted to see a botanical garden like those in Europe, with a library and a herbarium for the systematic collection of plant specimens. He and other scientists persuaded Shaw to make his garden more than a park.

Today the 156-year-old Missouri Botanical Garden serves visitors with myriad plants, three cafes, and half a dozen annual celebrations, including cultural festivals, flower shows, and concert series. It also claims more than 6.5 million scientific specimens, one of the world’s best botanical libraries, and the world’s largest online database of plant information.

A man named Peter H. Raven led the garden for four decades, bridging into this century, and greatly expanded its research activities. He championed efforts to preserve endangered plants, and co-wrote an internationally used textbook called “Biology of Plants.” Since 2010, the garden’s president has been Peter Wyse Jackson, an Irish botanist who has written 14 books and helped develop the United Nations’ Global Strategy for Plant Conservation.

So the garden’s goals are to appreciate, preserve, and study plants, not to breed them. But for pioneering domestication of perennial grains, legumes, and oilseeds, it banks information that Allison J. Miller thinks could be powerfully helpful.

Mohammed Allawala, right, from Austin College in Texas, helps plant seedlings from the genus Ratibida to see if they could be developed as food crops. The Missouri Botanical Garden and Saint Louis University will help The Land Institute canvass the world of perennials for more prospects. Scott Bontz photo.

Mohammed Allawala, right, from Austin College in Texas, helps plant seedlings from the genus Ratibida to see if they could be developed as food crops. The Missouri Botanical Garden and Saint Louis University will help The Land Institute canvass the world of perennials for more prospects. Scott Bontz photo.

Miller is a research associate at the garden. She is also an associate professor of biology at Saint Louis University, a Jesuit school founded in 1818, and the oldest university west of the Mississippi. Miller studies plant evolutionary biology, including how gene variants are distributed and change in populations, and how these apply to crop improvement. She is particularly interested in perennial crops and their wild relatives. She has studied grape, horseradish, pecan, big bluestem, and kudzu.

Miller first met Land Institute scientists in Rome in 2013, at a United Nations Food and Agriculture Organization conference to explore perennial grains. Domesticated annual grains now supply two-thirds of human food calories. Their cultivation causes the loss of millions of tons of soil each year, and damage such as groundwater pollution by wasted fertilizer. If perennial grains were grown in mixtures of species somewhat like prairie or forest, they could vastly surpass annuals in conserving soil and nutrients.

Filling the world’s diverse wild places are plant species numbering in the hundreds of thousands. The most economically important crop plants amount to only a handful or two of species, and the annuals among them command most of the world’s cultivated land. Perennials being developed as an alternative to annual grains are fewer yet: Land Institute scientists guide wheat, intermediate wheatgrass, sorghum, and a few plants from Asteraceae, the sunflower family. Colleagues in China and Sweden pursue perennial rice and barley. Miller wondered about plumbing the deep well of wild perennials and testing the best species as additional candidates for agriculture.

Over the past several decades, taxonomists have amassed large checklists of the world’s plants. The lists are kept at gardens and natural history museums like Missouri and London’s Kew. Already tallied are about 11,000 grasses, 20,000 legumes, and 23,000 Asteraceae. These are the three plant families that produce most of the seeds used as food for humans. About 40 percent of domesticated species are perennial, but none of today’s major grain crops. Close to 60 percent of flowering plants might be perennial. In the tropics it is about 80 percent. For many plants, including sorghum and tomato, being perennial or annual is not black or white. How the plant behaves depends on climate. “We’ll do our best to identify where species fall on that continuum,” Miller said.

Land Institute scientists had thought before about expanding the search for perennials to domesticate. The plan gelled in meetings that brought Miller together with institute President Wes Jackson. The first after Rome was at the US Botanic Garden, where participants discussed how to improve in the public eye something that might seem obvious, but which has been found lacking: the connection of plants and agriculture. The two scientists met again last November, after The Land Institute teamed with the Malone Family Land Preservation Foundation to dramatically boost spending for perennial-grain research.

Now Miller has funding from her school, the garden, and Malone to begin a program for the “global inventory and systematic evaluation of perennial grain, legume, and oilseed species for pre-breeding and domestication.” Miller, Land Institute scientists, and botanists from the garden will use the project not just to find perennials with promise as new food crops, but also to study how perennials compare with annuals under the pressure of selection for crop traits. This might identify principles vital to domestication of perennials.

The project is funded for three years. This is not long enough to find and evaluate every perennial prospect, but few grants are for more than that span. “I think you have to look at this as the first stage of a 20-year project,” Land Institute sunflower scientist David Van Tassel said. “I think you need to look at some of this as proof of concept.” Miller put it this way: “This is a three-year project, but it’s not a three-year project.” The work is new, and as the scientists learn and gather information, priorities might change. Knowledge gained can attract more scientists and power more research.

The Russian scientist Nikolai Vavilov traveled the world during the early 20th century to find where crops originated, and to gather diverse seed for breeding that would improve Soviet agriculture. He focused on traditional crops – annual grains. Now the job is to find perennials. Some might be closely related to existing crops, others might not.

At least to begin the new inventory, there need be no globetrotting. Miller wants to mine literature, databases, herbaria, and living collections. She has an undergraduate student making a trial of technique with one genus, a master’s degree student identifying leguminous candidates, and has hired a post-doctoral researcher who will help manage the three-year study. The team will gather information on the perenniality of species, their form, and where they grow. The scientists will collect seed from the most promising and develop long-term selection experiments. Two of the garden’s ethnobotanists, scientists of traditional plant culture, will contribute information on how people already use the candidate species, and two ecologists will help test germination of collected seed.

The filter for what Miller called this “broad but shallow first pass” will answer these kinds of questions: Is the plant perennial? Is it a grass, legume, or maker of oily seed like sunflower or flax? Is it herbaceous or shrubby, instead of outright woody like a tree? (Trees take years to reproduce, and have been domesticated with transplanted shoots, grafting, and clones, not with the annual or even more frequent breeding that can transform a grass into a grain crop.) This first screening will produce a global checklist of possible candidate plants. Publication will present the names to interested crop scientists.

Next, the researchers will cull their way to a short list of contenders by evaluating the traits of each species: How much seed does the plant produce? How much and what kind of oil and protein are in the seed? Where does the plant range? Can it grow where there is need? Can it grow with other species? How easily can it be found and gathered? Is it easy to grow? How does it adapt to different surroundings? Has it been used for food before? (In an American Journal of Botany article, Miller and other scientists noted that about 10 percent of known plants are edible, and about 7,000 have been cultivated for food.) How does it reproduce? Does it survive freezing? How large and what shape are its flowers, which plant breeders will have to work with? How does it germinate? How will it handle harvest and being made into food?

David Van Tassel tosses from the field bagged seed heads of a promising silphium plant. By such selection he quickly improved the wild perennial’s seed production. He would not stick with a species that responded slowly. In this work even a relatively speedy transformation can consume a scientist’s career. Scott Bontz photo.

David Van Tassel tosses from the field bagged seed heads of a promising silphium plant. By such selection he quickly
improved the wild perennial’s seed production. He would not stick with a species that responded slowly. In this
work even a relatively speedy transformation can consume a scientist’s career. Scott Bontz photo.

Miller is not a plant breeder, and development of criteria for judging will enlist Land Institute scientists. The plants picked should be the ones that are easiest to work with, and that have desirable traits beyond grain or oil yield, Van Tassel said. For examples he gave The Land Institute’s intermediate wheatgrass, which from the start made tasty and workable flour, and food products from which now have a trade name, Kernza®. There is also Asteraceae’s silphium, which, in addition to bearing big seed, can withstand drought.

The Land Institute’s initial inventory and evaluation of plants for domestication three decades ago passed over wheatgrass and silphium, and none of the handful of species selected then enjoy the main attention from breeders now. Eastern gamagrass was attractive as a close relative to maize, but its seed would be tough to make into grain. Perennial giant wild rye reportedly was eaten by Vikings, and was favored over wheatgrass, but it proved to make few seeds.

Meanwhile, Peggy Wagoner at the Rodale Institute evaluated candidates for perennial grain and settled on intermediate wheatgrass. “Peggy made a better choice than I did,” Jackson said. Rodale sent improved wheatgrass for The Land Institute to use in studies crossing perennials with annual crop wheat to make perennial wheat, and scientist Lee DeHaan recognized, by how quickly selection improved wheatgrass, that he had a perennial grain candidate in its own right.

Likewise, silphium originally did not produce many seeds, though they were relatively large. Selection by Van Tassel rapidly improved production. He said that if not for that speedy response, silphium should have been left to the wild.

“There’s sort of a scorecard that you might imagine for each species,” he said. The more good features a plant has, the less work they will require. With enough time and attention, probably any plant can be domesticated. The challenge is to find those that can be domesticated with relative ease and speed, and which would be most marketable.

For that search, Van Tassel argued, it is not enough to look at a row of 20 plants and judge a species by the average performance. Instead, across the variation in 50 times that many plants, seek the best. He grew thousands of perennial sunflowers, which normally have multiple branches, each bearing small heads, and found just one plant that flowered on a single stalk, like annual crop sunflower. From this individual came a breeding population.

Looking at herbarium books and spreadsheets one sees averages, or at best, ranges. Van Tassel is curious to also seek opinions from those who breed forage crops, ornamental plants, and prairie restoration varieties. These are people who have with their select plant species a deep familiarity. “There is some sort of holistic gestalt about a plant,” he said. Asking for this kind of appraisal might sound soft, and its methods hard to scientifically explain. At other times he has opposed it. “But in this case, I’m on the other side.”

Miller said that measures indicating domestication ease – how perennials flower, whether they self-fertilize or outcross, how many chromosomes they have, their seed size – are in the literature available at the garden. To fill holes, researchers will look at plants in the wild, and request seeds from seed banks and the US Department of Agriculture. “It’s hard to imagine a better place,” she said of the garden and its connections. “If it’s known, we’ll be able to get it.”

Jackson said the effort would far surpass the scope of what The Land Institute initially attempted. “This is global,” he said. But though there might be 50,000 herbaceous perennial species, Miller said survey of that number at this stage would be unrealistic. “There’s a limit to what we can do as information and seed gatherers,” she said. “All we can do is accumulate in an organized manner.” Her team will start with species and genera closely related to domesticates. “I say we’re casting a wide net, but to get off the ground we’ll start with subsets of groups.”

Discussion with Land Institute scientists about how to whittle the initial list, and how far, continues. Miller now has in mind arriving at a short list of 100 species, which she called “the tip of a taxonomic iceberg.”

After scientists have this in their grasp, there will begin collection of seed, germination trials in St. Louis, and shipment of seed to The Land Institute for growing out and study. Characteristics that could not be seen in the first pass will then appear. This is the stage called pre-breeding, when scientists know that many of their crosses will lead nowhere, DeHaan said. In breeding of plants already domesticated, the field is more refined, and the progress more incremental. Meanwhile, the botanical garden, which each year hosts about 1 million visitors, will use the candidates to make a display garden of “artificial prairie.”

Experiments at The Land Institute can begin even before the final short list arrives from St. Louis, Van Tassel said. The scientists already know some candidates to test. This year he oversaw planting of two coneflowers, Ratibida columnifera and R. pinnata (see cover photo).

The program will not only seek to decide which species to domesticate, but to understand in general and with practical applications how perennials evolve as they are domesticated, and how their change compares with that of annuals. An annual has no second chance, and must give its all to make seed in one year, while a perennial is a long-term planner. And although a perennial continues to grow and yield seed year after year, it changes not only in size, but also biologically, DeHaan said.

These differences might affect how populations respond when a scientist selects and breeds favored individuals. “Relative to annuals, we don’t know much about how perennial herbaceous plants respond to domestication,” Miller said. They might respond in some ways like herbaceous annuals, in others more like woody perennials. Land Institute scientists have learned from their handful of herbaceous perennials, but each researcher has been out to improve a particular species, not to formally develop and test a theory.

Even how annual crops respond to domestication is not well understood, at least genetically, DeHaan said. The domestication of annuals is now rare. Most plant taming came before science. A few novel crops, such as kiwi fruit, have been recently developed. But Jackson said, “The last time that humans added a major crop to the food inventory was 3,500 years ago.” Stories of how the grain crops came to be are written by archaeology and forensics. “We don’t have very strong support for any of the ideas,” DeHaan said. “We’re not watching what happened. We’re guessing.”

Among other issues that suggest domestication of perennials will pose new challenges is the kind of genetic diversity that comes when a plant must have another plant at hand to reproduce. Most perennials are such out-crossers, while most annual crop plants can self-fertilize. Selffertilization is an extreme form of inbreeding, with the unfit versions of genes quickly revealed, and their carriers culled. With out-crossers, inferior genetic variants from one parent can remain in a population if the other parent contributes a fitter version that is dominant. Although diversity is often a good thing, this kind cannot be afforded for a crop plant, whose populations must perform consistently.

Another theoretical issue for breeding of perennial grain crops is the tradeoff between making seeds for grain and making tissue such as roots so the plant can survive winter. Critics have said extra energy going to roots will keep perennials from matching annuals in seed yield. Land Institute scientists have made the case that not beginning each year from seed gives the perennial a head start. Some perennial trees already equal annual grains in their ratio of fruit to vegetation. But an herbaceous perennial grain, with aboveground growth dying back each year, might require a path different from those of apples or annual wheat, or even those of perennial pasture and hay crops, which have long been domesticated but are bred almost entirely for leaves and stems to feed ruminants, not for seed to feed people.

Miller and Van Tassel propose to grow closely related annual and perennial species at the same time, subject them to the same selection pressures, such as for seed size, and watch and compare how they change genetically and in form. “What I’d hope to see is an increase in seed size in both of them,” Van Tassel said. This might mean that a different domestication path is not necessary. Instead, it would show that just as wild annuals became productive domesticates, so can perennials, even if, because of a perennial’s more conservative life cycle, it doesn’t happen quite as fast.

Such experimentation could also reveal the differences between perennials and annuals in problems like genetic load, the accumulation of inferior genetic variants, and point to solutions. It could show that success like DeHaan’s in increasing seed production from intermediate wheatgrass was not a fluke. It could provide a general approach to successful domestication of perennials, and build support for building the inventory.

This experimental evolution could simultaneously pursue two or more traits, and see how they relate. For example, while one population of a species was selected, generation after generation, for larger seed size, another population could be selected to maximize the amount of carbohydrate stored in roots for winter, and a third for both characters. Does gain in one trait take from the other? How can this be worked through to reach a profitable balance? Land Institute sorghum breeder Stan Cox once said, “The history of plant breeding is the history of overcoming such negative relationships between traits.”

These tests might help answer the general question of whether it is better to select for all or many desired traits at once – plants flowering in sync, large seed, plants not dropping the seed – or select for single traits in different populations and then merge them. This is a question modern breeders of annual grains have never needed to ask, because they aren’t starting from scratch.

A related illustration of the variety of potential in plant genomes comes from University of British Columbia scientist Loren Rieseberg. He found that two species of sunflower, Helianthus petiolaris and H. annuus, gave rise to three hybrids, each favoring an extreme place unfrequented by the parents: H. anomalus in sand dunes, H. deserticola on desert floors, and H. paradoxus in salt marshes. “There isn’t just one outcome [from a crossbreeding],” Van Tassel said, “it depends on how you do the selection.” Human parents might relate.

The Land Institute lacks such extreme environmental diversity, but it has acreage for testing in hot summers and cold winters, one greenhouse and another in the works, and collaborators around the world, including in the tropics, who can help to grow plants. The Missouri Botanical Garden lacks agricultural facilities, but has a trove of information, including more global connections, to start and feed the research. Even together, the organizations won’t provide a combination of perennial grains to solve by 2018 the problems of farming marginal land in a place like sub-Saharan Africa, Miller said. “This is a massive undertaking,” she said. “We need to balance what we could do with no limits, and what we can do in three years.” But in bringing together two groups of people with complementary skills and resources, she finds a powerful and optimistic worldview for work that is good for both agriculture and the earth.

Share On: Share on Facebook Share on Twitter Select other ways to share